Instagram Tests Real-Time Algorithm Customization

Instagram began testing a real-time algorithm customization feature, according to a MediaPost report published June 30, 2026.

The test would let users reshape feed-ranking signals during a session rather than rely on the static interest categories Instagram introduced in earlier updates. Meta has not confirmed the test's scope, the surfaces it covers, or the markets included in the trial.

What does "real time" actually change?

Real-time controls differ from Instagram's existing "Interested" markers and topic follows in two measurable ways: latency and granularity. Current tools let users push content categories up or down through implicit signals like saves, follows, and dwell time. A real-time interface would expose those levers directly inside the app.

Practitioners tracking reach should expect variance in any campaign that runs while the test is live, since individual user feeds will deviate further from baseline ranking. Engagement benchmarks built over the past four quarters will not transfer cleanly during the trial window.

Why would Meta test this now?

Algorithm-transparency pressure has mounted across the platform's major markets. EU and UK regulators have pushed Meta on ranking opacity for three years; competitor TikTok now ships user-facing feed controls that Instagram does not match. A real-time dial answers both audiences without requiring Meta to publish its ranking model.

Meta has not commented publicly on the test as of the report date, and no Meta executive has confirmed the trial on record.

What should practitioners monitor?

Three signals will determine whether the test graduates to a full rollout:

• Surface coverage. Whether customization applies to Reels only, or extends to Stories, Explore, and the main grid.
• Granularity. Sliders for broad content categories, or per-creator weighting that lets users pin or mute specific accounts in real time.
• Persistence. Whether user-set preferences decay after a session or carry across logins.

Brands running reach-based campaigns during the test window should isolate creative variables and budget for CPM variance of unknown magnitude. Influencer teams will need to brief creators on feed-shuffling resilience, since follower-only reach is no longer a stable metric under real-time customization.

What does the reporting leave unanswered?

MediaPost's report did not specify geography, participant pool size, or whether creators have any visibility into the new controls. The feature does not appear in any public Meta roadmap documentation as of June 30, 2026. Third-party teardowns of recent Instagram builds have not surfaced UI elements matching the test description.

The development signals a broader platform shift toward user-facing algorithm transparency, and any campaign running in late June 2026 will function as a benchmark case once Meta publishes post-test metrics.