Turf

Core engineering on Turf, the all-in-one community platform SaaS companies use to boost user engagement, gather feedback and increase retention.

The problem

Most SaaS communities die quietly. A forum gets launched, a handful of people post, and within months it is a support queue with worse tooling. The difficulty is not building somewhere for people to talk — it is that a community only becomes self-sustaining above a density of activity that most products never reach.

That makes the software’s job less about hosting discussion and more about concentrating it.

Constraints

Attention is the scarce resource. Every notification, digest and prompt either brings someone back or teaches them to ignore the product, and the line between those is narrow.

Feedback also has to be more than a suggestion box. Collected feedback is worthless unless it can be aggregated into something a product team can act on, so structure has to be imposed without making contribution feel like filing a ticket.

What I built

Core engineering across the platform in Node, React and GraphQL.

GraphQL earns its place in a product like this: a community screen assembles posts, authors, reactions, and permissions in one view, and the alternative is either a chain of round trips or an endpoint per screen that ossifies as the product changes.

Outcome

Community infrastructure aimed at the actual failure mode — engagement that compounds rather than a forum that goes quiet — with feedback captured in a form a product team could use.