How to track hashtag performance on Instagram, week by week
Published
To track hashtag performance on Instagram, pull the posts under the hashtag for fixed date windows (one baseline window before your campaign, then one per week), and compare the same five numbers each time: how many posts appeared, how many different accounts posted them, the median likes and comments, how many of those accounts are new to the tag, and which posts led. Instagram's app shows you the posts under a tag. It does not show you how that tag is doing over time, so the history is something you build yourself, one window at a time. This guide walks through the whole method, says what each number can and cannot tell you, and gives you a sheet layout you can rerun every Monday.
Two things people mean by hashtag performance
The phrase covers two different questions, and they need different data.
- How did the hashtags on my own posts do? That is about posts you published, and the delivery data behind it belongs to Instagram. Only the account owner can see what Instagram's own Insights offers for those posts, and no outside tool can rebuild it.
- How is this hashtag doing? That is about a tag as a whole: a branded tag, a campaign tag, an event tag or a niche you watch. Everything needed to answer it is on the public posts under the tag.
This guide is about the second question. If you run a campaign hashtag and want to know whether people picked it up, kept using it, and engaged with what they posted, you are in the right place.
The numbers worth tracking
Five numbers carry almost all of the signal. Each one is computed from the posts themselves, so you can rebuild it from any list of posts that records a date, an author and the like and comment counts.
| Number | How to compute it | What it tells you | The trap |
|---|---|---|---|
| Posts in the window | Count posts dated inside it | Whether people are using the tag at all | One busy account can post twenty times |
| Unique accounts | Count distinct authors in the window | How far the tag has spread | Your own team's accounts inflate it |
| New accounts | Authors not seen in any earlier window | Whether the tag is still recruiting | Needs every earlier window kept |
| Median likes and comments | The middle value per post | What a typical post under the tag earns | Leave posts with hidden counts out of it |
| Top posts | Sort the window by likes, then comments | What content is working | One viral post says little about the tag |
Use the median. Engagement under a hashtag is lopsided: a handful of posts take most of the likes. One post from a large account can double the average for the week while the typical post earned exactly what it did before. The median moves only when the typical post changes, which is what you are trying to see.
Reach and impressions are missing on purpose. They are Instagram's internal figures and are not on public posts, so any tool that quotes them for a hashtag you do not own is estimating.
How to track hashtag performance on Instagram, step by step
- Pick one hashtag and write down the question. "Did people use #ourlaunch after the campaign started?" leads to a different chart than "Is #ourlaunch growing month over month?" The first needs a before and after. The second needs a long run of equal windows.
- Fix the window length before you collect anything. A week suits most campaign tags. A day suits an event. A month suits a slow niche. Every window must be the same length, or the comparison is meaningless.
- Take a baseline. Collect the window just before your campaign started, so the first post-launch week has something to be compared with. A tag nobody used before launch has a baseline of zero, and that is a useful number too.
- Collect each window by date. "Everything posted between Monday and Sunday" gives you a week. "The newest 500 posts" gives you a week on a quiet tag and an afternoon on a busy one, and your chart then measures your collection method rather than the tag.
- Store every window, including the empty ones. A week with no posts is a result. If it is missing from the sheet, the chart quietly joins the two weeks around it and the drop disappears.
- Compute the five numbers per window and put them in one row each, so the history reads top to bottom.
- Read the top posts by eye. The numbers say that something changed. The posts say what.
Setting up the sheet
One tab holds the posts, one row each, with every window pasted underneath the last. A second tab holds one row per window. These are the columns that matter on the posts tab:
| Column | Holds | Example |
|---|---|---|
| A | Date posted | 2026-09-29T14:02:11Z |
| B | Window label | 2026-W40 |
| C | Author | an account name |
| D | Likes | 41, or empty when hidden |
| E | Comments | 3, or empty when hidden |
| F | Post link | the post's URL |
On the summary tab, with the window label in column A, the formulas below work in Google Sheets and in recent versions of Excel:
- Posts in the window:
=COUNTIF(Posts!B:B, A2) - Unique accounts:
=COUNTA(UNIQUE(FILTER(Posts!C:C, Posts!B:B = A2))) - Median likes:
=MEDIAN(FILTER(Posts!D:D, Posts!B:B = A2)) - Median comments: the same, on column E
Leave a hidden count as an empty cell. MEDIAN skips empty cells, so a post whose likes are hidden simply drops out of the median. If you type a 0 in that cell, the sheet now claims somebody got zero likes, and your median falls for a reason that has nothing to do with the tag.
For new accounts, add a column on the posts tab that checks whether the author appears in any earlier row: =IF(COUNTIF(C$1:C1, C2) = 0, 1, 0) copied down, then sum it per window. That works because each window is pasted below the previous one, so "earlier rows" means "earlier weeks".
Reading the result
A few patterns come up again and again:
- Posts up, unique accounts flat. The same people are posting more. Fine for an internal team tag, a warning sign for a campaign that needed reach.
- Unique accounts up, new accounts falling. The tag spread and has now stopped recruiting. Most campaign tags look like this a few weeks after launch.
- Median likes up, posts flat. People are not posting more, but what they post is landing better. Look at the top posts to see why.
- A spike in one window and nothing either side. Usually a single event or a single large account. Check the top posts before calling it growth.
Compare windows. One post under your tag doing well is good news about that post. A tag is doing well when the typical week gets better and stays better.
When the manual way is enough
You do not need any tool for a small tag. If your hashtag gets a few posts a week, scroll its page in the app once a week, copy each new post's date, author and like count into the sheet above, and you have the same history. It takes a few minutes and costs nothing.
The manual way stops being enough when:
- the tag gets more posts per week than you will realistically copy by hand;
- you need the exact date and time of each post rather than "2 days ago";
- you track more than one tag, or more than one campaign at once;
- you want the weekly pull to happen whether or not somebody remembers to do it.
Before collecting anything, the free hashtag checker reads a tag and tells you how fast it moves, as posts per hour over the time span it measured, plus how much of it is carousels or video, how likes and comments are spread, and which tags travel with it. That one reading is often enough to choose the window length in step 2.
Collecting the windows automatically
The step that costs the most time by hand is step 4: getting every post from one date window, with its date, author and counts, into a list. The Instagram Hashtag Scraper is one way to do that. It is an actor on the Apify Store that reads one hashtag and writes its posts, newest first, to a dataset you can download as CSV or Excel and paste straight into the posts tab.
| Input | What to set for weekly tracking |
|---|---|
hashtag | Your tag, with or without the # |
onlyPostsNewerThan | The start of the window, as a date or as 7 days |
onlyPostsOlderThan | The end of the window, when you are backfilling an earlier week |
resultsLimit | A ceiling you are comfortable paying for; it is exact |
Each row carries takenAt, ownerUsername, likeCount, commentCount and the post url, which are columns A, C, D, E and F above. A hidden like or comment count comes back as null, never as 0, so it lands as an empty cell and the median handles it correctly without any cleanup.
It costs $0.0005 per delivered post, and nothing else: no fee per run and no fee per hashtag. A quiet week that finds three posts costs $0.0015, and a week that finds none costs nothing while still reporting that it ran. The run stops reading once the feed passes your start date, posts beyond your limit are not charged, and duplicates within a run are dropped and not charged. You can schedule the actor on Apify to run every Monday so the window arrives without anyone remembering.
The trade-offs are real. It needs a session cookie from an Instagram account you control, because Instagram serves hashtag posts only to signed-in requests. It reads one hashtag per run. It does not collect reels, comments, profiles or reach. The limits page lists everything it will not do, and the getting started guide shows the two cookie values it asks for.
The bottom line
Tracking hashtag performance on Instagram comes down to equal date windows and the same five numbers each time: posts, unique accounts, new accounts, median engagement and top posts. Collect by date, keep the empty weeks, and leave hidden counts blank. A small tag needs nothing more than the app and a spreadsheet. A busy one, or several at once, is where an automated weekly pull earns its place.
If you are still working out which numbers a hashtag can support at all, Instagram hashtag analytics: what the data actually gives you covers that and prices the main tools side by side. For pulling one exact date range, see how to search Instagram posts by date, and for a campaign built around submissions, how to run an Instagram hashtag contest.