Content Hub UX Analytics & A/B Testing

Autismbehaviorservices.com’s Social Stories — a genuinely valuable resource library for families and clinicians — were hard to find, slow to load on mobile, and had no real measurement in place. The team could see the content mattered, but had no evidence for which formats or topics were actually working, or whether the same stories were landing the same way with parents as with clinicians.

Watching real sessions before redesigning anything

Hotjar session recordings and heatmaps on the existing Social Stories pages showed visitors scrolling past the content they’d clearly come for because it was buried under unrelated navigation — a pattern no amount of GA4 pageview data alone would have surfaced, and the clearest single piece of evidence for consolidating into one hub.

Finding the visibility problem underneath the UX problem

Screaming Frog, SEMrush and Search Console showed which specific pages were losing visibility, informing the title, URL and schema fixes (FAQ, Video, WebPage) applied during the rebuild. Moz’s domain authority tracking against comparable autism-resource sites gave a baseline for whether the rebuild was closing the authority gap, not just moving rankings short-term.

Writing for two different readers on the same page

Families and clinicians come to the same Social Story wanting different things — plain, reassuring language for parents, clinical context for practitioners — so content across the hub was restructured to serve both without either group wading through the other’s version.

One hub instead of scattered pages

Consolidating videos, downloads and articles into a single WordPress/Elementor hub structure made both the user experience and the SEO signal (internal linking, topical clustering) far stronger than the previous scattered layout.

Testing with a result, not a guess

Title and CTA variants were run as proper A/B tests on GA4 event data (cleaned via SQL into Looker Studio), each evaluated to at least 90% statistical confidence before a winner was called — enough tests came back with no significant difference that “no change” was sometimes the honest, correct result, not every test a win worth reporting.

Results (4 months)

• Reached Page 1 on Google for “social stories”
• Organic traffic grew from roughly 4.8k to 13k+
• Lower bounce rate and higher CTA click-through on the statistically significant winners
• Hotjar-informed layout fixes directly reduced the scroll-past rate on the hub’s most-visited stories

Reflections

The session recordings did more work than the A/B tests themselves — they narrowed down what was worth testing in the first place, instead of running tests on hunches. Reporting the tests that found nothing significant alongside the ones that won was what made the dashboard credible to the team, rather than just a highlight reel.