What an Instagram analytics tool can see
Every Instagram analytics tool is built on the same three data sources, and each one has hard limits set by Meta rather than by the tool. Knowing which source a number came from tells you what it's worth.
Can any tool actually get me this number?
Instagram exposes three tiers of data. The Graph API covers accounts you manage and is the only source of reach, views, saves and shares. The public web gives you recent posts, likes and comments for any public account, and blocks quickly. Nothing returns a list of who follows an account - those endpoints were removed years ago. Any tool claiming otherwise is either scraping, buying data, or estimating.
Every Instagram analytics product - this one included - is built on the same handful of data sources, and the Instagram API sets limits on each of them that no product can lift. That means the interesting question about any metric is not how was this calculated but which source could it possibly have come from.
There are three, and they differ enormously.
The three tiers
Instagram Insights are the performance counts Meta reports to the owner of a Business or Creator account: reach, views, saves, shares, watch time and aggregate audience demographics. They exist only for accounts you manage, are read by tools through the Graph API, and no tier of access returns them for anyone else’s account.
The Graph API, for accounts you manage. Requires a Business or Creator account, and an explicit connection you authorise. This is the only place reach, views, saves, shares and watch time exist. Impressions was on that list until Meta removed it in April 2025 and replaced it with views. It gives full posting history rather than a recent window. Insights begin when the account became professional, so full means since then, not since the first post.
The public web, for any public account. Whatever a logged-out visitor can see: profile, follower and following counts, and roughly the most recent dozen posts with their like and comment counts. It rate-limits aggressively, so anything built on it is working with a small sample.
Paid third-party data. Companies that have accumulated data over years and resell it. Broader coverage, variable freshness, and you are trusting someone else’s collection method.
What that means per metric
What each tier returns, as of September 2026.
| Data | Accounts you manage | Any public account |
|---|---|---|
| Posts, captions, hashtags | Full history | Recent ~12, then blocked |
| Likes and comments | Yes | Yes, on those recent posts |
| Reach, views | Only here | No |
| Saves, shares, watch time | Only here | No |
| Follower and following counts | Yes | Yes |
| Audience age, gender, city | Aggregate only, needs 100+ followers | No |
| Who follows the account | No | No |
Two lines in that table do most of the work.
Saves and shares only exist for your own account
They are also the signals that matter most. A save is someone deciding your post is worth returning to; a share is someone spending their own reputation on it. Instagram has said it weighs both in ranking (Instagram’s ranking explainer, May 2023); how much weight is not published, since the API reports the count and not what the ranking does with it. And neither is visible from outside.
The consequence is awkward and worth stating plainly: “best performing post” means something different for your account than for a competitor’s. For yours it can mean the post that earned the most saves against the people who actually saw it. For theirs it can only mean the post with the most visible likes and comments against total followers. Those are not the same measurement, and a report that puts them in adjacent columns is wrong in a way that looks authoritative.
Nobody can give you a list of your followers
This surprises people, so: no Instagram API returns the identities of an
account’s followers. Not for competitors, not for your own account. The older
Instagram API that listed them was shut down in 2020; the Graph API that
replaced it has a followers_count field and no edge behind it, which you can
check against Meta’s IG User reference.
So when a tool shows you a list of followers with a “fake” or “real” label against each, it obtained that list one of three ways - scraping while logged in, buying it from a data reseller, or estimating from a sample of the people who engage with posts. All three are legitimate techniques with different costs and different accuracy, and the difference matters enough that you should know which one you’re looking at.
Scraping at volume is also the one that gets accounts disabled and IP ranges blocked, which is why serious tools sample rather than enumerate.
Audience demographics are aggregate, and thresholded
Age, gender and city breakdowns are available for accounts you manage - as percentages, never as individuals, and only once the account has enough followers for the aggregate to be non-identifying. Meta’s account insights reference puts the threshold at 100 followers. Below that you get nothing, which is correct behaviour rather than a bug.
Call quotas are a different kind of limit
The limits above are about which numbers exist at all. The Instagram API also sets a rate limit on how many calls an app may make for each account in a rolling window, and past that ceiling it returns errors instead of figures until the window moves on. What that does to a tool’s numbers is a delay or a gap, never a wrong figure: a reading arrives late, or a day is missing, and a tool that fills that gap with an estimate has made a choice you should be told about. Meta publishes the ceilings and the usage headers in its rate limiting documentation.
Three questions worth asking any tool
- Which tier did this number come from? If it can’t say, it doesn’t know.
- Is this metric available for the account I’m looking at, or is it being estimated because the real one isn’t reachable?
- When a figure isn’t available, what appears - an honest gap, or a plausible-looking substitute?
That last one separates tools you can hand to a client from tools you can’t.
Northfact labels every metric with its source and what it was measured against, and shows a gap rather than an estimate when the data isn’t there. Free during beta.
Also asked
Can an Instagram analytics tool see who follows my account?
No. No Instagram API returns the identities of an account's followers, yours included. The older API that listed them was shut down in 2020, and the Graph API that replaced it has a followers_count field with no list behind it. A tool showing a labelled follower list is scraping, buying data or estimating from a sample.
What Instagram data is public for any account?
Whatever a logged-out visitor can see: the profile, follower and following counts, and roughly the most recent dozen posts with their like and comment counts. It rate-limits aggressively, so anything built on it works from a small sample. Reach, views, saves, shares, watch time and audience demographics are never public.
What happens to a report when a tool hits Instagram's API rate limit?
The Instagram API caps how many calls a tool can make per account in a rolling window. Past that ceiling it returns errors instead of figures until the window resets, so a reading arrives late or a day goes missing. It never returns a wrong figure, only a delay or a gap.