Every distribution tactic eventually stops working. Cold email pulled strong response rates until everyone flooded the inbox and buyers tuned out. Today, LinkedIn is the go-to channel for reach, but Eugene Berson is already watching the sharper founders drift to Substack. After a career spent building go-to-market, Earlybird's newest Catalyst makes the point that the channel is never where the work lives: whatever works now has a shelf life, and your job is reading the data closely enough to move before everyone else.

That instinct comes from several stages Eugene has lived firsthand. He built distribution at Slack, Asana, and Miro, and now runs go-to-market at Lambda. He joins the Earlybird Catalyst Program to put that experience in front of our founders, aiming at the hardest part of scaling: the months right after product-market fit.

From "it works" to "it works predictably"

Eugene brings specific value to the Earlybird’s portfolio: the operating know-how to take a company from "it works" to "it works, predictably," plus a working read on the US market for those moving there. His experiences stem from a path that goes through Wall Street to the GPU cloud: four years in investment banking at Morgan Stanley and a boutique, then a long run in product-led SaaS at Slack, Asana, and Miro, where the market was "all knowledge workers." He now runs go-to-market at Lambda, which rents the latest Nvidia GPUs to AI companies.

Throughout his workline, the constant was never the product category but the moment when demand was proven, and the job became making growth predictable.

"Almost all my experience is the stage right after a company hits product-market fit in a specific segment. Take Slack. It was already around $75M in ARR when I joined, with real escape velocity in SMB and self-serve signups. Moving upmarket took a different engine: building out sales, evolving marketing focus from brand to demand generation campaigns, and standing up customer success for retention and signal back into product.” Eugene says.

Companies still struggle after product-market fit

Once a company lands PMF, not stalling is a major hurdle. Founders spend much of their first steps building up a motion that can win the first thousand users based on founder energy via word of mouth, and a self-serve product. This does not move an enterprise buyer. Bigger deals mean more stakeholders, longer cycles, and committee decisions, so the go-to-market machine has to grow up to match. Timing that build-out is Eugene’s cup of tea, but it is not everyone’s cup of tea.

There are no silver bullets in distribution

"Whenever you find something that works, it usually doesn't last. Cold email used to get strong response rates, then everyone flooded the inbox, and people tuned it out. LinkedIn is great right now, but I'm already watching people move to Substack, and it'll keep moving. So discipline isn't the channel. It's working backwards from revenue to the leading indicators, knowing which levers move them, and accepting that those levers keep changing."

What AI changes, and what it doesn't

AI is the newest lever everyone is reaching for, and Eugene has run it through the same test as any channel: work backwards from revenue, spend where the ROI is provable. But reality has shown to be irrational, and change is overdue.

"Two things are happening in the US at once, and they pull against each other. Hypergrowth companies are token-maxxing: giving each hire a token budget alongside salary, betting this leads to more code shipped. Meanwhile, the AI-slop reckoning is setting in. For deterministic use cases like coding and customer support, the models are good. Further from that, in go-to-market, there's no playbook yet and the ROI gets murky. So the office of the CFO is about to get involved: per-user budgets, caps, justify-to-exceed, not to cut spending but to make it predictable. An inflection point."

Stay tuned for more from the portfolio and the people building it. For another operating perspective, read Tilen Kegl's article on scaling your Content Engine with AI.