Three reputable 2026 benchmark reports put Instagram's median engagement rate at 0.30%, 0.48% and 1.62%. All three are correct. If you're about to compare your numbers against an industry benchmark, that's the first thing to understand — and almost no audit template mentions it.
The benchmark trap
Engagement rate has no single definition. Different research publishers divide by different things:
- By followers — the classic formula, and the one that punishes large accounts
- By reach — how many people actually saw it
- By impressions — total times it was displayed, including repeats
The same post produces three different percentages depending on which you use, and each is legitimate. Which is why Instagram appears in 2026 reporting at anything from around 0.30% to 1.62% depending on the source, and why comparing your dashboard number to a report using a different denominator tells you nothing at all.
Rule one of auditing Before comparing yourself to any benchmark, find out what it divided by. If you can't, the comparison isn't evidence — it's decoration.
The practical consequence: your own trend is more informative than any industry median. Is this month better than last month, on the same metric, calculated the same way? That question is answerable and actionable. "Are we above average?" mostly isn't.
What the 2026 numbers actually show
With that caveat firmly attached, the directional picture across platforms is consistent enough to plan around.
| Platform | Direction | What it means for your audit |
|---|---|---|
| TikTok | Highest engagement of the major platforms, easing slightly from its peak | Still the strongest organic reach if your audience is there |
| Rising — one of the few platforms improving | Under-invested by most B2C brands and some B2B ones | |
| Declining, propped up by Reels and carousels | Format choice now matters more than posting volume | |
| Facebook Pages | Historic lows; organic reach reportedly under 2% of followers | The clearest cut candidate for most brands |
| Facebook Groups | Reaching a large share of members, unlike Pages | If Facebook matters to you, the Group is the asset |
| X | Rebounded from a very low base, still lowest overall | Justifiable only for specific audiences or real-time relevance |
One aggregate finding worth carrying into the audit: analysis of tens of millions of posts found overall engagement falling roughly a quarter year on year. If your numbers dropped in 2026, the baseline moved — that's context, not an excuse, but it does mean a decline isn't automatically a failure of execution.
Audit the presence, not the profiles
Here's the structural gap in every checklist we reviewed. They audit the accounts you control, which is the part you already see daily and often the least influential part of your public presence.
What someone actually encounters when they look you up on a platform includes:
Accounts you forgot. The Pinterest profile from 2019, the regional account someone set up, the old brand-name handle. These still surface in search, still carry your logo, and still look abandoned. This is the single most common finding when anyone audits properly, and it's the easiest to fix — reclaim, redirect, or delete.
What your employees post. For most B2B companies this out-reaches the brand account by a wide margin. It's also uncontrolled, which is fine — the question is whether anyone has given them anything worth sharing. That's the substance of employee advocacy as a reach strategy, and it belongs in the audit whether or not you run a formal programme.
What customers and communities say. Search your brand name on each platform as an outsider would. What comes up — reviews, complaints, comparisons, questions nobody answered? This material increasingly feeds the AI answers buyers see when researching you, which makes it commercially consequential rather than merely reputational. Community threads are a large part of that picture, as covered in where forums and video sit in search now.
Impersonation and squatting. Accounts using your name that aren't yours. More common than most brands realise, and worth checking annually at minimum.
None of this requires a large budget to fix, which is worth saying — most of it is attention rather than spend, and that asymmetry is generally where smaller operators have the advantage, as set out in how smaller brands compete against bigger competitors.
Publer
Multi-platform social media scheduler with AI writing and content recycling features
Best for: Multi-Platform Scheduler
The checklist, organised by what changes a decision
Everything below earns its place by potentially changing what you do. Anything that can't isn't audited.
1. Inventory and consistency
- Every account bearing your name, including dormant and regional ones — with an owner and a login for each
- Profile completeness: bio, link, contact route, category, verification status
- Visual and naming consistency across platforms
- Whether the link in each bio goes somewhere current and works on mobile
That last one fails surprisingly often, and it costs you every click you did earn.
2. Performance, measured against yourself
- Your own engagement trend over twelve months, calculated consistently
- Performance by format, not just by platform — video, carousel, static, text
- Your top ten and bottom ten posts, and what actually distinguishes them
- Posting frequency against results, to check whether volume is doing anything
The format cut is the most useful and least common. Aggregate platform numbers hide that one format may be doing nearly all the work — short-form video consistently outperforms across platforms, and on LinkedIn document and carousel posts substantially outperform plain text.
3. Findability
A section that didn't exist in social audits five years ago and should now.
- Do you appear in in-platform search for your category's terms, not just your brand name?
- Are captions, on-screen text and alt text written so the platform can index them?
- Does your content get surfaced to non-followers, or only to people who already follow you?
Social platforms function as search engines now, and content optimised only for the feed misses the people actively looking. The mechanics are in why captions and on-screen text now matter more than hashtags.
4. Audience
- Does the audience you have match the audience you want? Follower count is irrelevant if the composition is wrong
- Growth rate and, more importantly, unfollow rate
- Overlap between platforms — are you reaching the same people three times?
5. Competitors and category
- Which platforms your competitors are actually investing in, judged by output not by presence
- What's working for them that you're not doing — and, more usefully, what everyone in the category is doing that you could deliberately skip
6. Operations
- Who publishes, who approves, who responds — named people, not roles
- Response time to comments and messages, including out of hours
- Whether a crisis process exists and whether anyone has read it
- What you're paying for in tooling and whether it's used
The output should be a decision to stop something
The most valuable finding an audit can produce, and the one no template asks for.
Effort spread thinly across six platforms almost always underperforms the same effort concentrated in two. And some platforms have declined far enough that maintaining a presence is a real cost with negligible return — Facebook Page organic reach sitting under 2% of followers is the clearest example, and brands have measurably cut posting frequency there in response.
So end the audit by answering three questions explicitly:
Which platform do we stop? Not reduce — stop, or formally park with a pinned post directing people elsewhere. An abandoned-looking account is worse than a closed one.
Which platform gets the freed capacity? Usually the one already performing best, because concentration compounds.
What are we doing differently, not just more of? If the answer to weak performance is "post more," the audit hasn't found anything.
An audit that concludes everything should continue roughly as it is has usually avoided the honest question about where results actually come from.
Measuring social when attribution doesn't work
Worth being straight about, because it's why social audits so often turn into arguments.
Social's main function for most brands is influence before intent — being known, being trusted, being remembered — and that is exactly what last-click attribution cannot see. Someone encounters you on LinkedIn in March and searches your brand in July; the credit goes to branded search. Judging social by attributed conversions consistently undervalues it, and the problem is getting worse rather than better, as covered in why attribution keeps getting harder.
Three practical substitutes. Track branded search and direct traffic as a proxy for awareness that produced no click. Add a self-reported "how did you hear about us" field to your forms — the answers routinely surprise. And judge the programme on whether total business outcomes improve as social investment changes, rather than attributing individual posts.
None of these is as satisfying as a clean attribution number. All of them are more honest than one.
What not to bother auditing
- Follower count in isolation. Composition and trend matter; the absolute number is a vanity figure that correlates poorly with revenue.
- Best time to post. Endlessly researched, and swamped by content quality and algorithmic distribution. Post when you can post consistently.
- Hashtag strategy on most platforms. Largely superseded by content-based indexing. Effort belongs in captions and on-screen text.
- Vanity comparisons to much larger brands. Their engagement rate is depressed by scale and their resources aren't yours. Compare to your own past.
A workable audit, in a day
- Inventory (45 min). Every account with your name on it. Include the forgotten ones. Assign an owner to each.
- The outsider test (20 min). Private window, search your brand on each platform, write down what you see.
- Pull twelve months of your own data (60 min). One consistent formula, noted down. Trend, format split, top and bottom ten.
- Check findability (30 min). Search your category terms in-platform. Are you there?
- Competitor scan (45 min). Where are they actually investing, and what's the category consensus you could break?
- Write five findings with consequences (60 min). Each one: what we found, why it matters, what we'll do, what would prove us wrong.
- Make the stop decision (15 min). Which platform, and where its capacity goes.
Roughly four to five hours. The reason to time-box it is that social audits expand indefinitely if allowed — there is always another metric — and a document nobody finishes changes nothing, which is the same failure mode as any end-to-end audit that becomes a project rather than a decision.
Once the decisions are made, the follow-through is a scheduling problem: what gets published, where, and how often, which is where a working content calendar and a proper distribution system take over from the audit.
The short version
Published engagement benchmarks disagree by several times over because they divide by different things, so check the denominator before comparing anything — and default to your own trend instead. Audit the whole presence, not just the accounts you run: forgotten profiles, employee posts, what customers say, and whether you're findable in in-platform search. Cut what isn't working rather than reducing it, since an abandoned-looking account is worse than a closed one. And end with five findings that each carry a consequence, plus one explicit decision about what you're stopping.
Posting everywhere and not sure what's actually working?
We audit social presence, cut what isn't earning its place, and concentrate effort where it performs.
Explore Social Media Marketing →Frequently asked questions
What is a good social media engagement rate in 2026?
The question is harder to answer than it looks, because published benchmarks disagree by several times over depending on the denominator used. Instagram's 2026 median has been reported as roughly 0.30%, 0.48% and 1.62% by different sources — all technically correct, because one divides by followers, one by reach and one by impressions. Before comparing yourself to any benchmark, confirm which formula it used and whether your analytics tool uses the same one. Broadly, TikTok leads major platforms, LinkedIn has been rising, and Facebook organic sits at historic lows.
How often should you run a social media audit?
A full audit once or twice a year is enough for most organisations, with a lighter monthly check of your own trend in between. The full version is worth triggering by a decision rather than a calendar — before a budget reallocation, a repositioning, a platform launch, or a hiring decision. Auditing more frequently than that tends to produce documents rather than changes, because the underlying numbers do not move enough in a quarter to justify a new conclusion.
What should a social media audit actually include?
More than your own accounts. A complete audit covers the profiles you actively run, any abandoned or forgotten accounts still carrying your name, what employees and customers post about you, what appears when someone searches your brand on each platform, and whether your content is findable through in-platform search rather than only through the feed. Most published checklists cover only the first of those, which is the part you already have most visibility into and often the least influential part of your actual public presence.
Should a social media audit end with abandoning a platform?
Frequently, yes, and this is the most valuable outcome an audit can produce. Effort spread thinly across six platforms usually performs worse than the same effort concentrated in two, and some platforms have declined enough that maintaining a presence there is a genuine cost with little return — brands have measurably reduced posting frequency on the weakest performers. An audit that concludes everything should continue as it is has usually not been honest about where the results are actually coming from.
How do you measure social media when attribution is unreliable?
Accept that a large share of social's effect is unmeasurable at the individual level and measure at a level where that does not matter. Track your own trend over time rather than comparing to industry medians, watch branded search and direct traffic as a proxy for awareness that produced no click, and use a self-reported source question on forms to catch what tracking misses. Judging social by last-click attribution consistently undervalues it, because its main function is influence before intent rather than capture at the point of purchase.
