Why should I compare each post against my own median, and how?

Compare each post with the median of your own recent posts, metric by metric, rather than with an industry or platform average. An engagement rate that is excellent for one account can be alarming for another, and the only honest baseline is what your account normally does.

Use the median rather than the mean, because one unusual post drags a mean around and barely moves a median.

Published benchmarks average across accounts that have little in common with yours: different sizes, sectors, audiences and posting habits. Measured against them, a small specialist account can look like a failure while doing well, and a large one can look healthy while slipping.

They also answer the wrong question. "Is this post above average for Instagram?" tells you nothing you can act on. "Did this post do better than what we normally do?" tells you whether to do more of it.

The median is the middle value when you sort your posts from lowest to highest. The mean adds them up and divides. With social numbers the difference matters, because one post that travels far outside your usual audience can be many times larger than the rest.

Take nine posts with reach between 7,000 and 11,000 and one with 80,000. The mean is pulled up to well above anything the other nine achieved, so nearly every post looks below average. The median stays near 9,000, which is what a normal post actually does.

1. 1Pick the metrics that match the goal. Reach for awareness, saves and shares for usefulness, comments for conversation, clicks or sign-ups for a hard ask. Three is plenty.
2. 2Read every post at the same age. A post read after one day and a post read after a week are not comparable. Pick an age, such as seven days, as a rule of thumb, and stick to it.
3. 3Choose the window. Your last ten to twenty posts on that account is a sensible rule of thumb: recent enough to reflect your current audience, long enough to be stable. Leave out the post you are judging.
4. 4Take the median of each metric. In a spreadsheet, MEDIAN over the column does it. Keep a separate median per platform; an Instagram post and a YouTube video are not the same thing.
5. 5Express the post as a difference. (post ÷ median − 1) × 100 gives the percentage above or below. That is the number worth reporting.
6. 6Wait for enough history. With only two or three measured posts, a median is barely a baseline. Say "not enough history yet" rather than reporting a comparison you cannot stand behind.

An Instagram post, read at seven days, against the median of the account's previous fifteen posts. These figures are illustrative.

• Reach: 12,400 against a median of 9,000, so 38% above.
• Saves: 214 against a median of 231, so 7% below.
• Comments: 38 against a median of 22, so 73% above.
• Reading: the post travelled and started conversation, but people did not keep it. The next post in that pillar might test a more practical format that people will want to save.

A spreadsheet with one row per post is enough: date, platform, format, pillar, ask, the age at which it was read, and one column per metric. Below it, a MEDIAN formula over the previous rows for each metric, and a difference column beside each post. Filtering by platform before taking the median keeps channels apart.

Once you have a few campaigns in the sheet, take medians within groups as well: all carousels, all posts in one pillar, all hard asks. A reel that beats the account median but trails the median for reels tells you something different from one that beats both.

Read small differences with care. A post a few per cent either side of the median did what the account usually does; that is not a result to act on. Look for posts well outside the usual range, and for patterns across several posts, before changing the plan. One post is an anecdote; five posts in the same direction are worth a conversation.

Follower count changes the picture over time. If the account has grown a lot, an old post's reach is not a fair comparison, and a shorter window, or dividing each metric by followers at the time, keeps the baseline honest.

CampaignFlow measures each post against your own median, not an industry average. Posts on Instagram, Facebook Pages, YouTube and Threads are re-read when you sign in, and new numbers are appended rather than overwritten, so the history a median is taken from stays intact.

Analytics are not metered on any plan. See the features page, or how to plan a campaign for where measurement fits in the run.