Threads Gives Users Control Over Feed Algorithm
Meta's Threads is rolling out a new feature called "Dear Algo" that allows users to directly influence and customize their content feed. The move responds to growing consumer demand for more transparency and agency over the content surfaced by opaque recommendation algorithms. This shifts some control from the platform's automated systems to explicit user preferences.
- The "Dear Algo" feature was directly inspired by a user-driven trend where people would post messages starting with "Dear Algorithm" in an attempt to influence their content recommendations. This provides a clear case study in leveraging emergent user behavior as a product discovery insight. - Unlike a permanent settings change, these feed adjustments are intentionally temporary, lasting only three days. This design choice positions the feature for timely events like sports games or avoiding TV show spoilers, encouraging repeat usage for momentary interests rather than long-term feed curation. - The user prompts are public posts by design, which allows others to repost a "Dear Algo" request and apply the same feed tuning to their own account. This turns a personalization feature into a social and community-building tool, allowing users to discover new content streams through others. - This approach to user control contrasts with competitors like Bluesky, which offers a "marketplace of algorithms," allowing users to choose from various custom feeds created by third-party developers. Threads' solution keeps content moderation and algorithm development centralized under Meta. - The rollout of "Dear Algo" coincides with Threads surpassing X (formerly Twitter) in daily active mobile users in early 2026 and Meta's broader strategy to invest heavily in AI infrastructure. - The feature provides an explicit, natural language layer of user intent on top of existing implicit signals like likes, shares, and dwell time that typically power recommendation engines. - While Threads has not detailed specific success metrics, the performance of such personalization features is often measured by increases in user engagement, session duration, content interaction rates (likes, replies), and long-term user retention. - Prior to this feature, a common user complaint on Threads was the feeling of a stale feed, with users seeing low-quality "hot takes" or content that was several days old, indicating a desire for more timely and relevant recommendations.
