Instagram competitor analysis from public data, done honestly

Instagram competitor analysis usually starts with the biggest account in the niche and a spreadsheet of follower counts. It usually ends there too, because a follower count tells you almost nothing you can act on by Thursday.

The useful version is narrower and more patient. It compares you with a handful of accounts your audience actually chooses between, reads only what public data can honestly support, and turns what it finds into posts of your own. It also looks at the ads those accounts are paying for, which Meta publishes for anyone to search and most people never open.

All of it works by hand, with a notes app and Meta's Ad Library. A tool comes in at the end, in one section, and you do not need it to follow any of the method.

Pick the accounts your audience compares you with

The account you most want to beat is rarely the one you should study. A creator with ten times your following is playing a different game, with different reach, a different budget and often a team. What works for them at their size says little about what will work for you at yours.

A better list has a few kinds of account on it, and five to eight accounts in total is plenty. More than that and the monthly routine below stops fitting in an hour.

What public data can and cannot tell you about another account

Everything in a competitor analysis comes from what Instagram shows the public, and that is less than it looks. Meta's developer documentation for Business Discovery, the route apps use to read other professional accounts, describes the basics: followers, the number of posts, and for each post its likes, comments and views. The same page, updated in August 2026, notes that the view count includes paid views as well as organic ones, so a boosted reel can look like a popular one.

What you never see is another account's reach, saves, shares, who its audience is, or how a post did with people who do not follow it. Any tool or spreadsheet that hands you those numbers for somebody else's account is estimating them, and an estimate of a competitor's reach is a guess with a decimal point attached.

Four things worth measuring, and what each one tells you

With five to eight accounts and the public numbers, four readings are worth the time. None of them needs more than a notes app and an hour.

Posting rhythm

Count how many times each account posted in the last four weeks and on which days. Rhythm is the most reliable thing you can read off a profile and one of the most useful, because a rival who posts five times a week to ordinary engagement is telling you something different from one who posts twice and gets a lot.

Formats

Note the split between reels, carousels and single photos, and whether it moved recently. A competitor who quietly switched from reels to carousels three months ago and stayed switched is usually reacting to their own numbers, which you cannot see but can infer from the fact that they kept going.

Posts that beat their own usual

This is the reading that matters most. For each account, take the middle value of likes or views across their recent posts, their median, and find the posts that did several times better. Those outliers are the closest you will get to seeing what that audience asked for more of. A big account's ordinary post can beat your best one and still teach you nothing.

Topics only they cover, and topics only you cover

List the subjects each account keeps returning to. Two lists come out of it. Topics several rivals cover and you never do are worth one test. Topics you cover and nobody else does are your corner, and the common mistake is abandoning them to chase the first list.

How to see competitors' Instagram ads in Meta's Ad Library

Meta's Business Help Centre (2026) says the Ad Library contains all active ads shown across Meta technologies, and that anyone can view and search it. Search a brand's name or Page, set the country, and you see each ad's creative and copy. See ad details adds information such as the advertiser's About section, and a Page's entry shows when it was created and its history of name changes.

Each ad in the library carries the date it started running and the platforms it runs on, and it says when one ad has several versions. Run length is the most useful field there. An advertiser keeps paying for an ad that works and switches off one that does not, so an ad that has run for months is one they have chosen to keep funding. Several versions of one idea side by side usually means a test, and the version still running a month later is usually the one that won.

What the library leaves out matters as much as what it shows. For an ordinary commercial ad it shows the creative, the copy and the dates, and no spend, budget or results. Meta's Transparency Center (2023) says spend, reach and funding entities are published for ads about social issues, elections or politics, which are kept for seven years. Ads delivered in the UK, the EU and associated territories carry more, including the dates the ad ran, the advertiser's targeting choices and its total reach in those places, and they stay for a year after their last impression.

Turn a rival's best post into your own version

Finding a post that beat its account's usual is half the job. The other half is making something of your own from it without copying. That is partly manners and partly results: a copied post arrives second, in somebody else's voice, to an audience that may already have seen the original.

1. 01Write down its structure in one line. Set the topic aside for a minute and describe the shape: a mistake and its fix, a before and after, a list counted down, a question answered in the first line. The shape is the part you are free to borrow.
2. 02Put a subject of your own into that shape. Choose something from your own work and your own corner. If the rival's post was a before and after of a kitchen, yours might be a before and after of a morning routine.
3. 03Write the opening in your own words. The hook is the part most likely to be copied word for word, and the part that most needs to sound like you. Write three openings of your own and keep the one you would actually say out loud.
4. 04Post it, then judge it against your own usual. Compare it with your own median and ignore the rival's numbers. If it beat your usual, the structure travels and you can use it again. If it did not, you found that out for the price of one post.

Give credit openly when you are clearly answering somebody else's idea, and never lift their words, their footage or their images. A structure belongs to nobody. A sentence belongs to whoever wrote it.

A monthly competitor routine that fits in an hour

Competitor analysis goes stale quickly and goes wrong when it is done daily, because checking rivals every morning turns into reacting to them. Once a month is enough, on a fixed day, with the same list in front of you.

The last line is the discipline. One test a month, judged against your own numbers, adds up to twelve considered experiments a year. Reacting to whoever went viral this week adds up to an account that sounds like everyone else on the list.

Where Synclify fits, and what it will not tell you

This is the section about our product, and the method above works without it.

On the competitors page you add the accounts you track and see them ranked, with their feed and their outliers: posts that beat that account's own median, which is the third measurement above, done for every account on the list. Each tracked account also has a vault of its content and its habits. Reading other accounts goes through Instagram's Business Discovery, so this part needs your Instagram connected through Facebook login.

How we know: a competitor outlier is a post that beat that competitor's own baseline, and an account with too few posts to have one is named rather than compared.

The second tab reads Meta's public Ad Library for the brands advertising against you: what each one runs and how long each ad has lasted, side by side. It needs no Meta connection at all. It holds no spend, reach or results, because Meta publishes none of those for ordinary ads, and the assistant refuses to estimate them however the question is put. If you run your own Meta ads, those are read separately, with their real spend.

How we know: competitor ads carry what each advertiser runs and how long each ad has lasted, and never spend, budget, reach or cost per result.

When an outlier is worth answering, the assistant can write a script built on its structure, in your voice and on your topic, borrowing the shape and never the words. It can also speak first when a tracked competitor posts, when a rival's post outperforms yours, or when a competitor ad is worth seeing.

How we know: the from_outlier format builds a script on the structure of a competitor's outlier, in your voice and on your topic, and never reuses their words.

Questions people ask about competitor analysis