Then 25,000 customers showed up to ServiceNow’s annual conference last week, the biggest crowd in the event’s history. They were not there to eulogize the platform; they were there because they needed more structure, not less.

Book a 1:1 call with me

What ServiceNow’s executives described from that conference was not a world where AI had simplified enterprise operations; rather, it was the opposite.

A CIO had 900 AI pilots running across his organization and canceled every one of them, because nobody owned them and he couldn’t tell what any of them were actually doing.

And a CTO at a major bank built 30 production-grade AI agents and couldn’t put a single one into production because he couldn’t answer basic questions about what they had access to. ServiceNow’s president described it plainly as AI chaos, and the room full of analysts went quiet.

This is what happens when you adopt fast and govern never. And the response from those enterprises was to go looking for systems with guardrails, with accountability, with humans in the loop who could actually be responsible for outcomes. The most sophisticated technology buyers in the world, after an 18-month experiment with ungoverned AI, came to the same conclusion: they want defined systems and expert people involved.

I find this clarifying rather than surprising. Our leadership trio (at FOMO.ai) has seen many cases where businesses rely on data, but the data they actually have is appalling. So what happens? A data scientist translates the ask and builds a response that includes an explanation and caveats, which, in turn, should facilitate a human judgment call.

But data scientists have looked like easy targets for layoffs over the last 2 years, and companies have built AI agents to sit on top of that same data instead. Anyone can now ask & receive data analysis back… the problems, of course, are (1) they do not have the training or understanding of how the data might be misleading, and (2) they put more trust in a response from an AI than they do their former human colleagues.

The SMB version of the CIO with the 900-pilot problem looks different on the surface. There’s no CIO, no compliance department, no formal audit process. But there are three ChatGPT subscriptions across a team of five, an AI chatbot on the website that nobody has reviewed in months, social content going out that the founder hasn’t read, and an AI SEO platform generating reports that look impressive and that nobody fully understands. The tools are everywhere, and the ownership is nowhere, which is the same problem the enterprises had, just without anyone whose job it is to notice.

What those enterprises discovered, and what I think SMBs are starting to feel, even if they can’t name it yet, is that the value was never in the raw AI output; it was always in the judgment layered on top of it. I’ve written before about how Google’s Performance Max, left to run without human intelligence guiding it, optimizes toward the average across every dimension that matters. The same logic applies to any AI system operating without domain expertise sitting above it. It produces output that is technically functional and strategically mediocre, because it has no way to know the difference.

The enterprises tried to skip that layer and spent the last year cleaning up the results. The platforms they’re now paying to fix the problem, the governance layers, the control towers, the context engines built on proprietary data, are essentially just structured ways of putting expert humans back in charge of what the AI does. ServiceNow built an entire product around this, and the commercial response, according to their own executives, was faster than anything they expected.

For smaller businesses, the solution doesn’t require enterprise software; it just requires the same underlying principle: someone who knows the business, understands the customer, and can tell the difference between output that cleared the technical bar and output that’s actually effective. That gap is where most of the value in marketing lives, and it’s exactly where ungoverned AI falls short every time, regardless of the size of the organization running it.

SaaS isn’t dying; it is growing into an outcome-based system (which Sequoia just called the $Trillion company) rather than a point solution. The local attorney really doesn’t care about the underlying SaaS, and they certainly don’t want just another software subscription that they have to give headspace to; they just want sales & leads, and to know that is coming from a system they can trust and people they can hold accountable.

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.

The moment the Soviet Union made clear it wasn’t going to the moon, America stopped going too. The threat dissolved, and so did the forcing function.

I’ve been thinking about that a lot this week while listening to The Daily’s interview with Keith Bradsher, the NYT’s Beijing bureau chief who has been covering China’s economy since 1991.

The piece is framed as a tariff story, but it isn’t really. It’s more of a story about what happens when you poke something and discover it’s a lot bigger and more solid than you thought.

About a year into Trump’s trade war, China’s overall trade surplus hit $1.2 trillion, bigger than the entire economy of most countries on earth. The tariffs shifted trade away from direct shipping to the US and toward more indirect routes, weakened its currency to make its goods cheaper everywhere else, and kept exporting components indirectly through third countries that assembled them before they reached American shores. The US plan sounded like it was trying to slow their machine, but it seems the machine just found new roads.

And the reason the machine is so hard to stop comes down to something that was years in the making, for reasons that had nothing to do with America at all.

By the mid-2010s, China had a problem. The one-child policy had quietly gutted its future labor supply. Young Chinese workers, the only children of parents who poured everything into their education, were graduating from college in record numbers and had absolutely no interest in factory work. There’s also a deep Confucian tradition that manual work is inferior to intellectual work, and with one child per family, parents weren’t about to let their child spend their life on an assembly line.

Factories were expanding, but bodies weren’t showing up.

So China automated. Not as a strategic flex, but as a survival response. And they did it at a scale and speed that now looks, in hindsight, like a master class. China today installs more industrial robots each year than the entire rest of the world combined. Germany, Japan, the United States — all now have fewer robots per 10,000 manufacturing workers than China. Countries that built their identities around industrial precision were outpaced by a nation that automated out of demographic necessity.

The factories Bradsher visited aren’t just robotic, either; they’re AI-integrated (you knew I was going to bring it back to AI eventually!). Cameras photograph every finished product and compare it with reference databases to flag flaws. AI tracks practically every step of the process. The result is that pretty much anything can now be made less expensively in China than anywhere else, and not just clothing and furniture, but cars, batteries, solar panels, and advanced technology products. The cheap labor narrative feels years out of date, given this is cheap AND advanced.

The tariffs, in other words, arrived at a party that China had already won. And by revealing just how dominant China’s manufacturing base had become, they arguably did more to expose Western industrial decline than to reverse it.

For founders and business owners, I think the lesson here isn’t really about China or tariffs. It’s about what forced necessity produces. China didn’t see automation as an opportunity. It saw it as the only available answer to a problem — no workers, no choice — and it committed completely. The result is a decade-long lead that trade policy alone cannot close.

We’ve built FOMO.ai on exactly this principle, and it’s shaped how we operate more than any strategy session ever has.

1. We (transparently) sell an idea to a client before we’ve fully solved it.

2. That commitment creates the forcing function.

3. The team now has a real problem, a real client, and a real deadline

What comes out the other side is invariably better than anything we would have built in a vacuum, whether it be a new process, a new product, or new learning that we carry into everything that follows. The discomfort of the commitment is the point! It’s what stops you from endlessly refining something in theory and never shipping it in reality.

China did this at a national scale. The one-child policy created a crisis, the crisis created a forcing function, and the forcing function produced the most automated manufacturing base on earth. They didn’t plan to be decades ahead; they allowed themselves to be pushed there.

The question worth sitting with is: what is your forcing function going to be? Because if you wait for the threat to arrive before you start building, whether that’s a competitor, a market shift, or a client you’ve already promised something to, you’ll be in the same position as the countries staring at China’s trade surplus right now. Technically still in the game, but playing catch-up on a lead that was built years before the pressure arrived.

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.

The headline is genuinely good news. Total capital raised hit nearly $120 billion in 2025, up 17% from the year before. Q4 alone brought in $36.1 billion, the strongest single quarter since 2022. Down rounds fell below 14%, their lowest rate in three years. By almost every surface measure, the market that reset hard in 2022 has found its footing again.

But the composition underneath that headline is where things get interesting.

Subscribe now

The numbers getting overlooked

Nearly half of all seed financings in 2025 were bridge rounds, meaning extensions of existing rounds rather than new ones. Companies are buying more time because the milestones needed to unlock a Series A have not yet arrived. This gets framed in most coverage as a sign of founder-friendly flexibility, and in some ways it is. But it also tells you something honest about the shape of the market: the bar between stages has risen considerably, and a lot of companies are doing the sensible thing by extending runway while they clear it. The time from seed to Series A now sits at a median of 2.1 years, up from 1.5 years five years ago. Series A to B takes 2.8 years on average, the longest interval on record. These are longer journeys with higher checkpoints, and founders who plan their capital accordingly are the ones navigating it well.

The AI concentration data is significant and worth being clear-eyed about if you are building outside of it. At Series E, AI companies commanded valuations 193% higher than non-AI peers, and captured 70% of all capital at that stage. A third of all investment from Series A through E went to AI companies. This is not a rising tide lifting all boats. It is a very focused beam, and most of the market is not standing in it.

The part that almost nobody seems to be discussing, and I find it the most useful data point in the entire report, is what happens to that AI premium at the seed stage; it nearly disappears. The gap, which reaches 193% by Series E, is close to zero at the beginning. The premium is being assigned as companies demonstrate traction and revenue, not handed out at inception because a founder put AI in their deck. Which is actually healthy, and also a useful signal for anyone trying to read where genuine value is being created versus where excitement is running ahead of evidence.

What this has to do with how you raise

About eleven years ago I was co-founder of an ad tech company, and raising money meant institutional meetings, lawyers, and more process than product. We needed millions up front just to have basic technology to compete. The priced round was not really a choice; it was a structural requirement of the moment. You absorbed the dilution, handed over more control than you felt comfortable with, and assumed this was simply the cost of building something real.

That assumption no longer holds for most founders, and the Carta data makes the case more clearly than I could from personal experience alone.

At FOMO.ai, every dollar we have raised came through SAFE notes with a post-money valuation cap. We started at a $7.5 million cap and raised it as we hit milestones and de-risked the business, moving to $12.5 million, then $15 million, then $20 million. We generated revenue within 60 days of launching. The cap table stayed clean, showing only the founders and a small cluster of core team members, and when institutional investors eventually came around, that cleanliness was the first thing they noticed. It is nearly impossible to maintain once you go down the priced round path early.

Read: Founders! Are You Sure You Need a Priced Round?

The short version is this: the post-money SAFE structure meant our investors were not diluting each other across multiple rounds, our legal overhead stayed close to nothing at the moment when time and money were most precious, and we kept control of the timing of any conversion rather than surrendering it to a forced date.

What I see in the Carta data is a market that has become very good at concentrating capital into a smaller number of larger bets, while the number of companies moving cleanly through the traditional Series A, B, C path keeps compressing. The founders doing well inside this environment are not necessarily the ones grinding through the old playbook. Many of them are the ones who figured out earlier that the priced round is often a solution to a problem that no longer exists in the form it once did, at least for companies whose ambition and capital needs sit somewhere south of the unicorn track.

If your total capital need is several million rather than tens of millions, if you are building toward a strong acquisition or exit rather than a public offering, if scale is not genuinely do-or-die in your market, then the dilution and control you hand over in a traditional priced round is a real cost with no guaranteed corresponding benefit. The Carta data shows the market bifurcating sharply between the companies that genuinely need that path and the ones that simply assumed they did because everyone around them was doing it and celebrating it online.

The comeback is real. The numbers prove it. But the invitation list is considerably shorter than the headline suggests, and the path that actually gets companies funded increasingly looks different from the one most founders are still being told to follow. That gap is worth understanding before you decide which road you are on.

Share

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.

Subscribe now

I’ve never believed that.

Maybe it’s because I’ve always been drawn to companies that put empathy front and center. But there’s a specific phrase from Disney that stuck with me and has come to help shape how we try to do things at FOMO.ai.

Technology companies like ours cannot afford to shrug off issues as “someone else’s fault.” If a customer is having a lousy time, even if something outside our control is to blame, we can choose to treat it as our problem to solve.

Where tech companies get it so wrong

Most investors in SaaS want a ruthlessly efficient model: strip out all the people, automate everything, and maximize margins. For years, the playbook has been about optimizing out the human element, and with the rise of AI, there is a lot of salivation.

But the reality on the ground looks very different. AI is changing at breakneck speed. The industry has massively oversold what its tools can deliver, especially the promise of “just sign up, and it works.” The result? People burned by bad experiences and underwhelming tools are coming back and asking for genuine support. They don’t want to feel left alone with a black box. And I actually agree with them.

I keep coming back to that Disney approach. When a child gets to the front of a 2-hour line and is told they aren’t tall enough to ride, instead of telling them it was their fault they didn’t use the measuring stick when they joined the line, they hand them a voucher to skip the line when they come back in 6 months’ time, a little bit taller. That’s what sets them apart. I want FOMO.ai to do the same at scale, even (maybe especially) when things go wrong for reasons outside our direct control.

How we use humans, and why it matters

We’re a technology company, and we’re always going to use technology to drive efficiency where it makes sense, but purposefully including humans in our processes isn’t just a “nice to have”, it’s become a core part of how we operate.

Our approach is pretty simple: we bring humans into the loop expressly to solve problems, answer questions, or provide reassurance when needed, much like that Disney cast member who can turn a small disaster into a memorable save.

Here’s the key: by focusing our human touchpoints where they have the most impact, we keep our margins healthy (higher than most traditional SaaS companies), avoid staffing bloat, and give people the kind of positive experience that builds real loyalty and trust. It’s both strategic and personal for me.

The idea isn’t to hold everyone’s hand through every click and keystroke. What we aim for is targeted intervention when someone’s stuck, frustrated, or simply needs to know that an actual person cares about getting them back on track. The result (and I’ve seen this time and again) is that the blend of human support and automation outperforms either approach in isolation.

A real example: when “not our fault” becomes “our problem” (and opportunity)

Let me give you a real story from the trenches. One of the most common things we do at FOMO.ai is help our clients generate sales and leads through AI search optimization. We’re deeply technical about it, but there are always peripheral factors that can drag results down, things that aren’t technically inside our scope.

Sometimes, a client comes to us with a situation like this: Their CEO just announced, out of the blue, “We’re launching a new service. Put it on a subdomain and get it live this week.” Suddenly, what should have been a straightforward optimization now has new technical complexity and knowledge gaps. They’re lost, stressed, sometimes even a little panicked.

Now, from a purist’s SaaS perspective, this is not our fault, this is client-side confusion, a lack of preparation. But I keep coming back to our differentiator: “If we want to be the trusted source, the company they really rely on, we have to step in and own the problem.”

Sometimes that means jumping into a private Slack channel, other times it’s a quick Google Meet to walk them through the solution. Yes, it takes extra time, but what we gain is both ethical, a sense that we actually care, and strategic: those are moments when the client’s trust goes way up and ultimately, so does our impact on their results.

James Dyson was the same. It is widely reported he once became angry in a meeting where he was been presented with a plan to optimize away all client calls. His attitude was that it’s a privelege to have a conversation with your client, even if the initial reason is negative.

And it’s not theoretical. Every time we help someone through those edge cases, we see meaningful improvements in customer retention and value. Our retention rate is significantly higher than most of the industry, and I credit much of that to consistently taking responsibility, even when it’s messy.

Human touch vs. efficiency: finding the line

I won’t pretend we’ve solved this balancing act perfectly. There’s a real tension here, and it’s something we wrestle with regularly. Having humans in the loop introduces the risk of scope creep. If every customer request and every unexpected issue leads to more and more real-time involvement, the costs can spiral.

At this point, I think we’re doing a good job. Our margins are healthy, customer satisfaction is strong, and we’re not burying ourselves in unscalable overhead. Still, it’s something I keep a close eye on. The trick is to build processes that foster empathy and support without breaking the model that keeps us growing.

What I’ve learned is that rigidity kills. If you set out to make every aspect of your product strictly “not our fault,” you not only make life worse for customers, you actually miss key signals about where your service is falling short. Conversely, if you over-index toward “solve every problem for everyone,” you lose focus and profitability. We’re all still figuring out where the line is, but I’m convinced that erring on the side of human empathy beats optimizing them out.

Why empathy is one of the only moats that last

I’ve heard the argument a thousand times: just keep automating, and soon, every AI company will run on autopilot. My worry? If that’s all we focus on, we make ourselves obsolete. Today, the differentiator is specialist knowledge, service, and trust. Maybe in ten years, the AI can handle absolutely everything, but even then, the companies that build relationships and offer empathy will be the ones left standing.

As AI gets better at mimicking needs, people’s desire for real human validation and reassurance only grows. Whenever I see a company that’s gone full “baseless,” all automation and no personality, I see a company waiting to be replaced by the next slightly better algorithm.

For me, this Disney quote isn’t just about good service; it’s a strategy for long-term survival. The more you lean into humanity and accountability, the stickier your customer relationships become. That’s a great extra moat to have.

What you can use today

I’ve learned a few things that I think apply well beyond my corner of the AI world. If you’re building anything, tech or otherwise, here’s what I’ve found helpful:

• Treat every user frustration, even if caused by something “external,” as your chance to shine. Pick the moments to step up and solve it anyway.

• Use human intervention not as a cover for failure, but as a force multiplier for success. Focus on moments where empathy and expertise matter.

• Stay vigilant about scope creep. Offer targeted support, not boundless hand-holding. Build feedback loops to know where you’re adding value.

• Measure results by retention and loyalty, not just short-term efficiency. The best customer is the one who sticks around when things get tough. All 10 of our first ever 10 clients over 2 years ago are still with us today, as are 90% of all the others.

• Most crucial: never let technology become your only identity. Bring personality, humor, and a sense of reassurance to every interaction.

Responsibility doesn’t stop at the edge of your product. If something goes wrong, whether or not it’s your fault, making it your problem to solve is the fastest way to build trust, loyalty, and long-term advantage. Empathy is not a cost center; it’s your biggest strategic asset in a world racing toward faceless automation.

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.

Instead of building software tools for marketers, we built a marketing team made of AI, not a platform or a copilot, but something that does the work and delivers the outcome the same way an agency would, except without the overhead built around human labor costs.

Share

Last week, Sequoia Capital partner Julien Bek published an essay that essentially described what we’ve been building. He called it “Services: The New Software,” and his central argument is that the next trillion-dollar company won’t sell software at all, but will sell the work itself. The distinction matters enormously when you think about where the real money sits. Businesses spend roughly six dollars on services for every one dollar they spend on software, which means every AI tool and SaaS copilot competing for attention right now is chasing the much smaller pile. Whereas the model that follows the labor budget is playing a completely different, much larger game.

What shaped how we built FOMO.ai was that the brands we most wanted to work with weren’t shopping for tools at all; they just wanted leads and sales driven by AI Search!

So we positioned ourselves as a marketing team, one that works for your company the way an agency would, except the underlying cost of doing the work keeps falling as the models improve, which means the service gets better and cheaper over time rather than more expensive. And unlike a fully automated system, we keep humans in the loop by design, because sometimes that’s simply practical (for now), and sometimes a client just prefers knowing a person was part of the process. (See: “Analysts are saying ‘we are cooked‘ but human behavior has something to add.”)

Building a model that can flex between the two turns out to be a genuine competitive advantage rather than a compromise. The shift felt counterintuitive in some investor conversations, where everyone was talking about how software multiples couldn’t be achieved with humans in the process. Two years later, now it’s what many investors are actively looking for.

Bek maps out the sectors he believes are ripest for this kind of disruption, and the pattern he identifies is consistent: work that is already being outsourced, heavily reliant on processing and pattern matching rather than deep contextual judgment, and where the buyer has always been purchasing an outcome rather than a relationship. These are categories where AI can now perform the core task at a level that is genuinely hard to distinguish from a professional doing it manually, and where the structural case for replacing an outsourcing contract with an AI native provider is far simpler than the case for replacing internal headcount. Marketing fits that description perfectly, and we’ve been stress testing it long enough with real brands to know where it works well.

The production side, the content, the optimization, and the distribution across channels are largely solved problems at this point. What remains is a 1:1 human connection when it is wanted, and when it is needed to give the client the reassurance that what we’re doing is the right thing, to answer questions, to help with adoption and change management… and when only a human will do.

The part of Bek’s argument I find most compelling, and most urgent for anyone thinking about building in this space, is what happens to client relationships over time. When a brand embeds an AI team into its marketing operations, the dependency that develops is operational rather than transactional, and it compounds in a way that a software subscription simply doesn’t. The brand training, the quality benchmarks, the institutional knowledge about what works for that specific audience, all of it deepens with every campaign. And because the value being delivered improves as the underlying models improve, every new capability release becomes a tailwind rather than a threat. That’s the complete opposite of what it feels like to be selling tools right now, where the next model version might quietly make your product redundant.

Our clients sometimes call us an AI-native agency (as does Y Combinator), an SEO agency, or even a performance agency because that framing maps onto something they already understand and trust. When I talk to investors, I lean into the tech side of what we’ve built, because the margins we’re driving are SaaS+ margins, and that’s rightfully what serious investors want to be putting money into. Both descriptions are accurate, and the tension between them is actually a pretty good illustration of why this model is so interesting. We’ve built the underlying technology and infrastructure of a software company, but we go to market as something that feels to clients like an agency that happens to be better, faster, and cheaper than anything they’ve worked with before.

What genuinely excites me about where we sit right now is the optionality we’ve built. Having proven the model works, we can move in any direction from here, vertically into specific industries, owning the full marketing stack for brands in those sectors, or horizontally across every marketing service that follows the same pattern, where the outcome is measurable, the work is intelligence-heavy, and the buyer cares about results rather than process. AI Search is where we started because it’s where the biggest shift in buyer behavior is happening and where the case for an autopilot model is most obvious, but the infrastructure we’ve built spans channels and categories in a way a single-point solution never could. The category doesn’t have a single, clean name yet, and we’re still navigating the same language question that every company in this space faces. But the underlying logic is sound, the data confirms it every week, and there’s something genuinely satisfying about watching one of the most respected firms in venture capital arrive at the same conclusion from a very different direction.

Share

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.

Either the robots are coming for everything, and we should panic, or the disruption is overblown, and we should relax.

Anthropic published research this week that sits uncomfortably in between those two positions, and I think that discomfort is actually the most useful thing about it. They built something called an "observed exposure index", that measures not just what AI could theoretically do to a given profession, but what people are actually using it for right now in real professional settings.

The gap between those two numbers is enormous, and depending on how you read it, that gap is either deeply reassuring or the most alarming part of the whole picture.

Language models could theoretically speed up 94% of all computer and mathematical tasks. In practice, only 33% of those tasks are currently being handled based on actual usage data. Office and Admin sits at 25% real-world coverage against 90% theoretical. Business and Financial at 20% against 85%. The popular framing is 'we are cooked.' The more accurate framing is that we are somewhere in the early chapters of a long book, and the red will keep growing.

Subscribe now

The paper names the scenario everyone in the knowledge economy should be thinking about: a “Great Recession for white-collar workers,” drawing a comparison to the 2007-2009 financial crisis when US unemployment doubled from 5% to 10%. So the alarm isn’t wrong, exactly, but the chart is only telling you what’s technically possible. It says nothing about what people will actually choose.

I wrote something on 84Futures that I keep coming back to in light of this research. The premise was that as automation reshaped the workforce, something unexpected happened: people began longing for the jobs they had left behind. Not for the paycheck or the grind, but for the small, often overlooked human interactions that gave work meaning. Doctors whose value shifted entirely to empathy and presence. In-person sales that came back because trust had become a premium in a world full of deepfakes. Craftspeople whose handmade goods became luxury items precisely because a machine could replicate them perfectly. I wrote it as near-future speculation, but the Anthropic data makes it feel much less fictional.

Workers in the most exposed occupations are more often female, better educated, and earn on average 47% more than the unexposed group. This wave is hitting knowledge workers first, not the trades. The person who graduated with debt and went into finance, legal, or marketing is more exposed than the electrician or the plumber. And yet the research also shows that workers in the most exposed occupations have not become unemployed at meaningfully higher rates than workers in jobs considered AI-proof, at least not yet. The disruption is happening at the entry point, not across the whole workforce simultaneously. Young workers aged 22 to 25 are were being hired less frequently (read: “Companies are reversing course on hiring graduates”) in professions with high AI exposure, suggesting that the technology is already reshaping entry-level employment.

That’s where the data ends, but it’s where the human question picks up, and it’s the one I find more interesting: when AI can do something, does that mean people will let it?

The 84Futures piece argued that nostalgia for human connection isn’t really about resisting technology as the Luddites did in industrial Britain; it’s about wanting to believe that a human was still necessary somewhere in the process. That instinct is going to shape which parts of the blue bars actually become red, and which ones stay theoretical forever because we collectively decide we’d rather pay more for the human version.

Which brings me to a parallel conversation I’ve been watching about a job title Jason Calacanis, one of Uber’s earliest investors, recently named. He called it the “agent maestro”: the person who can take a business process, explain it, and train an agent to execute it. Not a developer, not someone who codes, but someone who understands how a business works and can break that understanding into instructions an agent can follow and then iterate on over time. David Sachs added the broader frame that every company adopting new technology faces a massive change management problem, and the person who can lead that adaptation will have an extraordinary career ahead of them.

At FOMO.ai, we’re living this. The people who have become indispensable on my team aren’t necessarily the ones who understand the technology most deeply. They’re the ones who understand the workflow, can hand it to an agent or the dev team, and then catch where it’s not aligned to their vision. That skill sits right at the intersection of the Anthropic data and the nostalgia thesis: human judgment applied to machine output, which turns out to be exactly what neither a blue bar nor a red bar can fully capture.

Roughly 30% of workers have no AI exposure, including those in fields that require physical presence and that no language model can replicate. For everyone else, the question is whether you end up designing the system, being absorbed by it, or being chosen because of nostalgia. The chart shows you the territory, but what it can’t tell you is how much of that territory humans will actually surrender, and how much they’ll decide, quietly and collectively, they’d rather keep for themselves.

Subscribe now

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.

Subscribe now

The data confirms what I witnessed in real time. Only 30% of 2025 graduates secured a full-time job in their field, down from 41% the year before. Unemployment for young college graduates in the US hit 9.5% by September 2025, nearly double the general adult rate. Entry-level knowledge-work roles are vanishing, especially in fields most exposed to AI, with postings down roughly 35% since January 2023. The tasks AI does best happen to be exactly the tasks companies used to hand to their newest hires, and a lot of employers made a clean economic decision to stop making those hires.

They’re starting to regret it.

The companies that spent 2023 and 2024 cutting graduate intake, leaning on AI tools, and pushing junior work down onto senior teams are hitting a wall. Those tasks didn’t disappear when the junior hires did, they got absorbed by stretched senior employees, with the assumption that AI would fill the gap. Senior staff are burning out, and when they leave, the thinking is that AI will cover it. That logic looks clean in a restructuring deck, but in practice, it is producing flattened teams with little agency, endless rework cycles, and exhausted people juggling work at every level at once.

Here’s what’s becoming clear to the smarter operators. Graduates sitting on the outside of this market right now are not the liability that recent hiring decisions implied. For a company that has genuinely committed to AI, they may be the most interesting people to hire. Younger workers are far more AI-ready than their senior counterparts, learn faster, improve productivity more quickly, and cost significantly less. They haven’t spent a decade building workflows that need dismantling, they approach the tools without baggage, without the mental overhead of comparing everything to how things used to be done. The AWS CEO called replacing junior hires with AI one of the dumbest things he’d ever heard, precisely because graduates are the cheapest and most enthusiastically tool-native part of any workforce.

These two kids happen to be in England, which adds its own layer of difficulty. A lethally weakened post-Brexit economy, still absorbing consequences that the people who voted for it promised would never materialise, compounded every one of these trends, but the core dynamic is the same whether you’re graduating in London or New York.

This is shaping how I think about building our team. Some of the most efficient, most capable people we have are the ones shaping our intelligence, our ways of working, the unique insights that get clients into the top 20% of outcomes rather than the middle of the pack. The AI executes at scale, but the edge comes entirely from the human layer on top of it.

It’s a bit like Google’s Performance Max. Turning PMax on will find the middle of the market and settle there, optimising toward the average because that’s what’s in Google’s best interests! The brands getting exceptional results are the ones feeding it superior audience signals, sharper creative thinking, and better first-party data. The AI amplifies whatever intelligence you put in. If your ambition is average, there are easier ways to get there. If you want to lead, the quality of the human thinking wrapped around the AI is the difference.

That’s what makes this generation of graduates genuinely interesting as a talent pool. They’re not intimidated by the tools, and they haven’t calcified around ways of working that predate them. Paired with the right mentorship and a team that knows what good actually looks like, they can reach high-quality output faster than most hiring managers are currently willing to believe.

The job market right now is genuinely hard, and I’m not going to pretend otherwise for the people living it, but the companies worth working for are the ones figuring out that the answer to AI isn’t fewer humans, it’s better-positioned ones.

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.

If you’re not familiar, it’s the grueling process London black cab drivers go through to memorise 25,000 streets, every alley, every shortcut, every one-way system in one of the most complex cities on earth. It takes most people three to four years of riding around on a moped with a clipboard, and the dropout rate is enormous. What’s truly wild is that brain scans of cabbies who complete it show their hippocampus, the part of the brain responsible for spatial memory, is physically larger than average. The Knowledge literally reshapes their brains. Forever.

Subscribe now

I’ve always found that fascinating, and it made what happened next feel even more brutal. GPS and then Uber automated the expert part, the hard part, the thing that took years to learn and physically altered your brain. Suddenly, anyone with a car and a phone could compete. Employment in ride services went up 250 percent, but wages stayed flat because the hard part was gone, and anyone could do what was left. All those years of preparation, all that accumulated expertise, made redundant by a free app on a smartphone.

That story sits at the heart of something I think most of the AI conversation is getting badly wrong. The question people are asking is “will AI take my job?” and it’s too blunt and too binary to be useful. The question that actually matters is whether AI is taking the hard parts of your job or the easy parts, because those two scenarios lead to completely different outcomes.

We assumed, for a long time, that automation would work its way up from the bottom, taking the routine and repetitive work first and leaving the complex, expert work alone. And for a while, that was largely true. But AI isn’t respecting that old order. It’s coming from the bottom and the top simultaneously, automating the entry-level drudgery and the high-skill expertise at the same time, and increasingly leaving pressure on the people in the middle, the experienced practitioners whose value was always somewhere between pure routine and pure judgment.

To understand where you sit in that picture, it helps to look at what’s already happened to other professions.

When AI takes the hard parts, the expertise evaporates

Stock photo photographers spent years developing an eye, building libraries, and mastering light and composition. Then platforms commoditised the catalogue, and generative AI finished the job. The expertise didn’t become more valuable when automation arrived; it became irrelevant because the machine could produce the output without the craft behind it.

Travel agents carried deep destination knowledge, relationship networks built over careers, years of accumulated understanding about routing and visas and the right hotel for the right traveller. Search engines automated exactly that expertise and left behind a much smaller market of luxury and complex itinerary specialists, serving the clients whose needs still required a human.

Radiologists spent a decade learning to read scans. AI is now matching or exceeding human accuracy on certain image types, pressing directly on the hardest and most trained part of the job. The profession is not disappearing, but the nature of its value is shifting in ways that are genuinely uncomfortable and unresolved.

The cabbie is the clearest version of this story. When the hard part goes, competition floods in, wages flatten, and the years you spent becoming good at something count for far less than they used to.

When AI takes the easy parts, the human judgment becomes more valuable

There is a 1960s economic idea that almost never comes up in the AI conversation, and I think it belongs right at the centre of it. An economist named William Baumol observed that when technology makes some sectors wildly more productive, the sectors it cannot touch tend to get more expensive, not because they got better, but because everything around them got cheaper.

Personal trainers are a good example. Nothing about motivating a human body to move has gotten more efficient. But because everything around fitness got cheaper, the apps, the equipment, the YouTube workouts, the in-person trainer who actually knows you and keeps you honest has become more expensive, not less, because their time became the scarce thing.

Live music tells the same story. Recorded music got infinitely cheap, so the live experience became more valuable than ever. The scarcity shifted.

Skilled tradespeople, plumbers, and electricians, and the people who have to physically show up and solve a problem inside your specific wall, have seen their rates climb steadily for decades while software ate everything around them. AI does nothing to change that trajectory.

And the most interesting emerging version of this might be certain kinds of therapists and executive coaches. AI can now do a reasonable job of reflective listening and structured prompting. But the practitioner who has sat with 2,000 humans over 20 years and can read what someone isn’t saying out loud? That market is pricing upward because the thing AI cannot replicate is becoming the only thing that matters.

The teacher version of this stays with me. AI can teach calculus, probably better than most tutors. But the teacher who notices that a student stopped making eye contact three weeks ago and picks up the phone? That is not a productivity problem. That is a different kind of work entirely, and it is getting more valuable by the month.

The question that now matters most

For most of us working in marketing and business, the honest exercise is to look at last week and ask which parts of your work AI could already handle, and whether those were the parts that took you years to develop or the parts you could teach a new hire on their first morning. Because if AI is absorbing the drudge work and leaving you with the judgment and the relationships and the strategic calls, the economics are moving in your favour. The risk is in standing still while the nature of your value shifts beneath you, assuming that what made you good five years ago is still what makes you valuable today.

The people who come out ahead will be the ones who understand which parts of their work are being commoditised and which are becoming more expensive and more scarce, and who have the honesty and curiosity to keep moving toward the latter.

Share

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.

They were wrong, obviously.

Subscribe now

The internet did not stop getting better. Not for one single year after that crash. Email went from clunky to instant. Video went from buffering nightmares to streaming in your pocket. Connectivity got faster and cheaper every year. Online shopping went from a novelty to how most people buy things. We got maps that knew where we were, tools that let anyone build a website, video calls that replaced flights, and eventually, entire businesses that ran from a laptop on a kitchen table. The bubble popped, but the technology kept compounding, quietly and relentlessly, no matter what the stock market did.

I think about this constantly right now, because the AI bubble conversation is everywhere, and the numbers people are throwing around are genuinely wild.

Gartner says worldwide AI spending will hit $2.5 trillion this year. OpenAI has committed $1.4 trillion over eight years on data centers while generating about $13 billion in revenue. Meanwhile, a February 2026 study by the National Bureau of Economic Research found that 90 percent of firms reported zero measurable impact of AI on productivity. MIT said 95 percent of organizations investing in generative AI were getting zero return. Software stocks have been hammered. Oracle and Microsoft are down 22% and 14% this year, respectively.

So is AI overhyped? Are we in a bubble?

We can debate public company hardware capitalization rates all day long, but the whole question misses the point! The AI beneath all that noise is on the exact same trajectory as the internet after 2000, just much faster. Models are getting dramatically more capable with every release. Costs are plummeting. DeepSeek and Kimi proved that frontier performance is achievable at a fraction of the expected cost. And this month, Anthropic released a model so capable that software stocks cratered because investors feared AI might disrupt the entire enterprise software sector.

The fear is no longer that AI does not work; it’s that it might work too well!

Look at the human adoption

Here is what gets lost in the Wall Street noise: 800 million people use ChatGPT. 750 million use Google’s Gemini. They are planning meals, helping kids with homework, drafting emails, researching health questions, learning new skills, and a thousand other things that have nothing to do with data center capex or enterprise ROI. This thing only went mainstream 2 years ago!

And it keeps pushing further. OpenClaw, an open source AI agent that started as a weekend project a few weeks ago, already has over 145,000 GitHub stars and is being adopted from Silicon Valley to China. It runs on your own machine, connects to WhatsApp and Telegram, and manages your email, calendar, and web browsing autonomously. One developer called it “what Siri should have been, built by one guy and a repo instead of a $3.6 trillion company.” Too early for mainstream, but the trajectory is clear. Personal AI agents that actually do things for you are arriving fast.

Nobody went back to paper maps after Google Maps. Nobody went back to encyclopedias after Wikipedia. You cannot unadopt a technology that hundreds of millions of people have already woven into their daily lives, (although the concept of nostalgic jobs will be very real). The stocks can do whatever the stocks do. That part is irrelevant to the underlying momentum.

Some AI startups will struggle

We already saw this in 2025, when ChatGPT wrapper startups built on commoditized models started to drop. Builder.ai, once valued at $1.2 billion, filed for bankruptcy. Dozens more quietly disappeared.

The companies that survive will be the ones that build real vertical businesses, not thin layers on top of someone else’s API. At FOMO.ai, we took an entire industry, marketing services, and rebuilt it as a technology platform with humans layered in for quality and client confidence. We generate real revenue from real companies, we have established trust, we have solved for distribution, and we have a moat of years of human experience studied and translated into our DNA. That is a fundamentally different business from a startup that raised $20 million to put a nice interface on GPT and hoped for the best.

Bet on the Founder more than the model

Here is what I keep coming back to. Founders are uniquely suited to this moment, and I do not think enough people are saying it loudly enough.

You have complete control over your tools, your stack, and your processes. You do not need to submit a proposal to a steering committee or wait for IT to approve a new vendor. You do not need to run a six-month pilot program or build a business case for the CFO. You can just try things. Today. Right now. And that advantage is enormous, because the founders who are tinkering with these tools constantly, testing new models, integrating AI into their workflows, rebuilding how they operate every few months as the capabilities improve, those founders are building a compounding advantage that will be nearly impossible to catch later.

I see this in my own company every week. At FOMO.ai, we are constantly reworking how we use these tools, not because we are chasing shiny objects but because the models genuinely get meaningfully better on a regular basis, and the founders who are not keeping pace with that are going to get left behind. The gap between a founder who adopted AI eighteen months ago and kept iterating, and a founder who is still “waiting to see how this plays out,” is already enormous. In another year, it will be insurmountable.

You do not need to wait until you have a massive budget to start. You need curiosity and a willingness to keep showing up. Pick one or two tools, learn them deeply, and when the next model drops or the next capability unlocks, be the founder who is already there experimenting on day one. That is the muscle that separates the companies that thrive from the ones that vanish, and it has always been the muscle that matters most. We are wired to just get on and do it. This is our moment to do exactly that.

And don’t lose sight that if you’re an early stage, your investors are choosing YOU, not necessarily your business plan. All our businesses will evolve quickly now, and it’s our job to show others that change is what we thrive on, and no matter where this thing takes us, we’re going to keep turning into a real, growing business.

Share

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.

Leave a comment and tell me how you’re thinking about the topic of the bubble. And if you know another founder grinding through funding decisions, feel free to forward this along.

Subscribe now

But when I look at what we’re doing now at FOMO.ai, I realise the landscape has shifted. The truth is stark: a priced round is no longer the only, or the best, option for most founders, especially if you’re not playing the unicorn game. Most days, I find myself quietly grateful for this fact. And a bit bewildered that so many founders still feel compelled to chase the old milestones.

Here’s what I believe now: If you’re not laser-focused on building a unicorn, alternative funding structures, especially SAFE notes, offer founders more flexibility, cleaner cap tables, and the chance to keep control. You’re not obliged to play the “Series A, B, C” LinkedIn victory-lap game just because that’s what everyone before you did.

Bootstrapping First, But What If You Need Capital?

Let me be blunt: if you can bootstrap, bootstrap. That will always be the best path in most cases. Yes, I know some VCs even say this out loud, even if it’s bad for their deal flow. I once heard a VC in Boulder admit, “We want to invest in businesses that don’t need our money,” and that says a lot. The best businesses are the ones that don’t depend on outside capital, and every founder secretly knows it.

But real life is messy. Sometimes, you do need external capital to get out of the starting gates or to reach the next hilltop. That’s where founders get told, again and again, to default to priced rounds. I’m here to tell you there’s a better way.

SAFE Notes Changed the Game for Us

At FOMO.ai, every dollar we’ve raised has come through SAFE notes, always with a post-money valuation cap.

Here’s what that means for us. First, SAFE notes are genuinely easy. We use the Y Combinator post-money SAFE format, which slashed our legal costs and admin overhead. Sure, the Form D for the SEC is still required, but that’s a blip compared to the mountain of paperwork for a traditional convertible note or a full-priced round.

Why does this matter? Because when you’re an early-stage founder, every hour you spend with lawyers or wrangling admin is an hour you’re not talking to customers, building product, or recruiting the early believers who will actually move you forward.

But there’s more. SAFE investors don’t technically appear on our cap table. Instead, our table only shows the actual owners, me, my co-founder, and a small cluster of core team members. That’s it. Our investors are all individuals at this stage, and SAFE notes have also simplified things for them. No arcane legal rights to negotiate, no practical hurdles that scare away people who want to write smaller checks.

Critically, by choosing post-money SAFEs, our investors don’t get diluted by each other when we do multiple SAFE rounds. Unless we eventually convert to a priced round, they avoid that invisible dilution spiral that plagues pre-money structures. And frankly, as long as we stay with SAFE notes, that non-dilution advantage sticks with them right up until exit.

Raising Money Little by Little, And the Lessons That Came With It

Of course, nothing is perfect. Raising exclusively from individuals on SAFE notes has its unique set of headaches. Most of our investments were in increments of $25,000 - $250,000. That forced us to constantly be out in the world, pitching, courting, and convincing.

Looking back, I wish I’d invested more effort up front to put together a larger chunk of capital. It would have saved me weeks of time by avoiding the constant fundraising treadmill. There’s something to be said for gathering a big enough war-chest early, if you can do it without taking on more complexity than you want.

And if you do raise over time, as we did, you can keep increasing your valuation for less dilution down the line.

But there’s an upside. When institutional investors started circling, they liked what they saw. “You’ve only raised on SAFEs? Your cap table only shows the founders and a handful of core employees? That’s rare.” We had a clean, simple setup, one that’s almost impossible to maintain if you go down the priced round route early.

Founders and the Unicorn Trap

Every week, I see founders cheering online after closing their Series A, B, or C. There’s a cultural expectation in the US, probably driven by years of Valley folklore, that the path to success is paved with increasingly large, priced rounds. Get the seed. Get Series A. Hustle to B. Repeat. And if your “big number” goes up, you must be winning. Right?

I strongly question it.

If you’re building a business that truly needs $10 million-plus to have a shot, maybe you’re chasing a unicorn or playing in spaces where scale is do-or-die, then sure. Priced rounds are your reality, and SAFEs won’t get you all the way there.

But most of us aren’t actually building those businesses. At FOMO.ai, we’re not on the unicorn racetrack. We have a clear plan for an exit in the next 18 to 24 months at a multiple that will delight our investors. For that kind of target, SAFEs are plenty. If your total capital need is “several million” not “tens of millions,” ask yourself whether the dilution and control give-up of the legacy VC path really makes sense.

Truth is, if you can build a cash-generating business from day one, why chase increasingly big checks and shrink your own slice of the pie each time? Most of the time, it’s founder ego and external validation at play, not genuine necessity.

But SAFEs Aren’t One-Size-Fits-All, Read the Fine Print

Here’s where you need to be sharp: not all SAFE notes are created equal. So far, we’ve always used a valuation cap and stuck with the post-money format. That’s worked out because we understood exactly how it would convert, how shares issued lined up with shares authorised, and what it meant for us come exit or priced round.

But SAFE structures are flexible, which means you can get into trouble if you don’t understand the levers.

Some founders just use a discount, often 20 percent, which kicks in at exit or priced round. Others combine a valuation cap and a discount. Still others offer a SAFE that accrues interest and has a forced conversion on a specific date. Each option shifts the risk and upside between you and your investors.

My strong view: if you can, stick to just one variable, ideally the valuation cap. If you have to concede a second, say, adding a discount to the cap, it’s less ideal, but still manageable. The scenario I’d personally avoid is a SAFE with a forced expiry that triggers conversion, especially early on when your business is still finding its shape and timing is a moving target. Controlling the timing of conversion is strategic, so don’t surrender it lightly.

Valuations: More Art (and Nerve) Than Science

Let’s talk about setting the valuation cap, because this is where most founders stress out. I stumbled on a guy on TikTok who said that “an established company is valued by standard metrics like revenue, retention, gross margin, etc, but an early-stage startup’s valuation is determined in direct correlation to the size of the testicles on the founder.” Crude, a bit, but there’s real truth there. Early-stage valuations are subjective and largely depend on how confidently you defend your progress & vision.

We started at a $7.5 million post-money cap, which is about the norm. I’ve seen Founders go down to $4 million, others shoot for $10 million or more. I felt good about $7.5 million because my co-founder and I brought significant relevant experience, less risk for investors, and therefore more reason to push for the higher end. As we de-risked the business, began generating revenue quickly, and knocked off some high-uncertainty early milestones, we felt completely justified in bumping the cap to $12.5 million, then $15 million, then $20 million. Progress breeds leverage.

So, let the market validate your price to some degree, but don’t be afraid to hold your line, especially if you’ve got genuine traction or differentiation.

Why This is the New Normal (For Most Startups)

The shift didn’t just happen because founders woke up braver one day, the game itself has changed.

Back when I was part of that ad tech company, we needed significant funding because it took $millions to build even a basic tech platform. Today, thanks to AI and available tooling, we were able to generate revenue at FOMO.ai within 60 days of launching. That early revenue kept the lights on, and actively funded our technology. Building software is faster and cheaper than ever.

Today, the real struggle isn’t technology, it’s distribution.

(Read: The hard part has changed (technology —> distribution))

Getting your product seen and used is the big challenge, not building the thing in the first place. The old paradigm, big up-front money, institutional-only backers, priced rounds by default, just doesn’t fit that world anymore.

More founders in my circles are waking up to this. Instead of chasing priced rounds for validation, they’re choosing cleaner, more manageable cap tables and flexible funding that matches their actual business needs. Investors are getting it too. Some even worry about deal flow because founders aren’t queuing up for the expensive old model.

This is becoming the new normal, at least for companies whose ambition and capital needs sit somewhere south of moonshot territory.

What You Can Use Today

Here’s my distilled advice for founders right now:

• Bootstrap if you can. Real freedom starts there.

• If you need to raise, look at SAFE notes before defaulting to priced rounds.

• Keep your SAFE terms simple. One lever (ideally the valuation cap), occasionally two (add a discount if you must). Avoid forced conversion dates unless absolutely necessary.

• Use post-money SAFEs to keep your investors aligned and prevent them from being diluted by each other.

• Aim for a clean cap table. It makes you more attractive to future investors and saves admin pain later.

• Set your valuation cap with confidence, and let your progress justify raising it as you go.

• Understand your business goals. If you’re building for acquisition within a few years rather than chasing unicorn status, you don’t need the baggage (and dilution) of the “old path.”

• Raise bigger early if possible. Constantly doing lots of small rounds eats up time. See if you can get a critical mass up front.

If You Only Remember One Thing...

You do not have to follow the legacy playbook of priced rounds and endless dilution. For the majority of founders today, SAFE notes and, more broadly, flexible investment structures offer a smarter, cleaner, and more founder-friendly path. Your business type and ambition should drive your funding choices, not cultural inertia or ego.

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com. This post was helped by Hello Gordon, the easiest way to articulate your expert knowledge.

👉 Founder or operator? You can now schedule a 1:1 session with me.

At Chango for instance, we needed almost $3m just to get basic platform out the door, and even then it was helpd together by sticky tape to start.

That era is (largely) over.

Subscribe now

With AI accelerating development timelines, with no-code and low-code tools maturing, with open source libraries and cloud infrastructure available to anyone with a credit card, the actual building of software has become dramatically more accessible. A small team can ship in weeks what used to take a large team months. A single founder with the right tools can build a prototype over a weekend that would have required a seed round five years ago.

This should feel liberating, and in many ways it is. But it has also created a new problem. If everyone can build, then the product itself is no longer the differentiator. The companies that win are no longer the ones with the best technology. They are the ones that figure out how to get that technology in front of the right people, consistently, at scale, in a way that makes those people want to pay for it.

Distribution has become the hard part. And most founders are still operating as if the product is what matters most.

Building a tech company means getting eight things right at once

When you zoom out and look at what it actually takes to build a successful technology company, the list is humbling. You need to get the technology right, obviously. But you also need:

• distribution

• operations

• execution

• positioning

• timing

• team

• and capital.

All of these things need to work in some reasonable harmony, and a serious weakness in any one of them can sink you regardless of how strong the others are.

For a long time, technology sat at the top of that list in terms of difficulty and risk, but it’s clear that the balance has shifted. Technology has slid down the difficulty curve thanks to the tools and capabilities I just described, and distribution has climbed to the top.

This is something I think about constantly (at FOMO.ai), because we live it every day. We work with brands across industries, and the pattern is remarkably consistent. The companies struggling most are rarely struggling because their product is bad, but because they haven’t figured out how to reliably get their product in front of the people who would benefit from it. They have a technology problem that is mostly solved and a distribution problem that is wide open.

Or, how many times have we seen companies build a ‘solution awaiting a problem’.

And that realization, once you truly internalize it, changes how you think about everything from your go-to-market plan to your product roadmap to where you invest your time and money.

What twelve months of cold email taught us at FOMO.ai

I want to share something specific from our own experience.

We have been running cold email at FOMO.ai for about twelve months now. And I will be honest, the first few months were rough. The results were not where we wanted them to be, and it would have been very easy and very reasonable to shut the channel down and try something else. That is what most companies do. They test a channel for 60 or 90 days, decide it isn’t working, and move on to the next thing.

But we kept coming back to this fundamental belief that distribution is the hardest and most valuable problem to solve right now. Cold email was putting us in front of our target market. The conversations were happening, even if they weren’t converting the way we wanted. The distribution was working at a mechanical level, which is actually the hardest part to get right.

So instead of killing the channel, we did something that could have felt counterintuitive at the time. We looked at the distribution we had working, and we tweaked our offering to better align with it. Rather than forcing a specific product through a distribution channel and hoping for the best, we let the distribution inform what we should be selling and how we should be packaging it.

This reinforced something I now believe deeply: most founders build the product first and then go looking for distribution. The ones who win tend to find distribution first and then shape the product to match. Or at the very least, they treat the relationship between product and distribution as a two-way conversation rather than a one-way dictation. A company that is spending 80% of its resources on product and 20% on distribution is going to have a very hard time.

Define the box you’re playing in

None of this works, though, if you don’t know who you’re trying to reach. And I mean really know, not in a vague “we sell to marketers” kind of way, but with genuine specificity about the industry, the role, the company size, and most importantly the problem that is urgent enough and important enough to make someone respond to your outreach or click on your ad or pick up the phone.

This is the foundational step that makes distribution possible. You can have the best cold email sequence in the world, the most compelling ad creative, the sharpest sales deck, but if you’re pointing it at the wrong audience or trying to solve a problem that nobody is losing sleep over, none of it will land.

The way I think about this has evolved over time. I used to approach market definition as a product exercise, asking who would benefit from what we’ve built. Now I approach it as a distribution exercise, asking where can we reliably and repeatedly get in front of people who have a problem we can solve. Those two questions sound similar, but they lead you to very different answers, because the second one forces you to consider not just who your ideal customer is but whether you can actually reach them.

Anchor your story in something undeniable

Once you know who you’re targeting and what problem you’re solving, the next question is why now. Why should anyone care about this problem today instead of next quarter? Why should they prioritize this over the fifteen other things competing for their attention and their budget?

This is where timing and context become critical distribution tools. If you can anchor your message in something that your target market already knows is true, something they’ve been reading about in industry publications or hearing about at conferences or worrying about in their own leadership meetings, your outreach stops feeling like a cold pitch and starts feeling like a timely conversation.

The best companies I’ve seen do this find a macro level shift in their industry that creates genuine urgency around the problem they solve. Not manufactured urgency, not “limited time offer” urgency, but the kind of urgency that comes from a buyer recognizing that the world is changing and they need to adapt. When your cold email or your LinkedIn post or your sales deck can tap into that existing anxiety and then offer a clear path forward, the response rates change dramatically.

One of the things we’ve found most effective in our own outbound is leading with the shift rather than leading with the product. We talk about what’s changing in AI and marketing and search before we ever talk about what we do. That approach works because it meets the prospect where they already are, which is worried about something real and looking for someone who understands the problem deeply enough to help.

Pick your segment and commit

This one is closely related to the distribution question because the segment of the market you choose to serve dictates which distribution channels are even viable for you.

If you’re selling to solo practitioners or very small businesses at a low price point, you simply cannot afford to have salespeople running demos. The math doesn’t work. You need a self-serve, product-led motion where people can find you, try you, and buy you without ever talking to a human. That means your distribution has to be built around inbound channels, content, SEO, organic social, and paid acquisition with very efficient unit economics.

If you’re selling to enterprises at higher price points, you can afford a longer and more complex sales cycle, but you need to reach decision makers who are notoriously hard to get in front of. That means your distribution might lean more heavily on outbound, events, partnerships, and relationship-driven selling.

The point is that your positioning and your distribution strategy are deeply intertwined, and you can’t figure out one without the other. Too many founders pick a segment based on where they think the biggest opportunity is, without asking whether they can actually build a distribution engine to reach that segment. And too many founders try to serve multiple segments simultaneously, splitting their distribution efforts and diluting their messaging to the point where nothing resonates with anyone.

Commitment to a segment feels risky. But in my experience, focus is what makes distribution work. The more specific you are about who you’re talking to, the more your message can be tailored to their exact situation, and the more likely they are to feel like you built this thing for them.

(Side quest… proactive versus reactive)

Remember, when you do choose a segment, the work you put into scaling growth is your PROACTIVE distribution. However, you may stumble across some sales prospects who want to work with you, find you organically, and are not really different enough from your ICP to cause problems, so you take them on. That is REACTIVE distribution. If you don’t take this too far, this is a great way to organize your team around a common goal, while at the same time grabbing some opportunistic revenue.

Know the landscape before you spend a dollar

Before you invest serious time or money in any distribution channel, you need to understand what your prospects are already seeing from other companies in your space, because you are not operating in a vacuum. Your potential customers are being reached by your competitors through many of the same channels you’re planning to use, and if you don’t know what those competitors are saying and how they’re positioning themselves, you’re going in blind.

Spend an hour mapping it out. Look at competitor websites, read their reviews, pay attention to the ads they’re running, and the content they’re publishing. Try to understand not just what they’re selling but how they’re distributing it. Are they dominating a particular channel? Is there a channel where nobody is competing effectively? Are they all saying the same thing, which might mean there’s an opportunity to say something different?

This exercise almost always surfaces one of two opportunities. Either you find a segment of the market that is genuinely underserved, where the distribution landscape is wide open, or you identify a way to differentiate your message so clearly that you can compete directly and win. Both are valuable. But you can’t find either one if you’re not looking.

Make your message simple enough to survive a cold email

Here is something our twelve months of cold outbound taught me about messaging: if your value proposition can’t be understood in the time it takes someone to glance at a cold email on their phone, it is too complicated. That is the real test. Not whether it sounds impressive in a boardroom or whether your investors nod along during a pitch meeting, but whether a busy person who has never heard of you can read one sentence and immediately understand what you do and why it might matter to them.

Most companies overcomplicate their messaging because they’re trying to sound sophisticated. They use jargon, they layer in buzzwords, they try to communicate every feature and every differentiator in a single paragraph. And the result is that nobody processes any of it. The message gets deleted, the ad gets scrolled past, and the website visitor bounces in three seconds.

The discipline of simplification is one of the hardest parts of building a go-to-market plan, and it is also one of the most impactful. When we simplified our messaging at FOMO.ai, when we stripped away the insider language and said plainly what we do and who we do it for, our outbound performance improved meaningfully. People could actually understand what we were offering, which meant they could actually decide whether it was relevant to them.

Your messaging also needs to work across two very different scenarios. The first is overcoming apathy, where your prospect kind of knows they have a problem but isn’t motivated to do anything about it right now. The second is overcoming comparison, where your prospect is actively evaluating you against a competitor. Good messaging handles both. It creates urgency by connecting to something the prospect already cares about, and it differentiates by being specific about what makes you the better choice.

Execute one channel, learn, then scale

Once all of these foundational pieces are in place, the temptation is to go wide and try everything at once. Resist that temptation.

The smartest approach I’ve seen, and the one we follow at FOMO.ai, is to pick one or two channels, execute them well, collect data, and learn. Figure out whether your messaging actually resonates in the real world. Figure out which segments of your target market respond best. Figure out what objections come up and what questions people ask. All of that learning is gold, and you can only get it by being focused enough to actually pay attention to the signal.

Once you’ve found something that works, even modestly, then you can start to scale it and layer on additional channels. The founders who do well are the ones who own their early go-to-market motions themselves rather than immediately outsourcing to agencies or fractional hires. They develop a firsthand understanding of what works and why, and then when they do bring in outside help, they can hand over a playbook that is grounded in real data rather than assumptions.

(Our exception is cold email. We have a dedicated team managing it because it is SO hard these days.)

This is also where patience becomes a genuine competitive advantage. As I mentioned, we could have killed cold email after month three. Most companies would have. But by staying with it long enough to learn and adapt, we turned a channel that felt like it was failing into one of the most valuable distribution assets in our business. The lesson there isn’t that you should stick with something that isn’t working forever; the lesson is that distribution channels almost never work on the first try, and the companies that figure them out are the ones that stay long enough to iterate.

The real takeaway

The landscape has shifted underneath us. Technology is no longer the moat. In a world where anyone can build, the companies that win will be the ones that figure out distribution and have the patience to let it compound over time. That means your go-to-market plan isn’t just one workstream among many. It is arguably the most important strategic decision your company will make.

And it means that the foundational work, defining your market, understanding your timing, choosing your segment, studying the competition, simplifying your messaging, all of that boring, unsexy, strategic thinking, is actually the highest leverage work you can do.

Because without it, every dollar you spend on execution is a guess, and with it, every dollar works harder than the last.

Share

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.

Subscribe now

I hated that feeling, like I’d handed my business off to a robot who couldn’t care less. And I realized, ironically, just how much I miss the humans.

Here’s what I believe: while AI is incredibly powerful, the companies that succeed in the next wave will be those that blend artificial intelligence with accessible, accountable human expertise. Not just in theory, but built right into the DNA of how they operate. We’ve built FOMO.ai around this conviction, and the longer I’m in this space, the more certain I become that the hybrid approach is not a compromise. It’s the future.

AI Can Match Genius, But Not Sell It

One of my favorite stories to illustrate this tension comes from my days in advertising. Imagine being in the room when a creative team pitched Nike on “Just Do It.” The phrase, of course, is now part of brand history.

But think about this… These days, if you dumped all the prep materials, brainstorming notes, and transcripts from those meetings into an AI, I’m convinced the algorithm could eventually spit out “Just Do It” among its hundreds of suggestions.

That’s not the remarkable bit.

The real moment of value came when a real, live person stood up, put their name to the idea, and looked the Nike team in the eyes. They staked their professional reputations on a few bold words and convinced a roomful of skeptical executives that this was the right move. AI could have generated the line, but it couldn’t have sold the vision.

That’s the moment I always come back to when people ask if humans are really necessary in AI-driven work. The answer, for me, is a resounding yes.

Where AI Shines, Where Humans Lead

When we started, we made a deliberate choice: figure out which jobs AI does best, and which calls still need a human touch. For us, that breakdown is clear. Our platform helps clients earn sales and leads from AI-powered search. We let the AI handle the heavy optimization and the analysis, the kind of work that used to take two months in an agency is now completed in under an hour. It’s almost absurd how efficient it’s become; nobody wants to wait two months for information these days!

AI is also remarkably good at writing what I’d call “human quality” content, but for some clients, usually those with a BIG brand name, the last mile matters. So, we review content with humans, we help interpret strategy, and we guide clients not just through what’s happening but also WHY.

The trick is defining and protecting those roles. Let AI do the math, find the patterns, and generate at scale. Let people anchor the strategy, interpret nuance, and step in when intuition is required.

Why Speed Isn’t Everything (And What’s Missing from Pure Tech)

It’s easy to celebrate the speed that AI brings. Clients see results quickly, in a world where everyone wants their content found by the big AI models as fast as possible. The pace has become non-negotiable.

Fast results usually trigger delight, at least at first.

But there’s an undercurrent I’ve noticed: many clients assume that working with an “AI company” means dealing with something cold and faceless. They brace themselves for impersonal exchanges, ticketing systems, and the sinking feeling that nobody on the other side is invested in their success. So, when we reach out with a name, a face, and a direct line for questions, a private Slack channel, (and even a free Starbucks coffee,) they’re often caught off guard (in a good way).

Delight & surprise always pays dividends.

We don’t hide the humans behind layers of process. Sometimes our team acts as what some could call a “safety blanket.” That doesn’t just mean technical support, it’s also for things like, “I saw something scary on LinkedIn about AI and marketing. Does this affect me?” or “Am I making a huge mistake I’m not seeing?”

Here’s what I find myself repeating: the human element isn’t about coddling; it’s about trust and accountability. When things move at AI speeds, mistakes happen quickly, too. Sometimes, you just want a real person to tell you it’s going to be okay, or flag when it’s really not.

Dyson Had It Right

It is said that James Dyson, the inventor of the bagless vacuum cleaner, was in a meeting where the operations team was discussing new ways to optimize processes so their call centres received fewer calls.

His attitude was that they should consider themselves fortunate to have a customer reach out to them, and they should treat it as a positive opportunity.

YES!

It goes without saying that you should try and make a better product, and you should try and make calls efficient for both parties, but the point is that human interaction is meaningful, and can be turned into a future opportunity.

The Myth of Pure SaaS (and the Cost of Infinite Scaling)

I want to raise a flag here because I see a widespread delusion: that stripping out human involvement always means a better business. Investors (especially in SaaS) have hammered this home for years. Services, real, messy, human services, were seen as a flaw because you can’t “infinitely scale” people.

But here’s the uncomfortable truth: sometimes, the relentless pursuit of pure SaaS creates awful user experiences. My cap table story is one of many. Waiting days for a robotic answer to a simple problem isn’t just infuriating, it’s damaging. It devalues my investment and erodes trust.

I’m convinced that this investor-driven mindset, that any human touch is a negative, has outlived its usefulness. It’s not just me; I see it shifting. Some investors still say “no thanks” when they realize we involve people. They hear “human” and think “lower margins.” But our margins are still strong, better than some “pure” SaaS companies. Plus, user experience trumps theoretical scalability, especially when every competitor sounds and acts the same.

When every tool is faceless, customers start craving a real conversation. They don’t want another soulless ticketing system. Real solutions, delivered by real experts, are starting to look like an advantage again. I welcome this.

The Human Safety Net: It’s Not Optional

Some people, especially in tech, see human support as optional, a “nice to have” layered on top of a “real” AI platform. They’re wrong. Building ways for customers to connect with a person, even if they might never use it, changes how they feel about your entire company. It means when something feels risky or unclear, there’s someone actually accountable.

We do have offerings that are totally hands-off if a client wants that. Our AI can drive results alone. But we always keep the door open for human interaction when the stakes get high, or the questions become existential. Sometimes, our clients don’t use us for troubleshooting at all; they use us as a gut check or to interpret broad shifts happening in real time.

Making humans available isn’t just about solving immediate problems; it’s also an anchor for trust as our world becomes more automated and, yes, more uncertain.

Done-For-You Shouldn’t Mean Overpriced

Another lesson we’ve internalized: we can offer “done for you” services powered by both humans and our AI, giving clients the full agency experience but at a fraction of legacy agency costs. This flexibility works well for us. Clients who want to hand everything off can, but their experience is guided by real experts using the best AI we have. They get results, and know someone’s watching the details.

This spectrum, sometimes hands-off, sometimes deeply personalized, is what modern customers actually want. Not everyone wants to run everything themselves, nor do they want to be left in the dark. People value choice, and the closer we get to pure automation without any welcome mat for actual conversations, the more I see people pushing back.

What You Can Use Today

If you’re building, buying, or investing in AI-powered solutions, here’s my advice from the trenches:

1. Define the roles. Be crystal clear about what the AI should do (data processing, analysis, content generation) and which parts require human judgment (strategy, brand nuance, client reassurance). Know it will evolve quickly.

2. Don’t hide your people. Even a simple Slack channel or an email address to a real expert can build massive goodwill. It’s not just tech support, it’s forming relationships.

3. Speed plus expertise wins. Use AI for instant results, but back it up with someone who can interpret what’s happened and advise on next steps.

4. Beware the infinite scale trap. “Pure” SaaS looks good on paper but can destroy customer trust if support is non-existent or robotic.

5. Flex your offering. Some clients want totally self-serve, some want agency-style engagement. There is room for both, in fact, most crave optionality.

6. Be as human as you can be. Send that wedding card, send that food hamper to the sick client who’s been with you for 12 months.

If You Only Remember One Thing...

Technology moves fast, but trust is built slowly. Companies that combine AI efficiency with human accountability will win the next decade, not by scaling facelessness, but by making sure someone always picks up the phone.

Share

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.

Subscribe now

At the recent Web Summit Qatar, AI founders and VCs pushed back hard.

• Glean's founder (valued at $7B) said software won't die, it'll embed AI and keep going.

• Miro's founder said valuations are "crazy" and will correct, but gave it about two years.

• And Amino Capital's Larry Li sees the bubble deflating, especially for the big players.

The dot-com comparison keeps coming up. but people always take the wrong lesson from it.

Yes, a ton of companies died. But the internet didn't. Amazon didn't. Google didn't. Hundreds of real businesses with .com in their name didn’t. The real companies with real customers and real revenue survived and thrived.

The frenzied hype might have taken a beating but the technology didn't.

AI will be the same. The noise will shake out. The companies actually solving problems and generating revenue will still be here. (Hi, that's us.)

What I found most interesting: a non-AI founder on a panel said he's being benchmarked against AI companies growing 1,000% year over year. Imagine pitching your perfectly healthy business, and the investor is comparing you to a rocketship that's also on fire. That's the market right now, and I feel for those founders.

Oh, and OpenAI is reportedly set to lose $14 billion this year AND more than $340 billion chased startups in 2025, with 65%+ going to AI. The money is flowing, despite the overexcited headlines.

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.

For the last twenty years, the entire SaaS industry has run on a per seat pricing model. You pay per user, per month. The value you get from the software depends almost entirely on how well your people use it. The software helps you do the work, but a human still has to sit there and actually do it.

The thesis behind FOMO.ai was always different. We didn’t want to sell access to a tool and then hope our customers figured out how to get value from it. We wanted to deliver completed work. Actual marketing output. Content that’s written, strategies that are executed, campaigns that are live. The outcome, not just the environment to work in.

And I’m now seeing this same shift happening across the entire software industry.

Subscribe now

Software is starting to look a lot like a services business.

Think about what a services firm does. You hire McKinsey or Deloitte or a specialized agency, and they complete a project for you. They deliver an outcome. You don’t pay them per seat. You pay them because they promised to get something done, and then they did.

(Ironically, these same firms are already delivering work done by AI, and getting caught with their pants down because they used the tools badly and did not give humans a chance to check it.)

Now imagine software that works the same way. You don’t pay for access to a dashboard. You pay because the software completed a biotech drug discovery process, or designed an engineering project, or built an entire content strategy from scratch. That’s where things are heading, and quickly.

If you look at the total revenue generated by the global services economy and imagine even a fraction of that shifting to software companies who can deliver the same outcomes at 10x to 100x the efficiency, you start to understand why people in the venture capital world are losing their minds right now. The companies that figure out how to deliver on that value creation are going to generate outsized returns.

I see this every day in our own work.

We use AI at FOMO.ai to complete projects that I would have previously needed to hire an outside firm for. Research projects that would have taken a team of analysts months, we can now get meaningful output from in a matter of days. And in some cases, the AI produces insights that no services firm could have delivered because of the sheer volume of data it can process and the patterns it can identify.

That’s a wild thing to say out loud, but I believe it’s true. And it’s exactly why we built the company the way we did.

The job consolidation piece is even more interesting.

Something else is happening alongside this shift, and you can see it playing out in real time at companies of every size. Job functions are consolidating. A product manager, a UX designer, and a developer used to be three distinct roles with clear boundaries. Now you’ve got designers saying, “I can vibe code this myself,” developers saying, “I can handle the UX with a plugin,” and product managers saying, “I can do both of those functions.” Three jobs are competing to do the same work.

And then there’s the middle manager question. Think about what a typical middle manager at a large company actually does day to day. They go to meetings, they decide what meetings to schedule, they set the agenda, they create action items, and they follow up. A huge portion of that workflow is now being handled by AI meeting tools that listen, transcribe, generate action items, and track follow-through. All that coordination work that used to require a person is quietly being automated.

When one person can credibly do the work of three or four, the math changes dramatically. Companies become more profitable with smaller teams. Individual employees become far more valuable because their output multiplies. And the companies that figure out how to structure themselves around this new reality are going to generate returns that seem almost unfair compared to competitors who are still staffed the old way.

What this means for you.

If you run a business or a marketing team, this isn’t a theoretical conversation. The shift from per-seat software pricing to outcome-based pricing is going to change what you buy, how you buy it, and what you expect from the tools you invest in.

Without trying to sound overly dramatic, I think the question for every business owner right now is whether you’re going to restructure around this new reality or watch your competitors do it first.

👉 Founder or operator? You can now schedule a 1:1 session with me.

Share

The accountants would have hated it. The couple never forgot it.

Subscribe now

That story from Guidara’s book “Unreasonable Hospitality” captures something most businesses miss completely. We’ve gotten so good at measuring efficiency that we’ve forgotten to create moments worth remembering. Every interaction is optimized, every process is streamlined, and everything shows up neatly in our dashboards. But none of that makes people tell their friends about you.

The Gestures That Don’t Scale Are the Ones That Matter

At FOMO.ai, we do something that probably looks ridiculous on paper. When a sales prospect cancels because they’re sick, we send them a coffee credit or meal delivery gift. When clients get married, we send wedding gifts. These cost us maybe fifty bucks each time, and there’s no line item in our revenue forecast that tracks the return.

But here’s what actually happens.

The prospect who canceled and got a DoorDash credit? Six months later when they finally had budget, they came back to us instead of the three competitors they’d been considering. The client whose wedding we acknowledged? They’ve referred us to three companies in their network.

None of this shows up in a quarterly report, but it absolutely shows up in retention rates and the kind of word of mouth that paid ads can’t touch.

Most Companies Are Technically Correct and Completely Forgettable

The software industry has gotten really good at efficiency and really bad at humanity. Everything is automated onboarding flows, triggered email sequences, and chatbots that try to sound friendly but feel hollow.

Guidara’s approach was different. He looked for moments where his team could go beyond what was expected, practical, or even sensible from a cost perspective. The hot dog story works because it is connected to something the diners actually care about. When we send a meal delivery credit to someone who’s sick, we’re acknowledging that they’re human and that we care about them beyond the meeting they missed.

The Trick Is Knowing When to Be Unreasonable

Reading “Unreasonable Hospitality” doesn’t mean doing random nice things and hoping they stick.

Guidara and his team were strategic about when and how they went beyond expectations. They watched for moments that mattered and listened for clues about what would be meaningful to specific people. A generic surprise doesn’t carry much weight, but something that connects to what someone actually values becomes part of their story.

Think about the last time someone went out of their way for you in a business context. You probably still remember it because it was unexpected and personal.

Where AI Actually Helps (By Getting Out of the Way)

We’re building AI tools at FOMO.ai, and much of what we do focuses on automation and efficiency. But the reason we automate the routine work is so humans can focus on the unreasonable moments.

If your team isn’t drowning in repetitive tasks, they have the mental space and energy to notice the opportunities for genuine connection.

The AI can handle standard responses, data analysis, and routine content creation. What it can’t do is notice that someone mentioned their daughter’s college graduation in passing. It can’t decide that a prospect who’s having a terrible week deserves something to brighten their day. Those judgment calls require human intuition, empathy, and the willingness to do something without a clear ROI.

How to Start Being Memorably Unreasonable

You don’t need to transform your entire operation. Start by asking better questions. Where are you technically correct but emotionally flat? What do your customers mention in passing that reveals what they really value?

Then pick one moment and figure out how you could make it unreasonable in the best way. Maybe it’s sending a handwritten note after a particularly good call. Maybe it’s surprising a long-term client with something you know they’d appreciate.

Will Guidara built the best restaurant in the world by caring about details that seemed ridiculous. The question is whether you’re willing to be that ridiculous too. In a world that’s increasingly automated and efficient, being unreasonably human might be the most reasonable strategy there is.

👉 Founder or operator? You can now schedule a 1:1 session with me.

(Image: Nasjonalbiblioteket from Norway, CC BY 2.0 via Wikimedia Commons)

In 1914, Ernest Shackleton supposedly placed this ad to recruit for his Antarctic expedition:

“Men wanted for hazardous journey. Low wages, bitter cold, long hours of complete darkness. Safe return doubtful. Honour and recognition in event of success.”

Legend says he was flooded with applicants.

Whether the ad is real or apocryphal, the lesson behind it is bulletproof. Shackleton didn’t sell the dream. He sold the reality. And the people who showed up were genuinely ready for what was coming.

Now compare that to how most companies recruit salespeople.

The job posting says: “Dynamic, fast-paced environment. Unlimited earning potential. Join our winning team!”

What actually happens: Your first three months will feel like shouting into the void. You’ll hear “no” 47 times before your first “yes.” Your pipeline will be empty more often than it’s full. And some weeks, you’ll question every career decision you’ve ever made.

We don’t say that part out loud. And then we act shocked when people quit before the end of Q1.

Here’s what I’ve seen happen over and over again.

A company hires a salesperson. The interview is all sizzle. The product is amazing, the market is huge, the comp plan is generous. The candidate is excited. Everyone shakes hands.

Ninety days later, the new hire is gone. Not because they couldn’t sell, but because nobody told them what selling here actually looks like on a Tuesday afternoon when the CRM is empty and the phone isn’t ringing.

The hiring manager painted a picture of a Ferrari. The new hire showed up and found a go-kart. A perfectly good go-kart that another seller might have liked, but absolutely not what they were expecting when they accepted the offer.

And this keeps happening because the recruiting process itself is broken. Companies treat the interview like a sales pitch when it should be a Shackleton ad.

So here’s my challenge to anyone hiring salespeople right now: write your own Shackleton ad.

Not literally. But before your next interview, sit down and write the brutally honest version of the job. The one you’d never post on LinkedIn. The version that makes you slightly uncomfortable reading it back to yourself.

Something like:

“You will spend 80% of your time prospecting. Our brand is not well known yet, and nobody is coming inbound begging to buy. You’ll need to build your own pipeline from scratch. The product is good but still evolving, so you’ll occasionally sell something that doesn’t fully exist yet and then work with the team to make it real. Your first deal will probably take 60 to 90 days. If you need a base salary to feel safe, this probably won’t work for you. If you thrive on building something from nothing, keep reading.”

Is that going to scare some people away? Absolutely. And that’s the entire point.

The best salespeople I’ve ever worked with didn’t want the easy gig. They wanted to understand what they were walking into. They wanted the truth so they could decide if THIS was the fight they wanted to take on.

When you’re honest about the hard parts in the interview, three things happen.

The wrong people remove themselves from the process. That saves you months of wasted salary, wasted training time, and wasted morale on both sides of the table.

The right people lean in harder. The ones who read your Shackleton ad and think “that sounds like exactly my kind of challenge” are the ones who will still be with you a year from now.

And trust gets built before they even start. You told them the truth. They showed up anyway. That is a fundamentally different starting point than discovering the truth three weeks in and feeling like they were sold a bill of goods.

I talk to founders and sales leaders every week who tell me their biggest problem is hiring. The churn is too high, ramp takes too long, and the culture fit keeps missing.

Almost every time, when I dig in, the root cause is the same: the interview sold a version of the job that doesn’t actually exist in the real world.

Stop selling the role and start describing it.

Describe the worst Tuesday. Describe the deal that fell apart at the last minute because legal got involved. Describe what the CRM looks like at 5pm on a Friday when nothing closed that week. Describe the parts of the job that make your current top performer want to throw their laptop out the window, and then describe why they stay anyway.

That last part, the reason your best people stay despite everything, is your actual value proposition. Not the comp plan. Not the snack wall. The real reason people stick around when things get hard.

Shackleton figured this out over 100 years ago. Specificity attracts the right people. Honesty filters out the wrong ones. And the best salespeople in the world don’t want to be sold on a role. They want to be trusted with the truth about it.

The next time you write a job description or sit down for a first round interview, ask yourself: am I being Shackleton, or am I being every other job posting on Indeed?

Your churn rate already knows the answer.

Dax is the Co-Founder & CEO @ FOMO.ai, and the author of 84Futures.com.

👉 Founder or operator? You can now schedule a 1:1 session with me.