Habuild — Building the Funnel, Attribution to First Class

Product

Habuild is a wellness and habit-building platform - live online yoga, mindfulness, and fitness routines delivered through WhatsApp, YouTube, and Zoom. Today it's one of India's largest digital wellness communities, with 11M+ members across 100+ countries.

Problem

Habuild was spending on paid ads and generating real traffic, all of it landing in their database with no way to see it. There was no visibility into the one question that matters when paying for traffic: of everyone who clicks an ad, how many make it to their first class, and where does everyone else fall off. They were capturing data, but incorrectly, and couldn't leverage it.

My role

Owned funnel instrumentation and analytics end-to-end on Scale Labs' consulting engagement, working directly with co-founders Saurabh Bothra and Anshul Agrawal. Diagnosed where the funnel was leaking and shipped the fixes that closed the gap.

Scope

Event instrumentation, ad-to-signup attribution, funnel dashboarding, and conversion fixes across the full top-of-funnel journey - ad open through first live class. Payments sat outside scope: Habuild's payment flow at the time was entirely manual and separate, with limited data. Built a series of analytical views covering:

• Attributions
• Full Funnel
• Retention
• Cohorts

Approach

Mapped and instrumented the full funnel end to end, attributing each user back to the paid source they came in on: ad opened, site visit, intent signals (WhatsApp group, YouTube), register, first class. A number of concrete events and CTAs weren't being captured, so the funnel had blind spots before it could even be trusted - one example: the Register CTA sat in four spots on the site and all four fired the same event with nothing to differentiate them, collapsing four distinct placements into one number.

Once the funnel was properly instrumented and clean, one step stood out: people clicked Register in decent numbers, but clicking it didn't sign them up - it opened an embed where they then had to fill in their details, and very few followed through once that widget opened. The intent was there; the step between intent and completion was leaking most of it.

Fixes shipped to close the gap

• Put the form where the intent was - a simple name-and-phone field on the home page with a free-class tag, instead of routing users into the embed to register
• A push notification on every site open, prompting registration for a free class while the user was already there
• Reworked the information architecture so the onboarding path was straightforward and the main CTAs were obvious
• Embedded a sample class video on the home page so a first-time visitor could see what a session looked like before committing

Outcome

Week on week, users joining their first yoga class went up 20%. This was a top-of-funnel win specifically - getting more people through attribution, past the register leak, and into their first class. The complete analytical view stack also gave the team ongoing tracking they hadn't had before.

Habuild has since scaled to ₹300 Cr in annual revenue (latest, tentative) and 11M+ members across 100+ countries - a trajectory that reflects the company's full arc across a simple cohort- and habit-focused product, content, and go-to-market, not this engagement alone.