Figure: The redesigned Gistly homepage, from the hero through to the footer.
Figure: The Gistly blog index, a grid of article cards.
Context
Gistly's homepage is the front door for a hard-to-explain product: AI that scores recorded sales, support, and collections calls against quality rubrics. The people who land on it are QA leads and sales, support, and collections managers, buyers who need to grasp a technical, category-defining product quickly, then be moved toward a demo. The page has to teach and convert at the same time.
My scope
The copy came out of a wider marketing effort. My work was everything around it: the visual design, the hierarchy and composition, and how the content was organised and sequenced down the page. I argued to keep the customer testimonial when it was up for cutting, and I built most of the page in Webflow.
The problem
The previous homepage led with a broad promise and then a wall of everything: an "unlock the magic" feature grid, a phone walkthrough of how it works, and several dashboard shots. It explained a lot but pointed nowhere in particular: the primary action competed with exploration. The heatmaps later confirmed it: on the old page, people roamed the navigation and visuals instead of moving toward a demo.
Constraints
- Mixed intent. Visitors range from "what even is this" to "ready to book a demo". One page has to serve both.
- A fast, small team. The site changes constantly, so the design had to be a durable system, not a one-off layout.
- Features that are hard to picture. A capability only lands if a visitor can see it in their own workflow, so I created an animation for each feature in Adobe Illustrator rather than leaning on static screenshots.
The redesign
Five moves that took the page from a feature dump to a guided path, from the old design to the new.
01
Focused hero with a clear CTA
The old hero opened on a vague "empower sales and collections performance" and offered several competing things to do at once. I gave it one job: say what the product is, then offer a single obvious next step. The demo action leads, and the exploratory links step back behind it.
02
Outcomes over features
The old homepage listed capabilities like a spec sheet: automated auditing, sentiment analysis, transcription. Working from the marketing copy, I sequenced the page around what you get on day one and pushed the feature detail further down, for the people who want it — the same words, ordered so the payoff comes before the mechanism.
03
One block per capability
Instead of one long undifferentiated scroll, I organised the page into distinct capability blocks (conversation QA, agent assist, voice and chat agents, observability), each self-contained and each with its own call to action, so the page guides the visitor forward rather than dumping everything at once.
04
Show proof, not just claims
A complex, unfamiliar product is bought on trust. I argued to keep the customer testimonial on the page when it was up for cutting, and gave it and a real product dashboard room to breathe, so the value is demonstrated rather than asserted.
05
A single clear path to action
On the old page the primary action competed with exploratory navigation and decorative visuals. I tightened the hierarchy so the demo and the dashboard lead, and the eye is pulled toward acting rather than wandering.
Results
Measured year-over-year across GA4 and Microsoft Clarity, excluding the revamp month to control for seasonality. Homepage key events rose 475% (12 → 69) and the key-event rate went from 0.4% to 1.39%, a 246% lift: visitors completed meaningful actions far more often. Average engagement time per session rose 47% (42.8s → 62.9s), click efficiency climbed from 39% to 62%, and on the contact page visitors became more than 5× as likely to start a form as a year earlier.
Traffic grew too (active users up 47% year-over-year), but that reflected concurrent SEO and content work as much as the redesign, so I separate it out rather than claim it as design impact. What I attribute to the redesign is the shift in engagement quality, and the heatmaps confirm it: on the new page users move straight to the primary actions, where the old one scattered attention across navigation and visuals.
Heatmaps
Figure: Click map of the old homepage. The highest-ranked clicks land across the top navigation, with the densest hotspot on the nav links rather than on the demo button.
Before · Jun–Sep 2025 · 601 views, 236 clicksFigure: Click map of the redesigned homepage. The highest-ranked clicks land on the headline and on the Get a Free Demo and Explore Features buttons.
After · Dec 2025 – Mar 2026 · 686 views, 424 clicksWhat's next
The same analysis surfaced the next problems to solve. Blog traffic grew almost ninefold from SEO, but most of it bounced: the content had weak paths into the product. And the contact flow leaked: the form dropped users, and the Calendly booking sat below the fold. I proposed a dedicated book-a-demo path separate from the contact form, a persistent "request a demo" action in the nav that goes straight to booking, and stronger in-content conversion for the new blog audience.
What I took from it
The useful discipline here was honesty about attribution. It would have been easy to headline a single big number; the more valuable thing was to separate what the design moved (how people engage with the page) from what the whole team moved (how many people arrive), and turn that gap into the next roadmap.