Feature Request: Pre-Publish Performance Prediction AI-powered score and insights before a post goes live

[githb-ac]

Every post a social media manager schedules is a bet a judgement call that this piece of content, at this time, with this copy and creative, will perform well enough to justify the effort behind it. That judgement is currently made entirely on instinct and experience, with no data-informed signal available at the moment of scheduling. This feature request proposes a Pre-Publish Performance Prediction engine built into the TryPost composer that analyses a post before it is scheduled and returns an AI-generated performance score, a breakdown of the factors driving that score, and specific actionable suggestions for improving it giving users a data-informed second opinion at the most valuable moment possible: before the post goes live.

Summary

An AI prediction layer embedded in the TryPost composer that evaluates a drafted post against the account's historical performance data, platform best practices, content quality signals, and optimal timing to generate a predicted performance score before publishing. The score is accompanied by a plain-language breakdown explaining which elements are working in the post's favour and which are likely to suppress performance, along with specific one-click suggestions the user can apply to improve the prediction before scheduling.

Why This Matters

Performance prediction at the point of scheduling is the logical completion of TryPost's AI strategy. The platform already helps users create content with the AI copilot and analyse performance after the fact with analytics prediction closes the loop by bringing intelligence into the decision moment itself. For social media managers who are accountable to engagement targets or client reporting metrics, a pre-publish score is not a novelty feature but a genuine decision-support tool that reduces the variance in their output quality. It also creates a natural feedback loop users who act on prediction suggestions and see improved results develop a deeper trust in TryPost's AI layer and are significantly less likely to churn.

Proposed MVP

Performance prediction score displayed in the composer sidebar when a post is ready to schedule a clear numerical or letter grade with a confidence indicator

Score breakdown showing the individual factors contributing to the prediction including copy length, hook strength, presence of a clear CTA, image quality signal, hashtag relevance, estimated posting time quality, and content format fit for the platform

Plain-language explanation for each factor not just a score but a sentence explaining why that element is helping or hurting the predicted performance

One-click improvement suggestions that apply directly to the draft for example shortening the caption, adding a CTA, swapping a hashtag, or adjusting the scheduled time to a higher-performing window

Prediction history log showing the predicted score versus actual performance for past posts, allowing users to see how accurate the model is for their specific account over time and building calibrated trust in the feature

Platform-specific predictions the same post drafted for LinkedIn and Instagram receives separate scores reflecting each platform's distinct engagement dynamics

Learning loop where actual post performance data continuously refines the prediction model for each account, improving accuracy the longer the user stays on TryPost