WorkExpandi Marketing
Rebuilt Expandi's marketing site and brand, then ran it like a product: research, prototypes, A/B tests.
I joined Expandi as its marketing designer. The first job was a brand nobody had written down; the bigger one was a site that lost most visitors before they saw what the product did. I treated it as a funnel, not a gallery, and that is how I ended up in product.
- pages redesigned and rebuilt
- 35+
- sign-up abandonment
- 74% → 52%
- bounce rate
- 86% → 72%
- Expandi sells through its site and its content: playbooks, comparison pages, launch posts and ads.
- There were no brand guidelines, so every page and ad was drawn from scratch and looked it.
- Most visitors left the homepage without scrolling past the first screen.
- The product is hard to show. An automation tool has no hero shot.
- Several audiences on one site: sales teams, agencies, recruiters.
- Opinions about the homepage were plentiful; data about it wasn't.
- Wrote the brand guidelines and a design system: type, colour, components, the rules for ads and pages.
- Mapped the site's information architecture by audience, then wireframed and prototyped before building.
- Watched real sessions in Hotjar and Mouseflow, read the funnel in Google Analytics, HubSpot and Amplitude, and ran exit-intent NPS surveys.
- A/B tested headlines, layouts and calls to action in VWO, and the ads on Facebook, Google and TikTok.
- Sign-up abandonment fell from 74% to 52%, and the bounce rate from 86% to 72%.
- Scroll depth doubled, from 38% to 73%, and frustration clicks halved, from 22 to 10 per session.
- Sessions got two minutes longer, and self-serve sign-ups and MQLs grew. One finding saved budget: LinkedIn ads cost far more per lead than Facebook.
