Characteristic administration and experimentation is a broad set of capabilities that spans each software program supply and product administration.
With characteristic administration, engineering groups can interact within the follow of progressive supply whereby options are deployed with flags turned off, examined in manufacturing, after which progressively turned on to a progressively bigger viewers of customers till the brand new characteristic is totally launched. This mechanism can be utilized for brand spanking new options, however the most typical use case is safely managing steady updates.
Experimentation has a distinct objective. Product managers use experimentation to run A/B checks to match two related merchandise in manufacturing with KPIs corresponding to engagement, return visits, gross sales, dimension of buying carts, or telemetry on how rapidly a person can full a job.
The mix of characteristic administration and experimentation has been good for product groups. Nonetheless, as a rule, characteristic flags are the area of builders and experimentation is the area of product and advertising. That is inflicting a shift in how the 2 roles are being served.
For characteristic administration, builders are the first persona and, as such, look to characteristic administration capabilities instead and/or praise to conventional take a look at in manufacturing methods corresponding to blue-green and canary testing. This permits extra mechanisms to launch new code into manufacturing. The developer persona views a characteristic administration system nearly as if it had been a part of the applying structure or infrastructure.
For experimentation, a rising use case is personalization. The arrival of AI is enabling expertise designers to ship experiences which are based mostly on person conduct. The capabilities for A/B testing are there however they serve a extra refined objective: to tailor the expertise to the person’s wants. This isn’t the realm of builders – it belongs to classy expertise designers that work with social science and advertising specialists to create experiences that drive product utilization and retainment.
This brings us to the way forward for Forrester’s Characteristic Administration and Experimentation Wave. It straddles two very completely different personas and two distinctly completely different use instances. The evolution of those two use instances is going on quickly, however individually. For characteristic flags, it’s occurring in built-in DevOps platforms and better of breed instruments. For experimentation, it’s occurring in expertise design suites. Whereas there are glorious vendor choices that proceed to serve each use instances, by and enormous these use instances are being served by separate know-how markets.
For these causes, we’re retiring the Characteristic Administration and Experimentation Wave. We are going to proceed to cowl parts of it in different Waves. To realize purchaser perception when looking for characteristic administration capabilities, see our upcoming Wave on DevOps platforms. To realize purchaser perception on experimentation capabilities, see our Wave on expertise optimization options. To ask me questions on this alteration, shoppers can schedule an inquiry or steerage session with me.
