Social Media KPIs: Every Formula and Target Band
Social media KPIs are the handful of measurable values tied to a business goal that you commit to reporting and improving. Every KPI is a metric, but a metric only becomes a KPI when a decision depends on it.
That distinction is most of the work. What follows is the part almost no KPI guide publishes: for every KPI, the formula, a target band with a date and a named source, and what actually moves it — plus an honest note on the four KPIs where no published band exists, and the metrics to remove from your report entirely.
Table of Contents
- What Are Social Media KPIs?
- KPI vs Metric: What Actually Separates Them
- Social Media KPIs With Formula, Target Band and Source
- Why Four of These KPIs Have No Published Target Band
- Stop Reporting These: The Named Drop List
- How Do You Choose Which Social Media KPIs to Track?
- How Often Should You Review Each KPI?
- How to Report Social Media KPIs Without a Paid Dashboard
- Frequently Asked Questions
What Are Social Media KPIs?
Social media KPIs are the three to five numbers per channel that tell you whether social media is doing the job you hired it for. They group cleanly by business objective, and that grouping is how you avoid reporting 20 numbers that contradict each other:
- Awareness — are more of the right people seeing you? (audience growth rate, reach rate, share of voice)
- Engagement — do the people who see you care? (engagement rate, amplification rate)
- Conversion — does that attention turn into action? (link CTR, conversion rate)
- Revenue — is the attention worth money? (RPM, earned media value)
- Customer care — are you answering people? (first response time)
A good report carries one KPI from each group that matters to you, with last period’s number beside it. Anything that cannot be tied back to one of those five objectives is a metric you are collecting out of habit. For what to actually do when a KPI is flat, our companion hub on social media growth covers the levers; this page is about measuring them correctly.
KPI vs Metric: What Actually Separates Them
A metric is any number a platform will hand you. A KPI is a metric you have committed to a target on, and will act on when it misses. The difference is not statistical, it is organisational — the same number can be a metric for one team and a KPI for another.
Impressions are the clearest example. For a brand running an awareness campaign with a reach target, impressions are a KPI: miss the number and the media plan changes. For a support team measured on response time, impressions are context at best. Nothing about the number changed; the decision attached to it did.
Two tests settle almost every argument about whether something belongs in a report:
- The decision test. If the number moves 20% in either direction, does anyone do something differently? If not, it is a metric.
- The ownership test. Can one person name the input they would change to move it? A number nobody can influence is a weather report, not a KPI.
Social Media KPIs With Formula, Target Band and Source
Here is every KPI worth reporting, with its formula, a dated target band where one is published, and the source of that band. Seven rows say “no published band” rather than inventing one — that gap is explained in the next section, and the method for setting your own baseline is there too.
| KPI | Formula | Target band (source) | What moves it |
|---|---|---|---|
| Audience growth rate | (followers end − followers start) ÷ followers start × 100 | No published band — use your own trailing 90-day average | Cadence, reach, one breakout post |
| Reach rate | accounts reached ÷ followers × 100 | No published band — use your own 90-day median | Format, posting time, shares |
| Share of voice | your mentions ÷ (your + competitors’ mentions) × 100 | No published band | Launches, PR, campaign volume |
| Engagement rate, per follower (Instagram) | (likes + comments) ÷ followers × 100 | 0.37–0.55%, avg 0.48% (Socialinsider, 35M posts / 447,613 profiles); brand accounts 0.30% median (Quid 2026) | Format — carousels 0.55%, Reels 0.52%, images 0.37% |
| Engagement rate, per view (TikTok) | (likes + comments + shares + saves) ÷ views × 100 | 3.75–4.40%, avg 3.85% (Socialinsider, 2M videos / 214,507 profiles) | Hook retention, saves; smaller accounts score higher |
| Engagement rate, per impression (LinkedIn) | engagement ÷ impressions × 100 | 3.25–7.00%, avg 5.20% (Socialinsider, 1.3M posts / 16,645 pages) | Format — documents 7.00%, video 6.00%, links 3.25% |
| Engagement rate, per follower (Facebook) | (reactions + comments + shares) ÷ fans × 100 | 0.05–0.20%, avg 0.15% (Socialinsider, 25M posts / 130,683 pages) | Format — status 0.20%, Reels 0.18%, links 0.05% |
| Engagement rate, per subscriber (YouTube) | interactions ÷ subscribers × 100 | 0.21% median — a single figure, not a range (Quid 2026, 150 companies × 18 industries) | Comments, community posts, publishing rhythm |
| Amplification rate | shares ÷ followers × 100 | No published band | Usefulness of the post, not its polish |
| Link CTR | link clicks ÷ impressions × 100 | No published band — use your own 90-day median | First-line promise, link placement |
| Conversion rate | conversions ÷ link clicks × 100 | No published band — depends entirely on the offer | Landing-page match, UTM hygiene |
| RPM (revenue per 1,000 views) | earnings ÷ views × 1,000 | YouTube long-form $1–$30 est.; Shorts $0.05–$0.15 est.; TikTok $0.40–$1.00 per 1,000 qualified views est. No platform publishes a rate | Niche, audience country, format |
| First response time | median time from inbound message to first reply | No published band — set it as a service promise, not a benchmark | Staffing, notification routing |
Bands verified 24 September 2026. The Socialinsider benchmark studies carry their own disclosure that these are 2025 values presented in the 2026 edition; the Instagram and LinkedIn windows are January 2024–December 2025. The Quid 2026 benchmark report (formerly Rival IQ) publishes no month boundaries.
The engagement-rate denominator is different on every platform
LinkedIn’s 5.20% is not ten times better than Instagram’s 0.48% — they are different fractions. Instagram and Facebook engagement is measured per follower, TikTok per view, LinkedIn per impression. A report that puts those percentages in one column without the denominator is not comparing anything, and a benchmark applied with the wrong denominator will invent a crisis or hide one.
The rule is to compute your rate the same way the benchmark was computed, then compare only within that platform. Our per-impression LinkedIn source and per-follower Instagram source are separate studies for exactly this reason. The full arithmetic, platform by platform, is in our guide to calculating engagement rate, and the free engagement rate calculator applies the correct denominator per platform so you cannot make this mistake by accident.
Revenue KPIs: no platform publishes a rate
RPM is the only revenue KPI most creators need, and every RPM figure in circulation is an outside estimate. YouTube publishes the revenue split — 55% of long-form watch-page ad revenue and 45% of Shorts revenue — and never a resulting RPM. TikTok’s Creator Rewards terms set rewards at its “sole discretion” and reserve the right to cap them.
Two measurement points matter more than the estimate itself. RPM is not CPM: RPM is what you earn per 1,000 views after the platform’s share, across all views and all revenue types; CPM is what an advertiser pays per 1,000 impressions before the split. RPM is always the lower number, and confusing them is the most common reporting error in creator economics. And payouts run on qualified views, not public views — TikTok counts a view only at five seconds or more, once per account, on videos of at least a minute, so any estimate built on a raw view count comes in high. If you sell posts rather than earn on views, brands sanity-check a quoted fee against roughly $5–$25 per 1,000 people reached.
Why Four of These KPIs Have No Published Target Band
Reach rate, share of voice, amplification rate and conversion rate have no credible published benchmark, and printing one would be dishonest. Reach rate depends on how inflated your follower count is relative to your active audience. Share of voice depends entirely on which competitors you chose to count. Amplification rate is reported inconsistently across platforms. Conversion rate is a property of your offer, not of social media.
So set your own band instead, in four steps:
- Pull 90 days of that number for one platform, post by post.
- Take the median, not the mean — one viral post will drag an average somewhere you cannot repeat.
- Set the floor at the median and the target 20% above it. That is your band until you have a better one.
- Re-baseline quarterly. A band from last year’s follower count is not a band.
This is also why three rows you might expect to carry a band — audience growth rate, link CTR and first response time — do not. Numbers circulate for all three, but every version we chased led to a roundup with no named publisher, no sample size and no date, so we left the column empty rather than pass the figure on. The five engagement rows come from studies that name their sample (35 million posts, 2 million videos, 25 million posts) and are worth real weight. A benchmark without a sample size and a date is an opinion, wherever you read it — including here.
Stop Reporting These: The Named Drop List
These five metrics belong out of your report, and each has a replacement. This is the section most reports need and almost no KPI guide will commit to.
| Stop reporting | Report instead | Why |
|---|---|---|
| Follower count | Audience growth rate (%) | A raw count rewards the past; the rate shows this month’s work |
| Total impressions | Reach rate (unique accounts ÷ followers) | Impressions double-count the same people; reach does not |
| Total likes | Engagement rate with its denominator named | Likes scale with audience size, so they flatter big accounts and punish small ones |
| Total posts published | Reach per post | Volume is an input, not a result — and more posts often means less reach per post |
| Follower demographics (every month) | Nothing — check it quarterly | It barely moves month to month and fills slides with certainty you already had |
One nuance worth keeping: none of these numbers is useless, and several are inputs to the KPIs that replace them. The problem is reporting them as results. A deck that leads with total likes is describing the size of the audience, not the performance of the work.
How Do You Choose Which Social Media KPIs to Track?
Pick three to five per channel: one awareness KPI, one or two engagement KPIs, one conversion KPI, and a revenue KPI if you monetize. Three to five is the one number Hootsuite, Brandwatch and Sprout Social independently agree on, and the reason is practical — a report with 15 numbers gets skimmed and acted on by nobody.
Work backwards from the business goal rather than forwards from the dashboard:
| Business goal | Primary KPI | Supporting KPI |
|---|---|---|
| Get known in a new market | Reach rate | Share of voice |
| Build a warm audience | Engagement rate (correct denominator) | Amplification rate |
| Drive signups or sales | Conversion rate | Link CTR |
| Earn from content | RPM | Engagement rate |
| Keep customers happy | First response time | Amplification rate |
Then write the target, the baseline and the review date next to each one. A KPI without a baseline recorded on day one is unfalsifiable by the end of the quarter — and if you want a structured starting point, our free social media audit template scores where each channel stands before you set targets.
How Often Should You Review Each KPI?
Match the review interval to how fast the number can actually move: weekly for engagement and response time, monthly for growth, reach and CTR, quarterly for conversion, share of voice and RPM. Nobody publishes this table, which is why so many teams review everything monthly and act on noise.
| Review | KPIs | Why this interval |
|---|---|---|
| Weekly | Engagement rate, first response time | Both respond within days and both are fixable the same week |
| Monthly | Audience growth rate, reach rate, link CTR | Enough posts to have a median; short enough to change the next month |
| Quarterly | Conversion rate, share of voice, RPM, amplification rate | Slow-moving, and seasonal noise swamps a monthly read |
The failure modes are symmetrical. Reviewing audience growth weekly produces a number that swings on one post and tempts you into changing strategy for no reason. Reviewing response time quarterly means you learn in April that you were ignoring people in February.
How to Report Social Media KPIs Without a Paid Dashboard
Native analytics plus a UTM convention plus one template covers every KPI in the table above — the paid dashboards buy you time and history, not numbers you cannot otherwise get. That is worth saying plainly, because most KPI guides are published by reporting vendors.
The stack, in order:
- Native analytics, per platform — reach, impressions, engagement, follower change and video retention are all free. Export monthly; the platforms’ history windows are shorter than you think.
- UTM parameters on every link — without them, clicks and conversions cannot be attributed to a platform at all. Build them consistently with our free UTM generator and keep the naming convention fixed, because inconsistent tags are the most common reason a conversion KPI reads as zero.
- One report, same shape every month — KPI, this period, last period, target, one sentence on why. Our social media report template is ungated and pre-built with the formulas in this article; no email, no trial.
- An AI assistant for the write-up — pulling the numbers is mechanical, but explaining them is not. An AI social media manager can draft the narrative around the numbers you paste in.
A note on rigour that costs nothing: record the denominator next to every engagement figure in your report. “Engagement rate 3.9% (per view, TikTok)” is a number a stranger can check. “Engagement rate 3.9%” is not.
Reporting Nine Platforms From One Conversation
The reporting overhead of multi-platform social media is mostly collection, not analysis — nine native dashboards, nine export formats, nine different denominators, once a month.
Outfeed AI is chat-first: you ask for the numbers and the assistant pulls performance across all nine connected platforms in one conversation, then drafts the posts for the next period in the same place. It is not an enterprise analytics suite, and if you need social listening or multi-year historical dashboards, Brandwatch and Sprout Social do that properly. What it removes is the monthly copy-paste.
Starter is $29/month flat and Pro is $59/month flat with 3 team members included; per-seat billing applies only at the $149 Agency tier. Start a free 3-day trial with 100 AI credits, card required.
Frequently Asked Questions
What is the difference between a KPI and a metric?
Every KPI is a metric, but a metric only becomes a KPI when a business decision depends on it. Impressions are a metric; impressions become a KPI when you have committed to a reach target and will change your content if the number misses. The test is simple — if nothing changes when the number moves, it is not a KPI.
How many social media KPIs should I track?
Three to five per channel or campaign, which is the one rule Hootsuite, Brandwatch and Sprout Social all agree on. One awareness KPI, one or two engagement KPIs, one conversion KPI and, if you monetize, one revenue KPI is enough for a report anyone will read. Tracking 20 numbers means acting on none of them.
How often should I review social media KPIs?
Weekly for engagement rate and response time, monthly for audience growth, reach and CTR, quarterly for conversion rate, share of voice and RPM. The rule is that review frequency should match how fast the number can actually move — reviewing follower growth weekly produces noise, and reviewing response time quarterly means you find out too late.
Can I track social media KPIs without a paid tool?
Yes. Native analytics on each platform give you reach, impressions, engagement and follower counts for free, a UTM-tagged link plus your own analytics covers clicks and conversions, and a spreadsheet or free template turns those into a report. Paid dashboards buy you time and history, not numbers you cannot otherwise get.
What if my social media KPIs are not improving?
First check the KPI is measured the same way as the benchmark you are judging it against — Instagram engagement is per follower, TikTok is per view and LinkedIn is per impression, so the wrong denominator can invent a problem that does not exist. If the measurement is sound, change one input at a time and give it 30 days before judging.
Conclusion
Social media KPIs are only useful with three things attached: a formula, a target band that carries a date and a source, and a decision that depends on the number. The table above supplies the first two where a credible band exists and says so where it does not — which, on four of thirteen KPIs, is the honest answer.
Keep three to five per channel, record the denominator beside every engagement figure, review each number at the interval it can actually move on, and drop the five metrics on the list above from your deck entirely.
Next steps: compute your rates correctly with the free engagement rate calculator, build this month’s report in the ungated social media report template, and when a KPI is genuinely flat, work the levers in the growth hub linked above.
Sources (verified 24 September 2026): Socialinsider’s Instagram, TikTok, LinkedIn and Facebook benchmark studies for the five engagement bands and their sample sizes — their own disclosure notes these are 2025 values presented in the 2026 edition; the Quid (formerly Rival IQ) 2026 Social Media Industry Benchmark Report for the YouTube median, the 0.30% brand-account Instagram median and the year-on-year decline; YouTube Help for the 55%/45% revenue split and the RPM-versus-CPM definitions; TikTok’s Creator Rewards Program terms, US (updated 20 July 2026) for the qualified-view rules and the absence of a published rate; Influencer Marketing Hub for the independent RPM and brand-deal CPM estimates. Audience growth rate, reach rate, share of voice, amplification rate, link CTR, conversion rate and first response time are deliberately shipped with no target band: we could not find a benchmark for any of them with a named publisher, a disclosed sample size and a date, and an unattributable figure is worse than an honest gap on a page people plan against. No platform publishes a payout rate per view; every RPM figure here is an independent estimate and varies several-fold by niche and audience country.