Infinite scroll without breaking crawl

Context

Picsart is a creative platform and social editing app with over 150 million monthly active users. Photo, video, design, AI tools. Web and mobile.

I joined as Senior SEO Specialist from October 2020 to June 2021. I was the only person on the team focused on technical SEO.

Explore and hashtag pages sat at the center of discovery. Millions of dynamically generated URLs. Product wanted them to feel like a modern feed. Search needed them to behave like a crawlable archive.

Those two goals usually fight.

Problem

Infinite scroll was a product and CEO request. Users should keep browsing without hard page breaks.

Search engines still need pagination signals, stable URLs, and crawl paths that do not disappear behind client-rendered “load more.” Skip that, and you trade short-term UX for long-term indexation debt.

On a content graph this size, debt compounds fast. Crawling efficiency drops. Indexation gets noisy. Organic growth stalls even when demand is there.

The job was not to win the argument for classic pagination or for pure infinite scroll. It was to ship both: the experience product wanted, and the crawl graph Google needed.

Approach

I sat between product requirements and engineering constraints.

We designed a hybrid: infinite scroll for people, pagination and clear crawl paths for bots. Engineers owned the implementation. I owned the SEO contract: what must remain crawlable, how page relationships stay discoverable, what “done” meant before launch.

Then I built monitoring workflows so technical health was not a one-time audit. Crawl efficiency, indexation, and performance across millions of dynamic pages needed ongoing visibility, not a launch checklist that goes stale.

The concept was simple. Product UX and SEO are not a tradeoff if the architecture encodes both.

Outcome

3.2× clicks on Explore and hashtag pages after the technical work landed.

More important than the multiple: the content ecosystem gained a stable crawl model. Organic growth had a technical foundation that could scale with new hashtags, new Explore surfaces, and ongoing product changes.

For a team that had treated technical SEO as intermittent firefighting, that was the shift. From “fix the latest crawl issue” to “keep the graph healthy by design.”

What I’d do differently

I would have published a one-page SEO contract for feed-style templates earlier: infinite scroll rules, pagination requirements, canonical policy, and monitoring thresholds. Every new Explore or hashtag experiment would inherit it instead of reopening the debate.

I would also have instrumented bot vs. human render parity in the same dashboards as indexation. Hybrid solutions fail quietly when the crawlable HTML drifts from what users see. Catching that drift weekly would have reduced late surprises.