Feature delivery platform Split Software announced Thursday a new integration with Google Analytics to merge two silos of performance data into one. The new service combines website analytics data with code performance data to make it easy to see the impact of new features on key metrics like page load time, the company said.
Dave Karow, Split’s continuous development evangelist, said that new two-way data integration provides the ability to ingest data from and export data to Google Analytics. This functionality combines data that is usually in two separate systems.
“This tracking makes it easier to watch the metrics that your organization cares about,” he said.
SEE: Python programming language: A cheat sheet (free PDF)
Split’s platform ingests performance data and runs real-time statistical analyses on new features. This allows engineering teams to respond immediately to bad releases and measure changes in user experience. Split’s platform allows customers to release new features to a small set number of users, to gradually increase the rollout, and to monitor the impact on performance all in one deployment.
Karow said that it’s often difficult to catch a problem with a new feature before it is rolled out to at least 50% of users.
“It sounds bad to hear that one in five users would get an error but if you’re rolling this out to only 5% of your users, it is OK,” he said.
This progressive delivery method reduces the risk of code changes.
“Some things you just don’t know until you’re in production, like users doing things you don’t expect,” Karow said. “This allows development teams to confirm the impact of new initiatives with statistical rigor before declaring ‘done’ and moving on.”
The most important element of the platform from a developer’s perspective is that this gradual rollout can happen without a new deployment.
“Teams are literally able to change this via remote control,” Karow said.
This approach also gives technical teams performance data on new features to share with business colleagues. Having production data about the impact of new features can help development teams make the case for taking time to fix performance issues before releasing the new code to all users.
Combining data from Split and Google analytics gives development teams and business leaders more ways to collect and made data-driven decision, including:
Sending web performance, conversion, and business data from Google Analytics to Split to understand the impact of new features
Using average revenue per cart to see if a new feature improves business outcomes
Send a record of every feature flag in Split, called an impression, to Google Analytics for segmentation analysis
Send feature data from Split to Google Analytics to segment session or goal completion data by exposure to a new feature.
Instantly kill any feature in Split that causes a degradation in performance indicators measured by Google Analytics
How it works
Split’s philosophy of software engineering is progressive delivery, a modified version of continuous delivery and the next evolution for teams who use agile development, scrums, and DevOps. Progressive delivery includes canary deployments, A/B testing, and observability. Progressive delivery teams use feature flags to increase speed and decrease deployment risk and implement a gradual process for both rollout and ownership.
This phased ownership means that a dev team owns a feature when it is first released and is responsible for fixing any bugs. After the feature is released to production, the project manager owns it. When the feature is available to all users, a business team takes over ownership.
Karow said customers will now have access to the Split Tracker plug-in, which includes code to include in apps as well as documentation on how it all works.
“This integration represents a mix of new technology and really good advice from current Split customers that wanted this new capability,” he said.