Product market fit. Every startup finds itself in the relentless pursuit of it, but defining it and determining when you have it can be just as challenging. Some say it's customer pull: sales cycles are faster, retention is stronger, and product adoption is high. Many of us at Gigi worked at the Amazon ad tech company Perpetua in its earliest stages roughly a decade ago, and for many of us that was the first time we felt product market fit. Once you feel it, you know it. But perhaps worse, after you feel it, you painfully know when you don't have it.

The first version of Gigi was a legacy SaaS app that unified CTV media buying and measurement on Amazon's ad tech: the Amazon DSP and Amazon Marketing Cloud. In our first year we did ok, managing over $10M in CTV ad spend, but it was painfully obvious that we didn't have product market fit. This glaring clear absence was one of the primary reasons we pivoted Gigi to the product and company many know today.

Several months after our relaunch, in early 2026, we felt the signs. People wanted to buy AI. We found ourselves in sales cycles we had never anticipated. Customer churn ceased to exist, and our DAU/WAU ratio was on the level of a consumer product. This time it was different. But it was only recently that we discovered the science behind what led Gigi to product market fit.

Mark Roberge is a Managing Director at Stage 2 Capital and former CRO of HubSpot who takes a scientific, data-driven approach to breaking down product market fit and building repeatable go-to-market engines. He turns product market fit into a formula: PMF = P% of customers achieve E event(s) within T time. For example, HubSpot's early version was roughly: 80% of customers use 5+ features within 60 days.

We tried to apply this formula to Gigi. The primary way our customers interact with our product is through tasks. When onboarding a new agency, we sit down with them, learn their manual ways of operating the Amazon DSP, and break down their work into tasks, so Gigi now performs the actions that, due to bandwidth constraints, they only aspired to execute. Our customers' job transforms from laboriously pushing buttons to reviewing, accepting, and automating Gigi's work based on the new best practices we forge together. Since this is the primary interaction mode, we thought task acceptance, a combination of the number of tasks and the rate at which our customers accepted them, would be the sign of product market fit. But we were wrong.

Our GTM team does the work of building the tasks for our customers so their Gigi agent is turnkey. We do this to eliminate whatever friction we can from onboarding. Unfortunately, this lack of friction has the second-order effect of sacrificing potentially valuable customer engagement. And, to no surprise, we found the more we did up front, the less customers engaged early on. Accepting tasks is low lift. You just need to review Gigi's work and press accept. It doesn't require much customer ingenuity, and when you don't feel like you've had a hand in building the alchemy, it's tough to feel the magic.

We saw this in the earliest stages of our pilot with WPromote. Task acceptance and task acceptance rate were low, and broader Gigi engagement wasn't much better. But one media manager took an action that changed the trajectory of our partnership. For an at-scale pet brand, she asked Gigi to update budgets for 81 orders, 211 line items, and 271 creatives. Gigi did it successfully in 4.5 minutes. She immediately messaged our team: "woah, that would have taken me all afternoon." In that moment we realized tasks weren't the product interaction that showed PMF. It was chats in which our customers prompted Gigi to take action (bonus points when those actions number in the hundreds) and watched Gigi successfully take them. That was the magical moment that demonstrated the power of Gigi.

That media manager's magical experience catalyzed a dramatic uptick in Gigi adoption at WPromote. Every media manager quickly learned to prompt Gigi and rely on Gigi to take action and transform their work. WPromote is now one of our most leaned-in agencies.

So we changed our onboarding to engender these behaviours, and we've seen a marked improvement in product adoption and value demonstration. We now measure product market fit with a formula of our own: [X]% of customers prompt Gigi to take action within their first two weeks of using the product.

You can build the tasks for a customer, but you can't feel the magic for them.

We're proud to share our case study with WPromote this week here.