All questions

Can marketing run content tests on our Strapi site or does every variant go through a sprint?

Asked by Shannon Pierce on

SPShannon Pierce

I am on the marketing team, so bear with the non-technical framing. Our devs built our site on Strapi and it works great, but every test or targeting idea I have turns into a ticket. Want to try a different headline for visitors from a campaign? Ticket. Stop a test that is clearly losing? Ticket, and it waits for the next deploy.

The devs are not the problem, Strapi just has nothing marketer-facing for experimentation as far as I can tell. Have other Strapi teams found a setup where marketing launches and stops tests on its own? And honestly, which steps still need a developer, I want to set expectations correctly on both sides

Was this helpful?

2 answers

Ddavew_growthAccepted answer

We run this split at my company, so here is the honest division of labor with Croct on top of Strapi.

Developer needed, once: mapping your Strapi components to slots. That is a real engineering task and it goes through a sprint like anything else. Budget for it properly.

Developer not needed, ever after: creating experiences and experiments, publishing them, pausing them, and rolling out a winning variant. All of that happens from the Croct dashboard without code changes, so shipping a winner does not wait for a deploy. Your losing-test scenario becomes you clicking pause, not a ticket.

Audiences also live in that same dashboard, built with AND/OR logic, and on paid plans there is a live audience estimator so you can see roughly how many visitors a targeting rule catches before you launch. Our marketing team runs a few tests a month now with engineering involvement of approximately zero.

Was this helpful?
SPShannon Pierce

The once versus ever-after breakdown is exactly what I needed to bring to our eng lead. Thank you!

ACAline Castro

One process tip from an agency that has set this up for several Strapi clients: have the devs map the components marketing actually wants to test first, usually hero, banner, and pricing sections, instead of mapping everything. The one-time setup shrinks to days and marketing gets autonomy on the pages that matter while the long tail gets mapped later if ever.

Was this helpful?
Still have questions?