Why Growth Teams Shouldn’t Wait for AI-Ready Data to Chase Pipeline

Growth teams are asked for more pipeline before the data, attribution, and tooling underneath them are ready, and the standard advice is to fix the foundation first. That advice has only gotten louder as every revenue organization is working to make its data AI ready. A foundation built ahead of the market, though, rests on assumptions nobody has tested. A team that waits for it can spend weeks and months refining a theory about its buyer without hearing from one. The faster route to a sound foundation runs through live pipeline, because what leads actually do shows a team which parts of its data are worth fixing first.

Yehonatan Hahn, VP of Growth at Ascendx, runs growth across the portfolio of Salesforce ISVs the company acquires, small and profitable businesses building products on the platform. He came to the role through management consulting and a business operations and go-to-market career at the logistics software company Bringg, where the acquisition of a Salesforce ISV introduced him to that ecosystem. Hahn works with portfolio companies at very different stages of operational maturity, and that range shapes his view of how much foundation a growth team needs.

“Don’t let the foundation stop you from just going out and doing something obvious,” Hahn says. He still sets a strategy and builds a foundation, and he treats both as drafts that get revised once real leads respond.

Test first, because the market answers faster than a plan

Hahn describes the alternative to testing as an isolation room. A team sits inside it convinced the product solves a particular issue for a particular buyer, and ruminates on that idea for a very long time. Without a live test, the theory stays a theory. “A lot of growth is about trial and error,” Hahn says.

His version starts with the basics of what the product does, which problem it solves, and who will buy it. Once the team has a reasonable answer, it goes to market, collects feedback, and adjusts its assumptions based on what comes back. Along the way it learns where its pipeline comes from and what quality it’s bringing in. “Learning these things along the way is super valuable,” Hahn says.

Size the foundation to the question you’re asking

How much foundation a team needs depends on how mature the business is, and Hahn’s portfolio spans both ends. At a mature company, the growth team wants a unified picture of each lead before it entered the pipeline, including which campaigns it touched, how many, and who it spoke to. The questions get granular, covering lead source, first-touch, last-touch, and multi-touch attribution, and how many BDRs and AEs have talked to a lead compared with marketing emails alone. At a less mature company, the check is simpler. “You just say, ‘Hey, where’d you hear about us?’ And I assure you that’s 90% effective as well,” Hahn says.

“We adjust the strategy and also the scrutiny,” he says, based on where each business sits on the maturity scale. The test is to invest in the data that scale calls for. “You don’t need a Ferrari to go out and buy groceries.”

Double click into one segment at a time

Maturity also shapes how the team investigates a pipeline problem. Leaders at an early stage want total pipeline. A little later they want to know where it’s coming from. After that come response speed and the care leads get once they’re in. At each step Hahn tells the growth operations team to double click, and he’s firm about restraint. “It’s the ability to double click without over analyzing the situation,” he says.

His team compartmentalizes its questions and avoids monitoring an abundance of data fields all at once. When something looks off, it pulls out a sub-segment, such as the leads from a single source, and asks how they behave differently, what worked, and why something didn’t. “What literally happened to these 10, 20, 1,000 leads?” Hahn asks.

Let AI take the admin so selling keeps teaching the team

The data a team can learn from depends on what it captures, and Hahn doesn’t want the people closest to the buyer doing the capturing. His team uses AI to fill in fields that salespeople and BDRs aren’t necessarily recording day to day, and it tries to document as much as possible without overburdening the system. “We want their focus to be on selling, going out and finding new pipeline, not on just being admins of the system,” he says.

The stack behind this stays small. Apollo.io handles data enrichment, call and meeting recordings, and outbound email. Salesforce, including Marketing Cloud and Data Cloud, collects the data, and AI tools such as Claude scan it to keep it standardized. A team of people reviews the results, flags fields nobody has standardized, and converts them into pick lists so the problem doesn’t come back.

Hahn is wary of overlapping tools. “People that own software don’t realize what that software can do for them,” he says, so a pitch for a cool AI feature sounds great and leaves a company paying for duplicative tools. Hahn’s team maximizes the toolkit it owns and buys a specific tool only for a specific use case.

Shared credit keeps the feedback honest

A team can only learn from its pipeline if marketing and sales tell each other what they see. Hahn’s portfolio companies are small enough that this happens in the room, with marketing stakeholders in sales forecast meetings and sales stakeholders in marketing strategy. He says the split between marketing-sourced and sales-sourced leads doesn’t exist for them. Marketing touched every lead and opportunity in some way, and someone in sales likely shepherded it toward becoming an opportunity.

Hahn has promoted a joint go-to-market motion throughout his career, and he’s blunt about what credit fights cost. “I’m very anti just giving credit for the sake of people getting credit,” he says, because the room ends up arguing over whose lead it was while the lead sits there. His answer is cultural. “We get people to be more truth seeking rather than credit taking,” he says.

Hahn says incentives decide whether the two functions share information. At a lot of big companies, marketing works in a silo, reports how many leads it generated, and asks why sales isn’t converting them, while sales answers that the leads are poor. When marketing is incentivized on sales performance, he says, it gets more involved in what sales is doing. It also grows more careful about which events are worth attending and asks sales whether an event worked. “What we can’t verify in data, you verify a little bit anecdotally and see what you can move from there,” he says.

Start messy and keep a spreadsheet on the side

The obvious advice for revenue leaders, Hahn says, is to set the foundation and get the data AI ready, and he doesn’t argue with it. He adds that a team can start messy. “I’d much rather measure what we can in Google Sheets or Excel on the side as opposed to having everything ready in the system,” he explains, and that spreadsheet becomes the team’s database outside the foundation for a little while. His view holds at the big enterprise SaaS company he worked at earlier in his career and at the smaller businesses he works with now. In both, he says, it’s always a balance.

A messy start also gives the operations team something to work with. A growth team that’s moving surfaces problems, and ops hears about them from the people doing the work. “You’re in a much better situation if people are telling the ops team, ‘This is super messy, fix this,’ as opposed to everybody waiting patiently for something magical to happen,” Hahn says.