Introducing: Rule Scheduling in Feature Experimentation
It's time to achieve faster experimentation velocity. Schedule flags, experiments and targeted delivery rules. Turn an entire flag on or off. Play or pause a rule. And more.
Tom Burford
New to Feature Experimentation, Rule Scheduling lets you plan ahead of time and achieve faster experimentation velocity with confidence in scheduled changes. So if you’d like to release a feature at midnight, previously you would have to stay up until midnight to turn on the feature, and now you can simply schedule it to happen.
Users can schedule changes to flags, experiments and targeted delivery rules:
- Turn an entire flag on or off.
- Play or pause a rule.
- Change a rule’s rollout percentage.
For progressive feature rollouts you can automate the progressive rollout of a feature helping teams achieve smoother progressive deliveries with less opportunity for delays and hiccups.
Rule Scheduling will enable all users, including non-technical users, to schedule any rules options to their flags, rules or targeted deliveries.
Who is this for?
Product Managers: Time feature launches and automate progressive rollouts
Engineers: Maintain flag hygiene by automating flag lifecycle
Experimentation program managers: Automate your experiment calendar by scheduling experiment start and end dates
Features & Capabilities
|
Flag and Rule |
Traffic allocation |
Launch timing |
|
Schedule a Flag or Rule to toggle on/off |
Schedule traffic allocation percentages |
Scheduling your flag to turn on or off on a specified date/time allows you to plan your release without having to manually make changes at desired release times. |
|
Progressive rollouts |
Flag hygiene |
Observability |
|
Create a set of scheduled changes to your flag once, and Feature Experimentation slowly releases the feature for you. For example, you can schedule your flag to roll out to 20% of your user base tomorrow, 35% the day after, 50% the day after that, and increase that percentage until it reaches 100%. |
Plan for a flag end state ahead of time to ensure that rules are cleaned up. For example, a tech team wants to make sure that a flags rules are adjusted so that no experiment is running + rolled out at 100% after a month. This might be easily forgotten without scheduling. Instead of creating a ticket to make the change, they can schedule the end state ahead of time. |
|
What early customers love about Rule Scheduling
|
Experimentation velocity |
Smoother rollouts |
Better time management |
|
Faster experimentation velocity with confidence in scheduled changes |
Smoother progressive rollouts |
Plan experiment execution ahead of time |
Q: Is there a public API for Rule Scheduling?
A: Not yet, but this will be added in a future release.
Q: Is Rule Scheduling available for all customers?
A: Rule Scheduling is available to Scale and Accelerate customers.
Q: Can I schedule audience condition change and use public APIs to schedule changes in this feature?
A: Currently, users cannot schedule audience condition change and use public APIs, but this capability will be added soon.
Q: How close to current time and how far from the current time can I schedule a change?
A: A user can schedule a change in the next 3 mins and up to 3 months ahead from the current time.
