The situation

Heatmap is a revenue-based analytics tool. It had product-market fit, a strong customer base, and the classic post-PMF problem: activation had plateaued, churn was creeping up, and trial-to-paid conversion was stuck.

The Onboarding Loop is the framework I've spent years refining with SaaS teams. It’s based on human psychology, game mechanics, and bridges the gap between marketing and product experience. Activation, retention, and expansion are not three separate problems; they're three cycles that create the conditions to move users from one loop to the next.

At Heatmap, we ran the framework against three pressure points:

1. Users couldn’t explore the product on day one
2. Users didn’t know what their data meant, or what to do with it
3. Users would use one or two features and ignore the rest

Audit & Overhaul

We cleared the runway

Before we touched the onboarding, we had to clean up the surface area. The user-experience threw every feature at every user the moment they logged in. Beginners got lost. Power users couldn't tell the difference between what mattered and what didn't. The product had been built one feature at a time for users who had been with them since the beginning. For a new user, the interface was pure confusion.

We reorganized the product around progressive disclosure: a clean path for new users, with depth available on demand for people who wanted it.

This was prerequisite hygiene, not part of the onboarding. Every one of the loops below would have leaked straight through the old UI if we'd left it alone.

Result: New users felt welcomed rather than overwhelmed

Dashboard Before

Dashboard Before

Dashboard After

Dashboard After

Settings Before

Settings Before

Settings After

Settings After

Heatmaps Before

Heatmaps Before

Heatmaps After

Heatmaps After

Activation Loop

Users couldn’t explore the product on day one

It takes about a week for the application to populate with the users real data. This is the problem every analytics product eventually hits. The product needs data to show value, and new users haven't generated any yet. Heatmap's education center tried to fill that gap by describing features the user couldn't try, which is the worst of both worlds: a lot of talking, no proof. People left before they understood what Heatmap was actually going to do for them.

We watched session recordings of brand new users to the platform. We could see them excitedly clicking on the features but not being able to use them, then leaving. Clearly new users wanted to try the product hands-on, not read or learn about it.

We added the ability to load sample store data and replaced the education with a first-run flow built around the job the user hired Heatmap to do.

From the empty dashboard, the user picks the outcome they're after:

The product then configures itself around that outcome and shows what a successful result looks like for their store before their own data has finished populating. The user sees the endpoint first, then watches their own store flow into the same shape as their data begins to populate.

The shift was from "here's the product, come back later for value" to "here's value, now watch us fit it to you."

Result: Trial-to-paid conversion moved from 18% to 35%

Retention Loop

Users didn’t know what their data meant, or what to do with it

Activation got people to their first win. The retention loop had to turn that first win into a habit.

Heatmap had the opposite of an empty-product problem here. Data was streaming in, but users had no reason to come back and look at it. They'd log in, glance around, not know what to make of all their data, and then close the tab in frustration. Nothing was pulling them back to the product between sessions.

We engineered a habit system around three things.

A weekly insight digest. Sent to every active account, summarizing what moved in their store since the week before and what to look at first. Not a usage report. A short read that told the user something they didn't already know about their own data.

In-product "what changed since last visit" moments. The first screen on every return visit surfaced what had shifted in the data since they were last logged in. The blank dashboard was replaced with a lightweight trailhead: here's what moved, here's where to look.

An AI copilot, in beta, as the visible flourish. Still not public at time of writing, but live in-product for the accounts we were tracking closely. It turned "what am I supposed to do with all this data" into a conversation the user could actually have with the product.

Underneath those three, a handful of smaller interventions from the playbook did the quiet work.

We moved re-engagement off time and onto behavior. Instead of "it's been 7 days, come back," users got a nudge that referenced something real that had happened in their own store since they'd last checked in. Relevance over schedule.

We replaced most empty states with one visible next step tied to a real outcome. Not "Nothing here yet," not "Welcome, let's explore." Something closer to "Your checkout funnel lost 4% this week. See where."

We tied every progress indicator in the product to a real result the user could recognize, a leak identified, a fix queued, a conversion lift confirmed, rather than a setup task or a login count. Users who can see progress keep moving. Users who can't assume nothing is happening and stop opening the tab.

Result: monthly churn dropped from 8.2% to 4.9%

Expansion Loop

Users would use one or two features and ignore the rest

Most users were getting real value out of one part of the product and never touching the rest. They'd found their first win, built a habit around it, and settled into a workflow that used maybe a quarter of what Heatmap could actually do for them. The adjacent workflows, Funnels, User Journeys, and custom events, were the ones the revenue lived in, and users weren't finding their way there on their own.

To be clear about the scope of my work here: the product team built those features. My job was to onboard existing users into them.

We designed in-product onboarding sequences that activated existing users into each adjacent workflow at the moment it would solve a problem they were already hitting. Each sequence mirrored the first-run flow from the activation pillar: pick the outcome, see what success looks like with sample data, watch your own store land in that shape. Users who had hit the ceiling of their original workflow discovered the next one the same way they'd found the first one.

The pattern underneath all of it: treat every new workflow as its own onboarding moment. Users who finished one loop got pulled into the next, without a sales team needing to push.

Result: subscribers grew from 785 to 1,000+, and overall revenue climbed 250% within a year

How we measured it

Every change shipped had a metric attached to it before it went out the door, tracked in Mixpanel against a baseline we set before the Sprint began. I used Claude to generate weekly reports on the three pillar metrics so the signal didn't get buried in the dashboard.

What the founder said

"Bal helped elevate the product experience with clear, practical design direction. His work was a noticeable step up from what we had before, and he played a key role in shaping decisions as we moved toward release."

Dylan Ander, Founder and CEO, Heatmap

Churn fell from 8.2% to 4.9%

Trial-to-paid climbed from 18% to 35%

Business grew 250% within the year.

If this sounds like your product

Heatmap was a process of compounding loops. A Sprint to diagnose and rebuild, a retainer to keep the loop turning as the product and the users evolved.

Most post-PMF B2B SaaS companies between $500K and $5M ARR have the same three leaks: activation plateauing, users drifting away after the first look, and expansion barely improving or not at all. The Onboarding Loop Sprint finds and fixes the biggest one in 3 weeks.

If your users aren't activating the way you expected, let's find out why.