Why the GTM playbook stopped working.

The old SaaS GTM playbook is getting expensive and invisible. Three startups stopped optimising it and started building things buyers actually wanted.

The go-to-market playbook didn’t stop working because it was wrong. It stopped working because it was right, and then everyone started running it at the same time.

Every SaaS company is running some version of the same motion right now. Optimise every click, A/B test every subject line, gate your best thinking behind a form that 4,000 people will fill out and maybe 12 will actually read, and call it a strategy. It made sense when fewer companies were doing it. The returns have been shrinking for a while now, and most teams are just optimising harder.

Cold outreach response rates are cratering. Website heatmaps have gone cold. And the demos you spent months perfecting? Buyers are sitting through them and still choosing a competitor. Or worse, choosing nothing at all.

The 💜 of EU tech

The latest rumblings from the EU tech scene, a story from our wise ol' founder Boris, and some questionable AI art. It's free, every week, in your inbox. Sign up now!

Buyers know when they’re being sold to. They know the playbook because everyone is running it, and they’ve quietly started tuning it out. With AI handling more of their initial research, a lot of them aren’t even visiting your site to begin the process.

A study published in the Journal of Risk and Financial Management found that somewhere between 80% and 95% of a startup’s success depends on its marketing. Which means getting this wrong kills the company.

But here’s what I keep seeing across the founders I work with: the startups that are actually growing right now aren’t the ones who optimised the old playbook harder. They’re the ones who stopped running it entirely and built something buyers genuinely wanted to engage with.

Distribution happened because nobody was forcing it.

Three companies in particular show what that looks like, because they were bold enough to ask a different question.

Nobody wants to sit through your demo

A buyer comes to your website, clicks around, seems interested enough to book a demo. Your team celebrates. Then they sit through 30 to 60 minutes of being talked at, watching someone else drive the product they’re supposed to be evaluating. By the end of the week, they’ve watched two more demos from competitors. Everyone blurs together.

Varun Anand, co-founder of the data enrichment platform Clay, decided that was a waste of everyone’s time.

Anand had been tracking data enrichment complaints across online communities. Instead of spinning up a retargeting campaign or crafting outbound sequences, he went to where the pain already lived and invited people to try Clay themselves. The demos that followed weren’t really demos at all. Prospects signed in, shared their screens, and started working on their actual problems while Clay’s team annotated in real time over Zoom. No slide deck. No scripted walkthrough. Just: here’s the tool, solve your problem, and we’ll guide you if you get stuck.

“My goal is to blow their minds,” Anand said. “Can you solve someone’s entire problem in 30 minutes and have them do it? I think that is a bar that most software companies cannot achieve.”

The reverse demo did two things at once. Customers didn’t just understand Clay. They learned how to use it. And the sales team stopped needing multiple follow-up calls because the prospect had already done the work during the demo itself. Previously, they might spend several rounds of calls trying to close a deal. Now they could get there with little to no additional pitch.

That’s a fundamentally different relationship. The product became the pitch, and the pitch became the onboarding. Buyers left the call knowing exactly whether Clay solved their problem or not. No ambiguity. No “let me think about it and get back to you.”

Fifty hours of brainstorming, 100 million views

Clay proved the product could be its own marketing. Ramp, a spend management and accounting platform, proved something adjacent: the spectacle could be its own campaign.

You’ve probably heard some version of “go viral” tossed around in a marketing brainstorm. (If you have, I’m sorry.) But what Ramp actually did was way more deliberate than that. They put actor Brian Baumgartner, best known for playing Kevin in the US version of The Office, inside a glass box in New York City. Baumgartner sat at a desk surrounded by piles and piles of receipts, manually crunching numbers on full display for the passing crowds, playing the role of Ramp’s CFO. The whole thing was live-streamed.

A live counter tracked how many receipts Baumgartner processed versus how many Ramp processed at the same time. The contrast was the entire point.

Then luck showed up. Andy Buckley, who played Dunder Mifflin’s CFO David Wallace in the same show, turned up at the stunt set. Ramp’s team captured additional content with Buckley that went viral on TikTok. You genuinely cannot plan that. But you can create the conditions where it becomes possible.

None of the rest of it happened by accident, though. Ramp’s Head of Creative Experimentation, Kendall Hope Tucker, had connected with Rohan Kumar, the social lead for MrBeast, through cold outreach. That relationship turned into regular brainstorming sessions on what actually makes content spread. The final tally: over 50 hours of brainstorming and eight hours of live-streaming that generated more than 100 million views across social media platforms.

The stunt worked because it wasn’t pretending to be something it wasn’t. It was genuinely entertaining, it was specific to what Ramp actually does, and it gave people something worth sharing. That last part is the bit most marketing teams skip entirely.

The influencer play that didn’t feel like one

Clay made the product the pitch. Ramp made the spectacle the campaign. Gamma, an AI website and presentation builder, found a third way in: make other people’s communities the channel.

Working with influencers is a familiar playbook, and it has the same staleness problem as everything else. Massive followings. Scripted reads. The #ad and #sponcon tags that buyers have trained themselves to scroll past. Co-founder Grant Lee had been watching Gamma’s runway get shorter while sign-ups weren’t hitting the numbers the team needed. Big influencer partnerships weren’t a realistic option. (And honestly, they probably wouldn’t have moved the needle anyway.)

So Lee went smaller. Way smaller. Gamma started partnering with micro-influencers, creators with 10,000 to 100,000 followers who had tight, engaged audiences that actually trusted them. The data backs the move: according to Digital Applied, micro-influencers deliver over three times the engagement of macro-influencers at roughly 60% of the cost.

But the numbers were only half of it. The real difference was the relationship. Lee’s team onboarded influencers the same way they’d onboard a new hire. They invested in the creators’ understanding of the product and gave them room to talk about Gamma in their own voice. Lee described it as being closer to a teacher sharing a product they love with other teachers over summer break than a paid endorsement. That’s the kind of signal that actually breaks through.

The results confirmed it. Gamma has seen over 50,000 sign-ups in a single day and now has more than 50 million users. Over half of the company’s new subscribers come from word-of-mouth referrals. They didn’t sit around waiting for organic word of mouth to happen. They engineered it, through partnerships that felt organic because, in a lot of ways, they actually were.

Everyone learned the old playbook. Including the buyers.

These three companies did very different things, but they followed the same logic. Clay replaced the pitch with the product. Ramp turned the campaign into something people actually wanted to watch. Gamma figured out that the most effective recommendation is the one that doesn’t feel like a recommendation at all.

In every case, distribution happened because the thing being distributed was genuinely worth someone’s time. That’s the shift. The starting question changed.

The temptation, especially if you’re early-stage and feeling the pressure, is to default to the playbook that worked for the last generation of SaaS companies. (And honestly, parts of it still have value.) But the founders I’m watching most closely aren’t pouring budget into optimising a funnel that buyers have quietly learned to ignore. They’re asking what they could build, create, or make available that would be genuinely worth someone’s time.

Clay, Ramp, and Gamma are three of six AI-native companies that HubSpot for Startups profiled on exactly this question. The other three stories are worth reading alongside these.

That’s a way harder question than “how do we optimise our open rates.” But it’s the right one.

HubSpot for Startups works with early-stage founders building their first real go-to-market motion. If you’re pre-seed through Series A and trying to figure out what actually moves pipeline right now (not what worked in 2021), this is the program.

You can check it out here.