FeatBit

Remote Configuration

Try a new setting with a smaller audience.

Forty search results look better in a demo. Production datasets may tell a different story. Pilot a new string, number, or JSON value and keep a known-good configuration ready to restore.

Try FeatBit See the solution
Aa42{ }

A practical walkthrough with search-config

The story · A larger search page, without a larger incident

Forty results look better in a demo. Production has much larger datasets.

A team wants to try a richer search page. The application supports a remote JSON configuration, but doubling the page size could make large accounts slower. A new value deserves its own pilot.

Illustrative scenario · Real product configuration

Inside FeatBit · 3 steps

Pilot a configuration change in three steps.

Treat new values as a release: define them, control the audience, and retain tested defaults.

  1. 01Step 01 / 03

    Define stable and candidate JSON.

    Create a JSON flag with two named variations: Stable search and Candidate search. Set pageSize to 20 and 40 respectively, keep timeoutMs at 1500 in both, and use Edit JSON to edit each payload. Validate the contract in application code.

    Stable pageSize
    20
    Candidate pageSize
    40
    timeoutMs
    1500

    Values can change independently of a new deployment once the application supports the contract.

    Inside FeatBitLive app · Demo
    Swipe to inspect
    Variations → JSON · Candidate search and Stable search
    Inside FeatBitLive app · Demo
    Swipe to inspect
    Edit JSON → Candidate search payload

    20 results

    Try 40 results

    One validated configuration contract

  2. 02Step 02 / 03

    Target the candidate to the search beta.

    Reference search-beta in a targeting rule and serve Candidate search to matching users. Keep the first audience small while large datasets exercise the new value.

    Audience
    search-beta
    Rule serves
    Candidate search

    Only the beta receives the larger page size.

    Inside FeatBitLive app · Demo
    Swipe to inspect
    Targeting → Search beta configuration pilot

    search-beta

    Candidate JSON

    A value change has an audience

  3. 03Step 03 / 03

    Keep the stable values ready for recovery.

    Use Stable search as the default and OFF variation. Compare latency and search completion before widening the candidate, and reject unsupported values in the application.

    Default ON / OFF
    Stable search
    Before expansion
    Review latency and completion

    If larger pages slow down production queries, restore the stable configuration and investigate.

    Inside FeatBitLive app · Demo
    Swipe to inspect
    Targeting → Stable defaults

    Slow candidate query

    Stable JSON

    Return to tested values while investigating

The next decision

Tune from observed behavior, with validation in the application.

Compare latency and search completion for the pilot. Keep a sensible local fallback and reject unsupported or malformed configuration values.

Build your own release in FeatBit

Design configuration as an application contract

Validate remote values before using them. Keep credentials in your secret store, and remember that a new setting can only activate behavior the deployed application already implements.

From idea to implementation

Build your release path.

Create multivariate flag variationsTarget configuration by audienceExplore runtime control for AI

Keep the release moving.

All solutions

Segment-Based Releases

Try a search rewrite with an opt-in beta or a regional pilot. Use the same limited audience across related feature flags.

Feature Experimentation

Compare two experiences with a stable feature flag split. Follow Exposure and Measuring through the primary metric, posterior distribution, and guardrails.

Progressive Rollouts

Start an AI-assisted checkout with a small canary. Review errors and completion before more users receive the candidate.