Experimentation
Server-side tests
For changes that never belonged in the browser: pricing logic, search ranking, anything decided before the page is built.
Some changes cannot be made by editing a rendered page. Ranking a search result differently, changing what a pricing endpoint returns, or swapping a recommendation model all happen before any HTML exists. A server-side experiment covers those: your backend decides what to serve, and Pertento holds the experiment, the variants and the measurement.
This is the one experiment type that needs engineering time, and it is honest about that. There is no visual editor step. What you get instead is the same experiment record, the same targeting vocabulary, the same statistics and the same automation as every other test on the account, so a backend experiment is not a separate programme with its own spreadsheet.
Because the change lives in your codebase, the deployment story is different too. There is no one-click rollout to 100 percent: the winning branch is already yours to keep, and what Pertento records is the decision and the evidence behind it.
Running server-side and client-side experiments alongside each other is normal, and the collision model treats them the same way. Two experiments that touch the same audience are checked against each other whatever layer they run in.