Track competitor social media: what is visible, and what is invented

Someone asks how a competitor is doing on LinkedIn. An hour later you have twelve browser tabs, a screenshot in a Slack thread, and no way to tell whether they publish twice a week or twice a quarter. Next month the same question comes back and the hour is spent again.

This page covers how to track competitor social media honestly: what is visible without any access, what no tool can see about them, and how to turn the visible part into a benchmark you can update in minutes. For founders, marketing leads and agencies.

TL;DR

What is actually visible

Everything a competitor publishes is public by construction. That gives you their cadence, their formats, their topics, their posting days, and the public engagement each post received. It is a lot, and it is the part that answers the strategic questions.

Follower counts are public too, so their audience trend is sampleable. What you get is a curve, not a diagnosis. Read follower growth rate for why the percentage misleads on small accounts.

What no tool can see, whatever the pricing page says

Reach, impressions, saves, shares, profile visits and audience demographics are shown to the account owner behind their login. Meta's developer documentation requires an access token from the Instagram business user to read insights such as impressions and reach. Your competitor did not give anyone that token.

So any product showing you a competitor's reach is modelling it, usually from public engagement and follower count. That can be a useful estimate. It is not a measurement, and treating it as one is how a strategy deck ends up with a fabricated number in it.

If a tool shows a competitor metric that the competitor themselves has to log in to see, it is an estimate. Ask for the method.

What to benchmark, and what to ignore

Benchmark the things that are both visible and actionable. Cadence tells you what the category considers normal. Format mix tells you where they are investing production time. Public engagement per post, divided by followers, gives a rough comparison that survives different account sizes.

Ignore anything that needs their analytics. Ignore virality: one post that travelled says nothing about a program. And ignore raw follower counts as a target, because you cannot see how or when they were acquired.

The line that keeps this clean

Reading a public profile is what every visitor does. Collecting it at scale still has rules worth respecting: read what is public, do not attempt to access anything behind a login, do not build profiles of private individuals, and check the current terms of each platform before you automate anything.

Terms change, so treat any statement about them, including this one, as something to verify at the time you act on it. Where the public line sits, network by network, is in public social media metrics.

How to choose your approach

• One or two competitors, occasionally. Do it by hand once a quarter. A tool is not worth the setup.
• A category you track continuously. Sample their public profiles daily, and keep the history. The value is in the trend, and the trend needs consistent sampling.
• An agency pitching new business. Public benchmarks make a strong pitch, because a prospect can check every figure themselves.
• Anyone reporting to a board. State clearly which numbers are measured and which are estimated. See founder led marketing metrics.

A competitor benchmark in twenty minutes

You do not need a project for this. Pick three competitors, and for each one answer five questions from their public profiles.

1. Who publishes? The brand account only, the founder, or several people. This tells you what you are actually competing with.
2. How often, over the last month? Count the posts. Do not estimate, and do not trust the impression that they are everywhere.
3. Which formats? Text, carousel, video. Where the production effort goes is where their bet is.
4. What is the median engagement? Ignore their best post. The median tells you what normal looks like for them.
5. How fast is the audience moving? One follower count today, another in thirty days.

Write the answers in five rows. That table is worth more than any tool report, because every figure in it can be checked by anyone in the room in two minutes.

By hand, or with a tool

The honest split is that a snapshot answers a question and a trend changes a decision. If nobody has asked the question twice, do it by hand.

Turning the benchmark into a decision

A benchmark that changes nothing is a hobby. Three decisions come out of this data, and they are the only reason to spend an hour collecting it.

1. Set your own cadence. If the category publishes twice a week per person and you publish twice a month, the gap is your plan. Size it against your worst week, as posting cadence for teams sets out.
2. Decide where not to compete. If three competitors are all producing daily video and you have no video capacity, the useful conclusion is the format they are ignoring.
3. Give your own team a reference point. Someone who sees that the category publishes twice a week stops treating twice a week as an unreasonable ask.

None of those decisions need their reach. They need cadence, format and who is publishing, all of which are readable from a logged out browser in an afternoon.

Three ways competitor tracking goes wrong

• Copying a cadence without the capacity. A competitor with a content team of four is not a model for a founder publishing alone. Read their rhythm as evidence about the category, not as a target for one person.
• Chasing their best post. You see the outlier because it travelled. You never see the forty posts that did not, so the sample you are learning from is the least representative thing they published.
• Tracking too many of them. Three competitors give you a category. Twelve give you a spreadsheet nobody opens, and the hour it costs every month is real.

The good version of this work is small and repeated. Same three accounts, same five questions, same day each month. The trend then does the arguing for you, which is exactly what a one off screenshot never manages.

How often to refresh the benchmark

Cadence and format move slowly, so watching them weekly produces noise and a chore. Follower counts move daily, which is the one thing worth sampling often, because a curve built from two points a quarter apart hides everything that happened between them.

The practical version of that table is one automated reading and one calendar entry. Everything on the daily row should be collected without a person, and everything below it is twenty minutes once a month with the same five questions.

Capturing the readings is the continuous half of the job. Turning them into an argument, four times a year, is in social media competitor analysis.

How Groowth handles this

Any public profile can be added, including one you do not own. Groowth reads it once a day, counts the posts it finds, and keeps the follower history, so a competitor sits on the same board as your own team with the same numbers next to it.

It shows nothing that requires their login, on purpose, and it estimates nothing. There is also a free audit that reads a single public profile with no account at all, which is the fastest way to see what public data gives you before deciding anything.

Frequently asked questions