Innocent Instagram followers, mentions and creators

Every figure on this page is read from Instagram via our data provider and stored with the date it was taken. Where a reading does not exist yet we say when the next one is due rather than show an estimate. Engagement rate is (average likes + comments) ÷ followers. Growth compares the latest follower reading with the first one inside the last 30 days, once the two are at least five days apart.

Mentions are the public posts by other accounts that our searches found carrying one of the terms this brand is watched for: its hashtag, the posts it is tagged in and, for the busiest brands, its name. Each search returns a limited number of recent posts, so a count is a sample of what was posted, not every post on Instagram. Read it against other weeks and other brands, measured the same way. How we count. Innocent is collected every 14 days; a visit refreshes anything over a day old. Without an account the feed shows the last 7 days; an account opens the month.

Tone and subjects come from labelling each post once, on what its caption says about Innocent: one tone (positive, neutral or critical) and one subject from a fixed list, so the same thing is counted the same way every run. Every figure in that strip is a count of labelled posts. The read beside it is written by a model from the same posts, carries no figures and is marked as such. In Competitors, each brand's mentions are found by the same searches on the same schedule, and its share is its part of the posts we found for the whole set; followers and engagement are each brand's latest profile reading.

AI recommendations are the answers ChatGPT and Gemini give to the questions shoppers ask about each category Innocent sits in, asked once a month and recorded in order. Answers vary from one run and one person to the next, so each square is one answer on the date shown, not a score.