Social Media KPIs: 20 Metrics to Track, With Formulas

Social media KPIs (key performance indicators) are the handful of numbers you agree, in advance, will tell you whether your social media is doing its job. A metric is anything a network counts: likes, views, saves, profile taps. A KPI is a metric you have picked on purpose, tied to a goal, given a target and a date, and that you will act on when it moves. Every KPI is a metric. Most metrics are not KPIs.

That difference matters because the native dashboards show dozens of numbers and almost all of them go up and down for reasons you cannot control. Meta's reference for Instagram alone lists nine metrics for a single post and another set for the account, and every other network has its own list, often with a different name for the same idea. If you try to watch all of them you end up watching none. The point of choosing social media KPIs is to shrink that wall of numbers down to a few you can explain to someone else in one sentence each.

So when people search for a KPI for social media marketing, what they usually need is not a longer list. They need a short list that matches what they are trying to get from social in the first place: people who know the brand, people who interact with it, people who click through, people who sign up or install, and customers who get answers. This page gives you 20 candidates grouped by those goals. Pick three to six of them, not all 20.

One rule runs through the whole guide. Followers, interactions and reach are the headline numbers. They are the three that every network reports, that anyone can understand, and that tell you whether an account is growing, whether people care about what it posts, and how many people actually saw it. Everything else on this page is there to explain why those three moved, or to connect them to money.

We wrote this page on October 28, 2026. For every KPI, we went to the platform's own documentation to see what it actually counts, rather than repeating definitions from tool vendors that often lag behind platform changes. That meant Meta's developer reference for Instagram media insights, Instagram account insights, Threads insights and Facebook Page insights; LinkedIn's help pages on post analytics, video analytics and follower analytics; YouTube's impressions and click through rate FAQ and the YouTube Analytics API metrics list; Pinterest's analytics help; the TikTok video object in TikTok's developer docs; Mastodon's Status entity; and Telegram's channel statistics schema.

For traffic and conversion KPIs we used Google's own help on campaign URLs, the default channel group and key events, plus Apple's App Store Connect help on campaign links and App Store metrics. For what marketing teams actually track, we used Sprout Social's 2025 press release and Chief Marketer's write up of the same survey. For benchmarks we used the 2026 Rival IQ benchmark report, now published by Quid, and Buffer's 2026 engagement study. Each source is linked where it is used and listed at the bottom of the page.

The formulas are standard arithmetic, written out so you can paste them into a spreadsheet. The worked examples use a made up account, a small habit tracker app run by Alex, so you can see the maths with round numbers. They are illustrations, not benchmarks. Where a real benchmark exists, we name the study and the date, and we say how its denominator differs from yours, because that one detail explains most of the disagreement between published numbers.

A metric is a count or a rate the platform hands you. A KPI is a metric with four extra things attached: a goal it serves, a baseline (where it is today), a target (where you want it by a date), and an owner who will do something when it misses. Without those four, a number is just information. With them, it becomes a decision you have already half made.

Here is the test we use. Ask of any number: if this went down by a third next month, would we change what we post, where we post or how often? If the honest answer is no, it is a metric worth glancing at, not a KPI. Saves on a LinkedIn post, for example, are interesting but rarely change a plan. Followers gained on the network you are betting on, on the other hand, should change the plan quickly if they stall.

This is also why the same KPI list does not fit every account. A local bakery cares about reach in its city and direct messages asking about orders. An app founder cares about product page visits and installs. A B2B SaaS team cares about clicks to a demo page. A creator cares about followers and watch time. All four can share the headline trio of followers, interactions and reach, then branch into different KPIs below it.

If you have never written KPIs before, start by auditing what you have. Our social media audit template walks through pulling a baseline from each account, which is the first number every KPI needs. Without a baseline, a target is a guess.

If you only track three social media KPIs, track these. They work on every network, they are easy to explain to a founder or a client, and together they answer the three questions anyone asks about an account: is it growing, do people care, and how many people saw it.

Followers comes first because it is the one number that compounds. Each follower is someone the network will try to show your next post to. On its own, a total is a vanity number; tracked as followers gained and lost per period it becomes a real KPI. Instagram's account insights expose this directly as follows and unfollows, Threads reports followers_count, and LinkedIn's follower analytics show new followers over time and update the total once a day.

Interactions comes second. It is the count of things people did with a post: likes, comments, shares, saves, reposts, replies, quotes. Meta defines Instagram's total interactions as likes, saves, comments and shares, minus unlikes, unsaves and deleted comments. Interactions is better as a headline than engagement rate because it does not depend on which denominator you chose, and it adds up cleanly across networks.

Reach comes third. It is the number of unique accounts that saw a post at least once. Meta's documentation calls Instagram reach an estimated count that does not include repeat views by the same account; LinkedIn's equivalent is members reached, defined as distinct members and pages who viewed a post with repeat views not counted. Reach tells you the size of the audience you actually got, which is the denominator for most of the rates further down.

Put these three at the top of every report, per network, with the change since the last period beside each. A month where followers rose, interactions held and reach fell tells a very different story from one where all three fell together, and you can read it in ten seconds.

Tip: if you post with Spreadcast, its Analytics tab already leads with these numbers. It opens on followers (total, gained and growth), then interactions, with followers by network beside the chart, and lists likes, comments, reach and views for the posts it published, wherever the network reports them. You see the headline trio without exporting anything. See what Analytics shows.

Awareness KPIs answer one question: are more of the right people seeing us than last month? They sit at the top of the funnel, so they move first and fastest when you post more often or try a new format, and they are where most young accounts should start.

Followers and follower growth rate (KPIs 1 and 2) are the backbone. Use growth rate rather than raw gains when you compare periods or accounts of different sizes. Gaining 140 followers is a strong month at 1,100 and a quiet one at 50,000. Count losses too: Instagram's follows_and_unfollows metric splits the two, but note that Meta does not return follower metrics for accounts with fewer than 100 followers, so very new accounts have to read the number off the profile and keep their own log.

Reach (KPI 3) is the best single awareness number because it counts people, not displays. Views (KPI 4) counts every time something was shown or played, so one person scrolling past a post twice counts twice. On Instagram, views now replaces impressions: Meta deprecated the impressions metric for media created after July 2, 2024, and describes views as the number of times a post was played or displayed, a metric it still labels as in development. Facebook follows the same pattern with page_media_view, the number of times content was played or displayed, beside a separate count of unique viewers.

Profile visits (KPI 5) is underrated. A post that sends people to your profile is doing awareness work that likes do not show: those visitors are reading your bio and deciding whether to follow. Instagram reports profile_visits per post, and Threads reports views of your profile at the account level. Dividing visits by reach gives you a profile visit rate you can compare across posts.

Share of voice (KPI 6) is the one awareness KPI the networks do not hand you. It compares how often your brand is mentioned with how often named competitors are, over the same period. Most small teams count it by hand once a quarter with a saved search on each network. It is worth the effort only if you are in a crowded category where buyers compare a few names.

Which of these are social media KPIs for brand awareness that a small team should actually track? Reach and follower growth rate, per network, monthly. Add profile visits on Instagram if you post there, because it is the clearest signal that a post made people curious. If you are building awareness across many networks at once, our social media statistics page shows where the audiences are in 2026, and the Threads and Bluesky pages cover the two newer networks in detail.

Benchmarks help you sanity check, not set targets. The trouble is that published studies use different denominators. The 2026 Rival IQ report, published March 5, 2026, divides interactions by followers and puts the all industry median engagement rate per post at 0.30% on Instagram and 2.01% on TikTok. Buffer's 2026 study divides by reach instead and reports 2025 medians of about 5.5% on Instagram and 4.6% on TikTok. Both can be right; they are measuring different things.

So the practical rule is this: decide your denominator, then compare yourself with yourself. A KPI target such as "raise engagement rate by reach from 6% to 7% on Instagram this quarter" is far more useful than "beat the industry average", because you control the formula and the data on both sides.

One more nuance from Buffer's data: on Bluesky it does not publish a rate at all and reports interactions per post instead, a median of about 4 likes, comments and reposts combined. That is a fair way to track any small or new network where reach is not reported. If you want the full table of rates per network and per follower tier, the engagement rate calculator page keeps it current.

Most KPI lists leave out the one number that sits underneath all the others: did you actually post what you planned? It sounds too basic to be a KPI. In practice it is the first thing to break on a small team, and when it breaks every other KPI becomes unreadable. If reach fell in a month where you posted six times instead of sixteen, you have learned nothing about your content and everything about your calendar.

Track it weekly as posts published divided by posts planned, per network. Anything under 80% for two weeks running is a signal to cut the plan or change how the posts get made, not to try harder. A plan you hit at 100% with three posts a week beats a plan of daily posts you hit at 40%, because the first one gives you clean data and a steady presence.

The usual reason consistency fails is not ideas, it is production. Each post needs a design in the right size for each network, a caption that fits each network's limits and someone to schedule it. Our guides on batching a month of content in one afternoon and what to post when you run out of ideas cover the planning side, and how to automate social media posts covers the tools.

This is the gap Spreadcast is built for. Today you might design each post in a design tool, resize it for Instagram, LinkedIn and Pinterest, write three captions and schedule them one by one. With Spreadcast you write one line or pick a post type, it designs the post in your brand and writes a caption for each network, and it publishes to Instagram, Facebook, Threads, LinkedIn, Pinterest, TikTok, YouTube, Bluesky, Mastodon, Telegram and Discord on your schedule. For apps it goes further and posts on its own when a new version ships or the app hits a ratings milestone. Pro is $29 a month for 10 connected accounts with no cap on posts. See the plans.

If you have ever added up "impressions" across five networks and felt uneasy, you were right to. The word means something slightly different on each one, and a few of the differences are big enough to change a decision.

Start with views and impressions. Instagram now reports views as every time a post was played or displayed, and Meta's own reference marks it as in development. LinkedIn defines impressions as the number of times a post was displayed on screen and adds that the number is an estimate and may not be precise. A LinkedIn video only counts as a video view after 2 seconds of watching. YouTube counts an impression only for a registered thumbnail view on YouTube itself and leaves out external websites and end screens. Adding these together gives you a number that looks precise and means nothing.

Reach has a similar problem. Instagram reach and LinkedIn members reached both count unique viewers, but each network deduplicates only within itself, so the same person who saw your post on both is counted twice in any total. Mastodon and Bluesky do not report reach to you at all. Telegram reports average views per post rather than unique reach, which is closer to a views number than a reach number.

Thresholds and delays add a third layer. Meta withholds follower metrics on Instagram below 100 followers, Facebook Page Insights need 100 likes, and Threads demographics need 100 followers. Instagram account data can be delayed up to 48 hours, and Pinterest marks the last 48 hours as estimates that may go down. If you pull numbers on the first of the month, the last two days of the previous month are often still moving.

The fix is not a perfect formula. It is three habits. Compare each network with itself over time rather than with other networks. Pull monthly numbers three days after the month ends, not on day one. And when you do total across networks, total interactions and followers gained, which are counts of real actions, rather than impressions or views.

The most useful public data on which KPIs teams actually use comes from Sprout Social's 2025 Impact of Social Media survey. According to Sprout's press release, Glimpse surveyed 1,200 marketing leaders in the US, UK and Australia between May 30 and June 20, 2025. Most social teams measure success by engagement (68%) and conversion rates (65%). Chief Marketer's coverage of the same report adds revenue at 57%, efficiency at 55% and discoverability at 51%.

Two things stand out. First, engagement still leads, which matches what we see: it is the easiest number to get and the first one the dashboards show. Second, almost two thirds of teams already try to measure conversion, which means the reporting bar has moved past likes. If you report to a boss, an investor or a client, expect to be asked what social sent to the website or the App Store.

The same release says only 44% of leaders rate their social team as expert at measuring business impact, and that leaders often put this down to a lack of sophisticated reporting. Our reading: most teams do not lack a tool so much as the setup work, tagged links, marked key events and campaign links, that turns clicks into something a report can count. The next three sections cover exactly that setup.

The survey also shows why leaders care. Respondents credit social with customer acquisition (60%), customer loyalty (58%) and revenue (56%). Those are business words, not social words. A KPI report that ends at reach and interactions leaves the people paying for social to guess the connection; one that adds clicks, signups or installs makes it for them. If you are putting a budget around this, our guide to a startup marketing budget shows where the social line sits.

Native analytics end at the tap. To follow a visitor from a post to a signup, you need the link itself to carry its origin. That is what UTM parameters do. Google's URL builder help lists the ones that matter: utm_source for the referrer (instagram, linkedin), utm_medium for the channel (social), and utm_campaign for the product, promotion or launch. It also lists utm_content, which is handy for telling apart two links in the same post.

A tagged link for Alex's app looks like this in practice: the landing page address, then ?utm_source=instagram&utm_medium=social&utm_campaign=october_update. Use one source value per network and keep them lowercase. Google's help recommends one unique utm_source for each distinct platform, and warns that missing parameters show up as (not set) in reports.

The medium value decides which channel the visit lands in. GA4's default channel group puts a visit in Organic Social when the source matches its list of social sites or the medium is one of a fixed set of values, including social, social network and social media. Paid Social needs a social source plus a paid medium such as cpc or paid. If you write utm_medium=insta or utm_medium=post, the visit falls out of both social channels and your social numbers look smaller than they are.

Then mark what counts as a result. In GA4 you identify or create the event (a signup, a trial start, a purchase) and mark it as a key event. Google's help is clear that marking applies from the time you create the mark and does not change historic data, and that standard reports can take up to 24 hours to show the change. So mark key events before a campaign starts.

Once both pieces are in place, three KPIs from the table come free: sessions from organic social (KPI 14), key events from social, and social conversion rate (KPI 15). Open the Traffic acquisition report, filter to the Organic Social channel and add your key event as a column. That is the conversion slide of your monthly report.

If the business is an iPhone app, the conversion that matters happens on the App Store, not on your website, and Google Analytics cannot see it. Apple's answer is campaign links. A campaign link is your App Store URL with a provider token (pt), which identifies your developer account and stays the same, and a campaign token (ct), which you change for each campaign. Apple allows up to 30 characters for the campaign token.

When someone taps the link and lands on your product page, App Store Connect attributes impressions, product page views, first time downloads, usage, sales and subscriptions to that campaign. A download counts if it happens within 24 hours of using the link, and if a person taps several campaign links, the most recent one gets credit for later sales. Campaigns appear only after 24 hours and once they have first time downloads from at least five people, and each metric shows only once it reaches five in the selected range, so small campaigns can look empty for a while.

The KPI to track is first time downloads per campaign (KPI 16), with a campaign token per network or per post type: instagram_update, linkedin_launch, threads_tips. Then compare it with product page views from the same campaign. Apple's App Store metrics reference defines conversion rate as total downloads and preorders divided by unique device impressions, which is the store side of the same funnel.

Building those links by hand is fiddly, so we made a free App Store campaign link generator that builds them in bulk. For what a normal product page converts at, see App Store conversion rate benchmarks, and for what paid installs cost you to compare against, cost per install benchmarks. Once installs are flowing, the LTV to CAC ratio tells you what each one is worth.

Spreadcast was built first for this case. Paste your App Store link at Spreadcast's start page and it builds a brand kit from the listing, then watches the listing every day; when a new version goes live or the app reaches a ratings milestone, it designs the post with your real screenshots in an iPhone frame, writes a caption per network and publishes it, on full autopilot or after one tap of approval. Each release becomes a post with a link back to the product page, so the install KPI has something to count every time you ship.

A good social media KPI report fits on one page and can be read in two minutes. Anyone who wants more can open the tracker. We put five blocks on it, in this order.

First, the headline trio per network: followers (with gained and lost), interactions and reach, each with the change against last month and against target. Use plain arrows or words, not colour alone. This block answers "are we growing" before anyone asks.

Second, the goal KPIs: the two or three you picked for this quarter's goal. If the goal is signups, this is sessions from organic social, signups and social conversion rate. If the goal is installs, it is first time downloads by campaign and the product page views behind them. One line each on why it moved.

Third, the top and bottom posts of the month, three to five of each, with a thumbnail, the network, reach and interactions. This is the part people actually read, because it turns numbers into examples. Write one sentence on what the best ones had in common: a format, a topic, a hook.

Fourth, output: posts published against plan, response time and response rate. Keep it short. It is there so that a dip in reach in a month where you posted half your plan is read correctly.

Last, one decision for next month. Not a list of learnings, one change: post carousels on LinkedIn twice a week, move the Tuesday post to Thursday, drop Pinterest and add Threads. A report that ends in a decision gets read again next month to see if the decision worked. For carousel and image sizes per network when you act on it, keep our social media image sizes and character limits handy.

There are three ways to build a social media KPI dashboard, and the right one depends on how many networks you post to and how much time you will give it.

Native analytics only. Free and the most accurate, because the numbers come straight from the network. The cost is time: with five networks you open five dashboards, each with its own date picker, names and delays. This works for one or two networks and a monthly report.

A spreadsheet fed by hand. Still free, and it gives you the tracker above with your own formulas and history that does not disappear when a network changes its dashboard. Most small teams should start here. Budget about an hour a month to copy the numbers in, three days after month end so the late data has settled. Google Sheets or Excel both work; a chart per KPI per network is enough.

One place that pulls the numbers for you. Social media tools read the same platform APIs described in this guide and chart them together. The trade off is cost and the fact that a tool can only show what the API exposes, which is why tools sometimes lose a metric when a platform deprecates it, as happened when Meta retired impressions on Instagram.

If you already publish through Spreadcast, the third option comes with the posting. Its Analytics tab leads with followers per network and interactions, adds an engagement rate, and lists likes, comments, reach and views for each post it published, wherever the network reports them. If you work in Claude or ChatGPT, the Spreadcast MCP server can read the same analytics, so you can ask for last month's numbers in a chat and get the headline trio back without opening a dashboard.

A vanity metric is a number that looks good and changes nothing. Total followers on its own is the classic example, which is why this page tracks followers gained and growth rate instead. Raw impressions and views are the other common one. They are fine as context but too easy to inflate with a single lucky post, and they mean different things on every network.

Mixing denominators is the mistake we see most often in reports. Engagement rate by followers one month and by reach the next can look like a tenfold jump that never happened. Write the formula into the tracker and never change it mid year. If you must change it, restate the past months with the new formula so the line stays comparable.

Reporting totals across networks hides the network that matters. A combined reach number that rose 20% can hide a 30% fall on the one network where your customers are. Keep a row per network, then total interactions and followers gained if you need a single number.

Setting targets from someone else's benchmark is the next trap. A median from a study of large brands tells you little about an account with 800 followers that posts twice a week. Benchmarks are for a sanity check; your baseline is for targets.

Last, tracking too many KPIs. Twenty KPIs is a menu, not a dashboard. Teams that track everything usually act on nothing, because every month something went up and something went down. Three to six is the range that gets read.

If your KPIs keep pointing at weak creative rather than weak timing, the fix is upstream of analytics. Our guides on what makes a social media post work and keeping a brand consistent without a designer cover the design side of the numbers.

Most social media KPIs move for one boring reason: the account posts good looking, on brand content often enough, on the networks where its audience is. The KPI work above tells you whether that is happening. Spreadcast is for the part that usually stops it from happening, which is making and publishing the posts.

The before looks like this. A founder or a one person marketing team decides to post four times a week on Instagram, LinkedIn and Threads. Each post means opening a design tool, finding the brand colours and fonts, placing a screenshot, exporting three sizes, writing three captions that fit three limits and scheduling each one. Two weeks in, a release slips or a customer issue comes up, and the plan drops to one post a week. Posting consistency falls to 25%, reach follows it down, and the monthly report shows a decline nobody can explain from the numbers alone.

Now the after. You paste your App Store link, or your website, or a short description of the brand, at the start page, and Spreadcast builds a brand kit with your icon, colours and screenshots. You write one line in Create, or pick one of twelve post types for apps such as a testimonial, a how to or a feature, and it designs a single image post or a carousel in your brand, writes a caption per network and puts it on the calendar. You approve it, or let autopilot publish. For an app, new versions and ratings milestones turn into posts without you starting anything.

Then the KPIs read themselves. Analytics shows followers and interactions per network and likes, comments, reach and views per post, so the headline trio and the top posts of the month are already there when you write the report. If you prefer to work in Claude or ChatGPT, the same actions (design a post, write captions, schedule, approve, read analytics) are available through the MCP server.

What it costs: Free makes 3 posts to download with a small Spreadcast mark and no publishing, enough to see the design quality. Pro is $29 a month or $290 a year with clean posts, no cap, publishing to 10 social accounts across any mix of networks, autopilot or one tap approval, and posting from Claude or ChatGPT. Studio is $79 a month for 5 apps and 50 accounts, each with its own brand. If you would rather not subscribe, a Post Pack is $19 once for 10 posts designed and published, and a 30 Day Pass is $39 once for everything in Pro for 30 days. Paid plans have a 7 day full refund. For comparison with hiring help, our guide on what a social media manager costs prices the other options. Start with your App Store link or website.