Erez Rusovsky is a director of product management at CloudBees and one of the former co-founders of Rollout.io, a ground-breaking feature flag management platform acquired by CloudBees and evolved into CloudBees Feature Management today.
This is the third and final installment in a series of blogs by Erez. Keep reading to learn his thoughts on what we can expect to see in the world of feature management—in the near-term as well as the long-term.
Where Are Feature Flags Headed?
I think data is going to be a key play—being able to track and see the impact of a feature across the system performance side, an attribute, new feature deployments, business KPIs, user analytics, all of that. At the end of the day, one of the biggest struggles the software development team or product team has, is being able to attribute value to the work that they've invested in. Feature flags allow you to do that.
I think data is a big piece of the future of feature flags, but the integration with the broader software delivery platform will also be important. That's what we're doing at CloudBees. We're creating a software delivery platform. The idea is to cover all of these software delivery activities—the CI, the CD, the feature flags, having them be natively integrated. At the end of the day, nobody wants to do feature flags. People want to do software delivery; that's what they care about. They want to deliver software to their customers as fast and as meaningfully as possible, and they want to be able to learn from their data. I definitely think that the shortened feedback loop that feature flags have is going to get more and more momentum.
To summarize how I view this, definitely the platform play is key—feature flags as part of the software delivery toolchain. Then there’s the data piece, being able to actually measure impact by data. The last piece, which is the most interesting and probably the hardest one, is the automation piece.
Far in the future, I think a lot of feature flags are going to be automatically deployed based on data. When you have a new feature, it's going to be opened to a subset of the user base, and metrics are going to be in place. If it's behaving well, it's going to continue with the rollout until it hits a snag or it's going to reach 100% deployment. If something bad happens, it's going to roll back automatically. If you want to be science fiction for a second, the automation would even go to the side where the flag is automatically removed from the code after it serves its purpose.
This is more of a vision statement than a prediction for the near future, but I think we could get to a point where the software is adaptive. It turns features on and off based on who the user is, what they are doing and how the system behaves for them. I think that's the future of software, although again, I think it will take more than a few years.
To learn more about how and why to add feature flags to your CD processes, check out the CloudBees whitepaper, “Grow Continuous Delivery Maturity Using Feature Flags.”