Social Media Metrics and KPIs: The Complete Guide to Measuring What Matters

I've spent the last decade staring at social media dashboards — first as an in-house marketer, then as a consultant, and now building PostEverywhere. And if there's one thing I've learned, it's that most marketers are drowning in metrics but starving for insight.

You open Instagram Insights and there are 40 numbers staring back at you. TikTok Studio gives you another 30. LinkedIn throws in its own flavour. Multiply that by six platforms and suddenly you're in a spreadsheet hell where everything feels important and nothing actually is.

This guide is my attempt to cut through all of that. It's the reference I wish someone had handed me in 2016 — every metric that matters, every formula written out properly, and honest opinions on which numbers to chase and which to quietly ignore. Bookmark it. Share it with your team. Come back to it when your boss asks "so what does engagement rate actually mean?" and you need a proper answer.

TL;DR

• There are six categories of social media metrics: reach, engagement, audience, conversion, video, and brand health. You don't need to track them all — you need the right ones for your goal.
• Engagement rate by reach is the most meaningful engagement metric for organic content. Anything by followers is increasingly misleading in 2026 because follower counts barely influence distribution anymore.
• The 2026 median engagement rates from Socialinsider's 70M post study: TikTok 3.70%, LinkedIn 5.20%, Instagram 0.48%, Facebook 0.15%, X 0.12%.
• Vanity metrics (followers, likes, impressions in isolation) aren't useless — they're just not KPIs. Treat them as diagnostic, not goal-level.
• Every metric should ladder up to a business goal: awareness, consideration, conversion, or retention. If it doesn't, stop tracking it.
• The Avinash Kaushik framework — amplification, applause, conversation, economic value — is still the cleanest way to think about social ROI twenty years after he wrote it.
• You need one primary KPI per campaign, not twelve. Pick the one, optimise for it, report everything else as context.

Table of Contents

• The six categories of social media metrics
• Reach metrics
• Engagement metrics
• Audience metrics
• Conversion and ROI metrics
• Video-specific metrics
• Brand health metrics
• Vanity metrics vs actionable metrics
• Which metrics map to which goal
• 2026 platform benchmarks
• Platform-specific deep-dives
• FAQs

The six categories of social media metrics

Before we get into formulas, let's zoom out. Every social media metric — and I mean every single one — falls into one of six buckets:

1. Reach metrics — how many eyeballs saw your content
2. Engagement metrics — what they did once they saw it
3. Audience metrics — who they are and how your audience is growing
4. Conversion metrics — whether they took a business action (clicks, leads, sales)
5. Video metrics — how people consumed your video content specifically
6. Brand health metrics — what people say about you when you're not in the room

The mistake I see constantly is marketers treating these as interchangeable. They're not. A campaign optimised for reach will usually tank your engagement rate. A campaign optimised for conversions will usually underperform on reach. You cannot maximise all six simultaneously — pick your battle and track the rest as context.

Let's go through each category properly.

Reach metrics: who actually saw this?

Reach metrics answer the first, most basic question: did anyone see this post? If the answer is no, nothing else matters. You can have the world's best creative and the most compelling CTA, but if the algorithm doesn't distribute it, you're shouting into a void.

Impressions

Definition: The total number of times your content was displayed on a screen, including multiple views by the same person.

Formula: It's a raw count — no calculation needed. Your platform analytics report it directly.

When to use it: Impressions are your top-of-funnel awareness metric. They're the denominator in most other calculations (CTR, engagement rate by impressions, CPM). On their own they don't tell you much — someone scrolling past your post in 0.4 seconds counts as an impression — but they're the foundation for everything else.

Reach

Definition: The number of unique people who saw your content at least once. Reach ≤ Impressions, always.

When to use it: Reach is the honest version of impressions. If you have 100,000 impressions but only 20,000 reach, the same 20K people saw your post five times each — which is fine for an ad campaign but probably means something is broken with your organic distribution.

The ratio of impressions to reach is called frequency (Impressions ÷ Reach). For paid campaigns, frequency of 2-4 is usually healthy. Above 7 and you're fatiguing your audience.

Video views

Definition: Depends entirely on the platform — and this is where marketers get burned. A "view" on TikTok counts after 0 seconds (an impression, essentially). A view on YouTube requires 30 seconds. On Instagram Reels it's 3 seconds. On LinkedIn it's 2 seconds with 50% of the video in view.

Never, ever compare video views across platforms as if they're the same unit. They're not. It's like comparing miles to kilometres without converting.

Want to stop manually pulling reach and impression data from six dashboards? PostEverywhere's analytics pulls it all into one view — reach, impressions, frequency, trends — across every platform you post to. Start your 14-day free trial, no credit card required.

Engagement metrics: did it actually resonate?

This is where things get interesting — and where most people calculate things wrong. Engagement metrics are the signal that your content didn't just appear on screens, it made people stop scrolling and do something.

The three engagement rate formulas (and which to use)

Here's the thing nobody tells you: there isn't one engagement rate formula. There are three, and they measure different things.

Engagement rate by reach:

Engagement Rate (Reach) = (Total Engagements ÷ Reach) × 100

This is the one I recommend for organic content in 2026. It tells you: of the people who actually saw this post, what percentage engaged? It's the purest measure of creative quality because it isolates content performance from distribution luck.

Engagement rate by impressions:

Engagement Rate (Impressions) = (Total Engagements ÷ Impressions) × 100

This is what most paid media platforms default to. It's slightly less flattering than by reach (because impressions ≥ reach) but it's the standard for ad reporting, so know it.

Engagement rate by followers:

Engagement Rate (Followers) = (Total Engagements ÷ Followers) × 100

This is the old-school formula from 2015 when organic reach was roughly proportional to follower count. In 2026 it's basically meaningless on algorithm-driven platforms (Instagram, TikTok, X) because follower count has almost no relationship to distribution. I still see agency reports using it because clients recognise follower counts. Don't be that agency.

My honest take: use engagement rate by reach for organic reporting, by impressions for paid reporting, and abandon by followers except on LinkedIn where the follower graph still actually matters.

If you want to skip the math, our free engagement rate calculator does all three formulas side by side so you can see which one your agency is quietly using.

Likes (approvals)

The softest engagement signal. A like is a two-calorie commitment — it takes half a second and requires zero thought. Useful as a directional signal, useless as a KPI. If you're measuring success by likes in 2026, you've already lost.

Comments

Comments are 10-100x more valuable than likes because they require actual effort and they signal to the algorithm that your content sparked conversation. Every algorithm in 2026 — Instagram, TikTok, LinkedIn, Facebook — weights comments heavily in their ranking systems. If I had to pick one engagement metric to optimise for, it'd be comments per post.

Shares (amplification)

Shares are the gold standard engagement. When someone shares your content, they're putting their personal reputation behind it. It's the strongest possible endorsement and it's what makes content actually go viral. Shares should be tracked separately from likes and comments because they're a completely different quality of signal.

Saves

Saves are the quiet superhero metric on Instagram, Pinterest, and TikTok. A save means "I want to come back to this later" — it signals high intrinsic value and the algorithms notice. If you're making educational or reference content, saves should be one of your top KPIs. They massively outperform likes as a distribution signal.

The Avinash Kaushik framework

Back in 2011, Avinash Kaushik (ex-Google) wrote a piece called Best Social Media Metrics: Conversation, Amplification, Applause, Economic Value. It's fifteen years old and it's still the cleanest framework I've ever seen for thinking about social metrics. Here are the three rate formulas it gave us:

Conversation Rate:

Conversation Rate = (Comments ÷ Followers) × 100

This measures how well you're prompting dialogue. Avinash's insight: replies aren't accidents. Content that generates conversations is content that matters.

Amplification Rate:

Amplification Rate = (Shares ÷ Followers) × 100

This measures how much your audience is willing to broadcast your content to their own networks. High amplification = high organic distribution = lower paid media needs.

Applause Rate:

Applause Rate = (Approvals ÷ Followers) × 100

This is the likes-per-follower rate. Avinash included it for completeness but explicitly warned it's the weakest of the three. He was right in 2011 and he's more right in 2026.

I still run these three calculations for every client report I touch. They tell a story that aggregate "engagement rate" can't — specifically whether your audience is talking, sharing, or just mindlessly tapping hearts.

Audience metrics: who are these people?

Audience metrics are about the humans behind the numbers. They're lagging indicators (slow to change) but they're strategically essential.

Follower growth rate

Follower Growth Rate = (Net New Followers ÷ Starting Followers) × 100

The honest version of "did we gain followers." Raw new follower count is misleading because gaining 1,000 followers means different things for a 5K account vs a 500K account. The rate normalises it.

Benchmark: 1-2% monthly growth is healthy for most brands. 5%+ monthly growth means something is working unusually well — dig in and figure out why.

Audience demographics

Age, gender, location, language, interests. These aren't KPIs, they're diagnostic. You check them to answer questions like: "Our engagement dropped — did our audience shift?" or "Should we post at a different time because 60% of our followers are now in APAC?" Speaking of which, our best time to post guide breaks down optimal posting windows by audience timezone.

Audience quality

Nobody officially measures this but you should. It's the ratio of "people who actually care about your brand" to "people who followed during a giveaway and will never engage again." Proxies:

• Engaged follower rate (followers who engaged in the last 30 days ÷ total followers)
• Follower-to-customer conversion rate
• Email signup rate from social

A 10K audience of genuinely interested people destroys a 100K audience of giveaway hunters. Every time.

Conversion and ROI metrics: does this make money?

This is the category your CFO cares about. Everything above is interesting — this is what keeps your budget.

Click-through rate (CTR)

CTR = (Clicks ÷ Impressions) × 100

The percentage of people who saw your content and clicked the link. For organic social CTR benchmarks range wildly — 0.5% is decent on Instagram, 2%+ is good on LinkedIn, paid social ads average 0.9% across platforms.

Always tag your links with UTM parameters or you won't be able to attribute clicks properly. Our free UTM builder generates properly structured UTM URLs in about four seconds.

Conversion rate

Conversion Rate = (Conversions ÷ Clicks) × 100

Of the people who clicked, what percentage did the thing you wanted (bought, signed up, downloaded). Conversion rate lives on your landing page, not on social — but if your social CTR is good and your conversion rate is bad, your targeting is better than your offer.

Cost per click (CPC)

CPC = Spend ÷ Clicks

What you paid for each click. Paid social CPCs in 2026: LinkedIn $5.58, Instagram $1.28, Facebook $0.97, TikTok $1.00, X $0.38. LinkedIn is always expensive because B2B intent is always expensive.

Cost per mille / thousand (CPM)

CPM = (Spend ÷ Impressions) × 1,000

What you paid to reach 1,000 people. CPM is the foundation of brand awareness ad buying — when you're not optimising for clicks, you're optimising for cheap impressions to the right audience.

Cost per acquisition (CPA)

CPA = Spend ÷ Conversions

The real one. CPA is what every ad campaign eventually gets judged on. If your CPA is lower than your customer lifetime value (LTV), you have a profitable machine. If it's higher, you're losing money on every new customer and making it up in volume — which is a famous way to go bankrupt.

Return on ad spend (ROAS)

ROAS = Revenue ÷ Ad Spend

Expressed as a ratio or multiplier — a ROAS of 4 means you made $4 for every $1 spent. For e-commerce, 3-4x is usually the profitability threshold; below that you're likely losing money after COGS, shipping, returns, and overhead.

Social referral traffic

This lives in Google Analytics, not your social dashboards. Go to Acquisition → Traffic Sources → Social and see how much actual website traffic each platform is driving. You'll often find the platform driving the most engagement isn't the one driving the most traffic. That's useful.

Tired of manually calculating CPC, CPA, and ROAS across five ad platforms? PostEverywhere's analytics tracks paid and organic performance in one place with automatic ROI calculations. See pricing.

Video-specific metrics: the new must-haves

Video is now 80%+ of all social content. You cannot evaluate a TikTok, Reel, Short, or YouTube video with the same metrics you'd use for a static image. Here are the ones that actually matter.

Watch time

Total minutes watched across all viewers. This is YouTube's single most important ranking signal and increasingly matters on TikTok and Instagram Reels too. A 60-second video watched fully by 1,000 people generates 1,000 watch minutes. That beats a 10-second video watched by 10,000 people for pure watch time (1,667 minutes vs, well, 1,667 minutes — the math is the same but longer content compounds harder in recommendation systems).

Average view duration / retention curve

Average view duration is the headline number. The retention curve is the story. Open any video in TikTok Analytics or YouTube Studio and you'll see a graph showing what % of viewers were still watching at each second. Dips in that curve tell you exactly where you're losing people — usually it's the intro, a boring transition, or the moment someone realises you're about to pitch them.

Video completion rate

Video Completion Rate = (Viewers to End ÷ Total Viewers) × 100

The percentage of people who watched all the way through. For short-form (under 30 seconds) aim for 50%+. For long-form (over 3 minutes) aim for 30%+. Anything above these means the algorithm will reward you with more distribution.

Hook rate

Not a standard metric but I use it constantly: the percentage of viewers still watching after the first 3 seconds. If your hook rate is below 70% on short-form content, the problem isn't your whole video — it's your first frame.

Brand health metrics: the stuff that's hard to measure

Brand health metrics are the ones executives love and analysts hate, because they're fuzzy, they're delayed, and they're genuinely hard to calculate. But they matter — sometimes more than everything else combined.

Brand mentions

How many times your brand was mentioned on social in a given period (with or without a tag). You'll need a social listening tool (Brandwatch, Sprout Social, Brand24, Meltwater) because platform-native analytics won't show you untagged mentions.

Share of Voice (SOV)

Share of Voice = (Brand Mentions ÷ Total Industry Mentions) × 100

Your percentage of the total conversation in your category. If you and your four top competitors were mentioned 1,000 times combined and you got 250, your SOV is 25%. Growing SOV correlates strongly with market share over 12-24 month timelines. It's one of the few brand metrics with actual business rigour behind it.

Sentiment

The ratio of positive to negative to neutral mentions. Good sentiment tools do this automatically with NLP. Manual tracking is a nightmare for anything over 500 mentions/month. Watch for negative spikes more than overall ratios — sentiment crises are more informative than sentiment averages.

Net Promoter Score (NPS)

NPS = % Promoters − % Detractors

Technically a survey metric not a social one, but I include it because NPS often correlates with social brand health. Promoters are people who score you 9-10 out of 10; detractors are 0-6; passives (7-8) get ignored. Scores range from -100 to +100. Above 50 is excellent, above 70 is world-class.

Vanity metrics vs actionable metrics: the honest take

Here's where I'm going to upset some people.

"Vanity metrics" has become a lazy insult. Marketers throw the term at follower counts and like counts and pretend only CPA and ROAS matter. That's wrong. Vanity metrics aren't useless — they're just not goal metrics.

The real distinction is:

• Goal metrics (KPIs): The 1-3 numbers you're actually trying to move. These directly ladder to a business outcome.
• Diagnostic metrics: Everything else. You watch them to understand why your goal metric is moving up or down.

Followers are diagnostic. Likes are diagnostic. Impressions are diagnostic. Reach is diagnostic. All of them tell you something useful when you're debugging a drop — but none of them should be the headline number in your monthly report.

If your monthly exec report leads with "we gained 5,000 followers and got 40,000 likes" and doesn't mention pipeline contribution, you are making a career-limiting decision. Lead with conversions or revenue influenced by social. Lead with CPA trending down. Lead with SOV growth vs competitors. Then, as supporting evidence, bring in the vanity metrics.

Which metrics map to which goal

Here's the cheat sheet I wish someone had printed for me in 2016.

One KPI per goal. One goal per campaign. If someone asks you to optimise for "all of them," your job is to push back and force a choice — because optimising for everything means optimising for nothing.

2026 platform benchmarks: what "good" actually looks like

Benchmarks are always the first thing people ask for and the last thing that actually matters — because averages hide everything interesting. Still, you need a baseline. Here are the median engagement rates across platforms for 2026, pulled from Socialinsider's 70-million-post benchmark study released earlier this year.

A few things worth noting:

• LinkedIn leads in 2026. The B2B platform finally overtook TikTok on median engagement rate this year — largely because LinkedIn's algorithm rewards thoughtful long-form posts and the user base has matured massively.
• Instagram collapsed. The median is now under 0.5%. In 2019 it was over 2%. Organic reach has been throttled across the board and engagement rate by followers is now a borderline meaningless metric on the platform.
• Facebook and X are functionally dead for organic engagement. You can still get reach on both — but expect paid to do most of the heavy lifting.
• TikTok remains strong but is softening. It was 4.5% in 2024 and it's 3.70% now. Still the highest organic engagement of any mass platform.

For more context, Rival IQ's annual industry benchmark report slices these numbers by industry (B2B, SaaS, e-commerce, non-profit, etc) and is the best free resource I know for sector-specific benchmarks. And our engagement rate benchmarks post goes deeper on the by-industry data.

You can also run your own account against platform benchmarks using our social media benchmarks tool or run a full audit with the social media audit tool.

Platform-specific deep-dives

This guide is the hub. Each of the posts below takes a single platform and goes properly deep on which metrics the native analytics actually show you, how the algorithm weights them, and what to optimise for.

• Instagram metrics and KPIs — Reach vs impressions, why saves matter more than likes, and the metrics the Instagram algorithm actually rewards.
• TikTok metrics and KPIs — For You Page signals, video completion rate, and why follower count is basically meaningless here.
• YouTube metrics and KPIs — Watch time is king. Everything else is supporting evidence. Deep dive on CTR, AVD, and session watch time.
• LinkedIn metrics and KPIs — Why LinkedIn is the one platform where follower graph still matters, and the dwell-time metric you can't see but needs to exist.
• Facebook metrics and KPIs — How to measure anything meaningfully when organic reach has collapsed. Includes the Facebook algorithm signals that matter.
• X / Twitter metrics and KPIs — Impressions are the default, but engagement rate by impressions is what actually matters here.
• Threads metrics and KPIs — The newest platform, the most basic analytics, and the metrics you have to calculate manually.
• Pinterest metrics and KPIs — Pinterest is a search engine pretending to be social. Saves and outbound clicks are the whole game.

Track these metrics in your scheduler

Measuring metrics in eight different native dashboards is how strategies die quietly. PostEverywhere's social media analytics tracks every metric in this guide across all 8 platforms in one place, so you can compare engagement rate on Instagram to CTR on LinkedIn without twelve open tabs.

If you're focused on a single platform, jump to the relevant scheduler: Instagram scheduler, TikTok scheduler, YouTube scheduler, LinkedIn scheduler, Facebook scheduler, X scheduler, Threads scheduler, or Pinterest scheduler — each has built-in analytics for that platform's specific KPIs, plus best-time-to-post data derived from your own historical engagement.

FAQs

How do I calculate engagement rate?

There are three formulas depending on what you're measuring. Engagement rate by reach (most accurate): (Total Engagements ÷ Reach) × 100. Engagement rate by impressions (standard for paid): (Total Engagements ÷ Impressions) × 100. Engagement rate by followers (legacy): (Total Engagements ÷ Followers) × 100. I recommend engagement rate by reach for organic reporting in 2026. Use our calculator if you want to run all three side by side.

What's a good engagement rate in 2026?

Depends on the platform. Per the Socialinsider 2026 benchmark study: TikTok 3.70% median (5%+ is good), LinkedIn 5.20% (7%+ is good), Instagram 0.48% (1.5%+ is good), Facebook 0.15% (0.5%+ is good), X 0.12% (0.3%+ is good). Context matters more than the number — beating your own historical average is more meaningful than beating industry median.

Which metrics matter most for ROI?

For paid campaigns: CPA and ROAS, full stop. For organic campaigns with commercial intent: conversion rate, CTR, and social referral traffic (measured in Google Analytics, not platform-native). Don't try to assign ROI to pure awareness campaigns — use reach, SOV, and brand lift studies instead.

What's the difference between reach and impressions?

Reach is the number of unique people who saw your content. Impressions is the total number of times your content was displayed, including repeat views by the same person. Impressions will always be equal to or greater than reach. The ratio (Impressions ÷ Reach) is called frequency — for ad campaigns, frequency between 2-4 is usually healthy; above 7 you're fatiguing your audience.

Are followers still a useful metric in 2026?

As a KPI? No. As a diagnostic metric? Yes. Follower count has almost no influence on organic distribution on algorithm-driven platforms (Instagram, TikTok, X) because those platforms now show content to non-followers based on interest signals. The exceptions are LinkedIn, YouTube (for subscribers), and to a lesser extent Facebook — where the follower graph still genuinely affects distribution. Track follower growth rate as a lagging indicator of content quality, not as a primary goal.

What are vanity metrics and should I ignore them?

Vanity metrics are metrics that feel impressive but don't directly tie to business outcomes — things like follower count, raw like count, and impressions in isolation. You shouldn't ignore them but you shouldn't lead reports with them either. Think of them as diagnostic: useful for understanding why a KPI moved, but not the KPI itself. Your monthly exec report should lead with conversion metrics or brand health metrics, not follower growth.

How often should I check my social media metrics?

Three different rhythms: Daily — monitor for anomalies (a post going unexpectedly viral, a sudden drop that might indicate a platform issue, crisis spikes in mentions). Weekly — review content performance and spot trends across recent posts. Monthly — full analysis, benchmark comparisons, strategy adjustments. Obsessive daily checking of every metric is a productivity trap — set up alerts for anomalies and do real analysis on weekly and monthly cadences.

What's Share of Voice and how do I calculate it?

Share of Voice (SOV) measures your brand's percentage of total industry conversation. The formula: Share of Voice = (Your Brand Mentions ÷ Total Industry Mentions) × 100. You'll need a social listening tool to capture untagged mentions (Brandwatch, Sprout Social, Brand24, Meltwater). SOV correlates strongly with market share over 12-24 month timelines and is one of the few brand health metrics with rigorous business evidence behind it. Track it quarterly against your top 3-5 competitors.

Wrapping up

Here's what I want you to take away: metrics are a tool, not a trophy. Nobody at your company should be impressed that you track 47 KPIs — they should be impressed when one of them moves meaningfully and you can explain exactly why.

Pick 3-5 KPIs that ladder up to business goals. Track them relentlessly. Track everything else as context. Report on them with honesty — including the ones that went down. Build the muscle to say "this campaign failed and here's what we learned" because that's what separates marketers from dashboards.

And for the love of god, stop reporting follower counts as wins. It's 2026.

Ready to stop spreadsheet-wrangling your metrics and get everything in one clean dashboard? PostEverywhere's analytics unifies reach, engagement, conversions, and video metrics across every platform — with ROI calculations done automatically. Start your 14-day free trial today. No credit card required.

For further reading, I genuinely recommend Hootsuite's social media metrics overview and Sprout Social's metrics guide — both keep their content updated and both treat the subject with more rigour than 90% of the other stuff out there. And if you haven't read Avinash Kaushik's original framework piece, do it this week. Fifteen years old and still the clearest thinking on the topic.

Now go measure something that matters. And if you're using PostEverywhere to schedule your content, generate it with AI, or figure out your best time to post — the analytics are built right in. Pricing starts at $19/month, and there's a 14-day free trial on every plan with no credit card required.