The timelines around releases are an important part of the human process that plays out: you need to set expectations with stakeholders and make sure everyone on your team has a clear understanding of when any given release is expected to kick off, submit, release, and roll out. Most teams try to create some kind of shared release calendar, but there are so many dependencies and inevitable deviations from the happy path that it’s tedious to maintain.
Given that Runway has all the context needed to pull together an accurate picture of release schedules and milestones — and keep it all updated for you — we’ve added new release schedule views in the platform. Both per release and at app level (across all releases), Runway populates a calendar with key lifecycle events like kickoff, submission, and release, as well as info on every step of a phased or staged rollout. In addition to seeing the actual timing of events as they occur and any upcoming targets that are explicitly scheduled (e.g. with a recurring release cadence you have configured), Runway can even predict events for upcoming releases based on timing of your past releases.
Many teams already rely on Runway’s rollouts functionality to build (and act on) a more complete picture of app health, combining signals from crash reporting tools, observability and product analytics platforms, and the app stores. But we know teams are using even more kinds of tools to capture data like this, and many have their own data pipelines set up in-house.
Now, you can stream any and all custom data into Runway using a new API endpoint to surface and use in the rollouts context, just like any other data from Runway’s out-of-the-box integrations. Runway will surface your custom data on time series graphs and you can create health metrics with thresholds set for alerting and rollout automations.
Even with our new release schedule views, it’s likely that folks on your team who spend less time in Runway will rely on calendars outside the platform for information on release timelines and milestones. On the flipside, folks relying on Runway’s calendars might worry that they’re missing important one-off or ad hoc events which are relevant to a release but added only to their team’s external calendar.
Our newest integration type allows you to connect your team’s external calendar to Runway to ensure everything stays aligned. In one direction, you can have Runway automatically sync your Runway release schedules to your external calendar, adding and updating events as needed. In the other direction, Runway can also pull in any additional events created in your team’s external calendar so that your Runway calendar views will always show a complete view of your schedules.
(Currently supported for Google Calendar, with Outlook on our radar — reply to this email to let us know you’re interested!)
Some of the trickiest app health issues to track down are the ones caused by feature flag changes as opposed to new binary releases. Flag changes are more dynamic and harder to stay on top of, and it can take teams quite a while to connect the dots when a flag change made by someone, somewhere on the team causes a new crash or breaks an important flow in the app.
With your feature flagging tool integrated in Runway, we now perform continuous diffing on your flags to capture not just “on/off” changes but other updates to progressive rollouts, variations, audiences, etc. In addition to surfacing the resulting diffs in feature flag lists in the platform, we also overlay flag changes on your rollout graphs. That way, if you see a spike or dip in an important metric, you have immediate context on any flag changes that were made around that time and some hints as to a possible root cause.
While Runway’s existing metric types give you plenty of ways to monitor health signals from observability and analytics tools, and alert and automate rollouts based on thresholds you set, we understand that teams often want to consider more complex aspects of app health. Certain funnels, conversions and computed values were harder to get at.
Runway now lets you build more complex health metrics based on existing building blocks. All you need to do is select one or more of your existing metrics in Runway, then enter an expression describing how those metrics should be put together. For example, to look at a conversion from user signup to checkout, you might select your “signups per DAU” and “checkouts per DAU” metrics and enter the expression “checkouts per DAU”/”signups per DAU”. Any mathematical expression is valid, and you can even combine metrics from different integrations for a powerful new way to monitor health.
When distributing pre-production builds, many use cases call for an easy way to share builds with folks who may be a bit further from your immediate team — perhaps internal stakeholders who wouldn’t otherwise spend time in Runway, or even people outside your company. At the same time, you don’t want to overprovision access or jeopardize the security of builds that should be locked down to a specific audience.
Now, you have the option of enabling public access on any of your build buckets in Build Distro. When you enable this on a bucket, you’ll get a public link which allows anyone to access and install any builds in that bucket. No other buckets or areas of Runway will be accessible to users accessing via these public links.
In addition to the entirely new calendar integration type highlighted above, we added two new options at existing integration points in Runway. Dynatrace is our latest stability monitoring integration, for surfacing crash-free and adoption rates in the rollout context, and Bitbucket Pipelines is a new option for CI/CD.
