Social Media Metrics and KPIs in 2026: What to Track, How to Calculate It, and What to Ignore

Social media metrics are the numbers platforms report about your posts and accounts. KPIs are the few metrics you choose to judge success because they connect to a business goal. Most teams need three KPIs: one for attention (reach or views), one for resonance (engagement rate, saves or shares) and one for action (clicks, leads or sales), reviewed monthly against their own baseline.

The distinction matters because every network now reports dozens of numbers, and the definitions keep changing. A metric becomes a KPI only when you have decided in advance what a good result looks like and what you will do if it moves. Everything else is context. This guide covers how the platforms define their metrics (quoting their own help pages), the formulas, how the major benchmark studies differ, which KPIs fit which goal, business and network, a step-by-step setup and the mistakes that make social reports meaningless.

Key findings

• "Views" replaced impressions on Meta. On 14 November 2024, Meta announced that views would become the primary metric for Facebook reels, videos, posts and stories, calculated the same way as on Instagram, replacing plays and, for photo and text posts, impressions.
• LinkedIn now reports saves and sends. LinkedIn's post analytics include members reached, in-network and out-of-network impressions, saves, sends, profile viewers and followers gained per post.
• Engagement rates differ by denominator. Hootsuite lists six engagement rate formulas, each with a different denominator (reach, followers, impressions, daily followers, views, spend), which is why benchmark studies disagree.
• Median engagement by reach or impressions is far higher than by followers. Buffer's 2026 report (52 million+ posts) puts LinkedIn at about 6.2% and Instagram at about 5.5%, while Socialinsider, which divides by followers, puts Instagram at 0.48% and Facebook at 0.15% for 2025.
• YouTube gives a CTR range. YouTube says half of all channels and videos have an impressions click-through rate between 2% and 10%.
• Replies move results. Buffer's analysis of nearly 2 million posts found replying to comments was associated with about 42% more engagement on Threads and 30% more on LinkedIn.
• Executives and teams see value differently. In the 2025 Sprout Social Index, almost half of 900 practitioners said executives only somewhat understand the value of social. Good KPIs close that gap.

What's changed in social media metrics in 2026

Views are the new common currency

Meta's switch to views means a Facebook photo post now counts each time it appears on screen, including repeats, and Meta warns that view numbers may be higher than previous impression counts. If your year-on-year report compares 2024 impressions with 2026 views, the jump may be a definition change, not a performance change. Meta also replaced watch time with minutes viewed and average minutes viewed.

Shares and sends matter more

Meta's Instagram Feed ranking explanation lists the likelihood that you will send a post to someone in a direct message as one of its predictions, and LinkedIn now reports sends on LinkedIn for every post. Private sharing has become a visible, trackable signal of value.

In-network vs out-of-network reach

LinkedIn splits impressions into those from people who follow or are connected to you and those from people who are not. A rising out-of-network share means your posts are being recommended to new people, which is one of the clearest discovery metrics any network offers.

APIs expose more metrics

The Threads insights API now returns views, likes, replies, reposts, quotes and shares for each post, and clicks and follower counts for accounts, though Meta marks views and shares as in development. Scheduling and analytics tools can therefore report Threads alongside the older networks.

Retention windows are finite

LinkedIn keeps discovery and engagement counts for 1,000 days, members reached for 400 days and demographic breakdowns for 180 days. Platform analytics are not an archive. Export monthly if you want year-on-year comparisons.

Metrics vs KPIs: the difference

A metric is any number. A KPI is a metric with three things attached: a goal it serves, a target, and an action you will take if it misses. "Engagement rate" is a metric. "Instagram engagement rate by reach of at least 5% each month, because our goal is community, and if it drops two months in a row we will review our format mix" is a KPI.

The four layers of social media metrics

Think of metrics as a funnel with four layers. Each layer answers a different question, and you need at least one metric from each layer that matches your goal.

Attention metrics

Reach

The number of unique accounts that saw a post or your profile during a period. LinkedIn calls it members reached and describes it as distinct members and Pages who saw your post, excluding repeat views, and notes that it is an estimate. Reach tells you how far content travelled. See our glossary entry on reach for platform differences.

Impressions and views

The total number of times content was displayed, including repeats by the same person. Definitions vary: LinkedIn counts a Page post impression when at least 50% of the post is on screen for 300 milliseconds or more, according to its Page analytics help; Pinterest defines impressions as the number of times Pins were on screen; Meta now calls this views. Impressions are always equal to or higher than reach. More in our impressions entry.

Video views and watch time

Each network defines a view differently. LinkedIn counts a video view at two or more continuous seconds, including replays. Meta counts a view each time a reel or video is played. YouTube's analytics focus on watch time and average view duration, which its help page defines as the average time viewers spent watching. Comparing raw views across networks is misleading; compare watch time and completion within one network instead.

Follower growth rate

Formula: (new followers minus lost followers) divided by followers at the start of the period, times 100. Growth rate is more meaningful than the raw follower count, which mostly reflects history. LinkedIn now shows followers gained per post, which tells you which posts actually grow the account.

Resonance metrics

Engagement rate

The share of people who interacted with a post. The formula you choose changes the answer dramatically, so pick one and use it consistently.

What counts as an engagement also varies. LinkedIn's Page engagement rate counts clicks, reactions, comments and shares divided by impressions. Pinterest's counts saves, Pin clicks and outbound clicks divided by impressions, as its analytics help page explains. Write down your definition. Our free engagement calculators for Instagram and LinkedIn apply the formulas for you, and the engagement rate glossary entry covers the details.

Saves, shares and sends

Saves mean someone wants to come back to the post. Shares and sends mean they valued it enough to pass it on. Both take more effort than a like, and both feed ranking systems. If you can track only one resonance metric beyond engagement rate, track saves on Instagram and Pinterest and shares or sends elsewhere. See save rate.

Comments and replies

Comments show conversation. Read them, do not just count them. Ten thoughtful comments are worth more than fifty one-word replies. Track your own reply rate too: Buffer's data links replying with sizeable engagement lifts on Threads, LinkedIn and Instagram.

Action metrics

Link clicks and click-through rate

Click-through rate (CTR) is link clicks divided by impressions, times 100. It tells you how well a post turns attention into a visit, and it is the key metric for traffic goals. LinkedIn reports visits to links in each post, counting repeat clicks; Pinterest reports outbound clicks. On YouTube, impressions CTR measures how often a thumbnail impression became a view, and half of channels fall between 2% and 10%. See click-through rate.

Profile visits

Useful wherever links in posts are limited or deprioritised. LinkedIn reports profile viewers from each post. A spike after a post means it made people curious about you.

Direct messages and enquiries

For service businesses, a message asking "how much for..." is often the most valuable metric of all. Track the count weekly, even if manually.

Outcome metrics

Outcome metrics live outside the social platforms, in your website analytics, CRM or shop. To connect them to social, tag every link with UTM parameters. Google's Analytics help page says you should always use utm_source, utm_medium and utm_campaign, and that values are case sensitive. Our UTM builder keeps tags consistent.

• Conversions from social: sign-ups, purchases or bookings from social-tagged sessions.
• Conversion rate: conversions divided by social sessions, times 100.
• Revenue from social: total value of conversions attributed to social.
• Cost per result: spend divided by conversions, for paid campaigns.

For turning these into a return-on-investment figure, see our social media ROI guide.

Metric definitions by network

The same word means different things on different networks. Use this table when you combine data from several networks in one report.

For network-specific deep dives, see our guides to Instagram metrics, LinkedIn metrics, TikTok metrics and X metrics.

Input metrics: the numbers you control

Most social metrics are outputs: you influence them, but the audience and the algorithm decide. A few input metrics are fully in your control, and they explain a surprising share of the outputs. Track them alongside your KPIs.

Active weeks is the most underrated metric in social media. In Buffer's 26-week study of more than 100,000 users, people who posted in 20 or more of the 26 weeks earned about 450% more engagement per post than those who posted in 4 weeks or fewer. If your engagement drops, check your active weeks before you rewrite your content strategy.

Publishing success rate matters too. A post that fails to publish because a connection expired is a silent gap that no analytics dashboard will flag as a problem. In Postprism, every post carries a status (draft, scheduled, publishing, published or failed), and opening a post shows the result for each network with the error message the platform returned. A failed network can be retried on its own once the cause, such as an expired connection, is fixed, without re-sending to the networks that already published. Checking for failed posts once a week takes a minute and keeps your active-weeks number honest, because a post that never went out does not count, however good it was. Our guide to scheduled posts not publishing covers the common causes.

Benchmarks: why the studies disagree

Benchmarks are useful for sanity checks, but they vary wildly because each study uses a different formula and sample. Here is how three large studies compare.

Buffer's figures come from more than 52 million posts sent through its tool; Hootsuite's from its 2026 industry benchmarks; Socialinsider's from 70 million brand posts from 2024 to 2025. Same networks, very different numbers. The lesson: never compare your by-followers rate with someone else's by-reach benchmark. Your own trend over time is a better yardstick than any industry average.

Benchmarks by industry

Hootsuite's 2026 industry table shows how much the sector matters. On Instagram, nonprofits average 5.5% and government 5.0%, while consumer goods and retail sit at 3.6%. On LinkedIn, dining and hospitality lead at 4.5%. If you use a benchmark, use one from your industry and with your formula.

Metrics by content format

Different formats earn different kinds of engagement, so judge each format by the metric it is built for. A carousel should be judged on saves and swipes, a video on watch time, a link post on clicks.

The figures come from Buffer's analysis of content formats. The practical lesson is that a link post with a lower engagement rate than a carousel is not necessarily failing. If its job is traffic, judge it on clicks and what those visitors did next.

Metrics by account size

Account size changes which numbers are meaningful. Small accounts have noisy data; large accounts have averages that hide what is happening.

Under 1,000 followers

Percentages swing wildly on small numbers, so track counts: comments, saves, enquiries and new followers per post. One thoughtful comment from a potential client matters more than any rate. TikTok's own explanation of its recommendation system says follower count is not a direct factor in what it recommends, so small accounts can and do get reach; judge each post on its own.

1,000 to 10,000 followers

Rates start to stabilise. Engagement rate by reach and follower growth rate become useful monthly KPIs. This is the stage to set your baseline.

10,000 to 100,000 followers

Reach and out-of-network share matter more, because your existing audience is no longer the main source of growth. Track which posts bring new followers, not just which get the most likes.

Over 100,000 followers

Averages hide outliers. Report medians, segment by format and pillar, and focus on outcome metrics. Engagement rate by followers tends to fall as accounts grow, so use by-reach rates for fair comparisons over time.

Community and listening metrics

Not every useful number sits in post analytics. Three community metrics are worth adding once the basics are in place.

• Response time: how long it takes you to reply to comments and messages. The 2025 Sprout Social Index found 73% of consumers say they will buy from a competitor if a brand does not respond on social, which makes response time a commercial metric.
• Reply rate: the share of comments you answer. Buffer's 2026 report found replying was associated with more engagement on six networks, from about 5% on Bluesky to about 42% on Threads.
• Share of voice: your mentions divided by total mentions of you and your main competitors, times 100. It needs a listening tool, but it is the best awareness metric for crowded markets.

Sentiment (the share of positive, neutral and negative mentions) is useful too, but treat automated sentiment scores with caution and read a sample yourself each month.

Organic vs paid metrics

Keep organic and paid metrics in separate columns. Paid campaigns buy reach, so mixing them inflates organic performance and hides whether your content is working on its own. For paid, the core metrics are cost per result (spend divided by conversions), cost per click and cost per engagement, which Hootsuite defines as amount spent divided by total engagements. For organic, the cost is mostly time, which is why our ROI guide counts hours as well as money. A good habit: boost only posts that have already performed well organically, then compare the paid cost per result with your organic baseline.

Setting targets without industry benchmarks

Given how much benchmark studies disagree, the most reliable target is your own history. A simple method:

1. Take the last three months of each KPI and calculate the median.
2. Set next quarter's target 5% to 15% above that median, depending on how much you are changing.
3. If you are starting from zero, run one month with no target, then use that month as the baseline.
4. Revisit targets quarterly, not monthly. A single month is too noisy.

This approach also keeps you honest when a platform changes a definition. If Meta's views jump because they now include repeat appearances, your baseline jumps with them, and the comparison stays fair.

Choosing your KPIs by goal

Pick KPIs from the layers that match your goal. Here are sensible defaults.

Rule of thumb: one KPI per layer you care about, three in total. If a metric would not change any decision you make, it is not a KPI.

KPIs by type of business

Solo creators and founders

Keep it to follower growth rate, engagement rate by reach and one action metric (newsletter sign-ups, enquiries or profile visits). Review monthly. Anything more is time taken from creating.

Local businesses

Reach in your area, saves or shares, and enquiries (messages, calls, bookings). Ask new customers how they heard about you; it is the cheapest attribution there is.

B2B companies

LinkedIn impressions and out-of-network share, comments and reposts, and leads or demo requests tagged with UTMs. Pair social reports with CRM data so you can see which posts influenced pipeline.

Ecommerce brands

Reach, saves and revenue from social-tagged sessions. Track cost per result separately for paid social. See social media for ecommerce for more.

Agencies

Agree three KPIs per client at onboarding, in writing, and report on exactly those every month. Changing KPIs mid-contract is how agencies lose trust.

How to set up your KPI system: a step-by-step method

1. Write one goal per network. "LinkedIn: generate demo requests. Instagram: build community." Not every network needs the same goal.
2. Pick three KPIs per goal from the tables above: attention, resonance, action or outcome.
3. Choose one formula for engagement rate and write it in your report template.
4. Tag every outbound link with UTM parameters, using one naming convention.
5. Record a baseline. Average the last three months for each KPI. That is your starting point, not an industry benchmark.
6. Set targets as a percentage improvement on baseline, for example 10% more reach in a quarter.
7. Decide the action for each KPI if it misses: "if CTR drops below baseline for two months, test new hooks and link placement".
8. Export monthly. Platform retention windows are finite; keep your own record.
9. Review monthly, decide quarterly. Look at numbers monthly, change strategy quarterly, so noise does not drive decisions.
10. Share a one-page scorecard with stakeholders: three KPIs, trend, what you learned, what you will do next.

Worked example: calculating the metrics for one post

A physiotherapy clinic posts an Instagram carousel, "Five desk stretches", with these numbers after seven days: 4,200 reach, 5,100 views, 180 likes, 22 comments, 95 saves, 41 shares, 3,900 followers at the start of the week, and 63 link clicks from the bio link tracked with UTMs.

The high save rate tells the clinic that practical, reference-style content works. Next month they plan two more "save this" carousels and compare save rates. The 1.2 views per person means some people came back to the carousel, which fits the save behaviour.

KPIs by industry

Industry changes both the goal and the realistic range. Hootsuite's 2026 industry benchmarks show the spread: on Instagram, healthcare averages 4.5% and finance 4.1%; on LinkedIn, consumer goods and retail averages 4.3% and government 2.8%. Use the industry figure for context, then set targets from your own baseline.

Local businesses should also remember where their audience is. Pew's 2025 survey found 71% of US adults use Facebook and about half use it at least daily, so Facebook reach and messages often matter more for a neighbourhood business than follower growth on a newer network.

Worked example: a B2B LinkedIn month

A 20-person HR software company posts 12 times on its LinkedIn Page in a month, with the goal of demo requests. Its three KPIs are impressions, engagement rate (LinkedIn's definition, including clicks) and demo requests from UTM-tagged links.

Two posts, both explaining a specific HR rule with a short example, drove six of the nine demo requests. The team adds two more "explain one rule" posts to next month's calendar and drops one generic industry-news post. That is a KPI report doing its job: three numbers, one clear decision each.

The quarterly KPI review

Monthly reports steer; quarterly reviews decide. Once a quarter, spend an hour on four questions: did each KPI beat its target, which pillars and formats drove the best results, are these still the right KPIs for our goals, and what will we stop doing? Write the answers down and carry them into next quarter's content plan. Our social media audit checklist is a good companion for the deeper annual version of this review.

Leading vs lagging indicators

Most teams look only at lagging numbers, then wonder why their reports never change what they do. A metric that arrives a month after the post is useful for judging strategy, not for steering next week. The fix is to pair every lagging KPI with one leading indicator you can act on quickly.

Outcome metrics like revenue are lagging: they move weeks after the content that caused them. Attention and resonance metrics are leading: they move within days. Watch leading indicators weekly to steer, and judge lagging indicators quarterly. Saves and sends are good leading indicators for future clicks; profile visits are a good leading indicator for follower growth.

Metrics to treat with caution

• Follower count: mostly history, and easily inflated by giveaways that attract the wrong people.
• Likes in isolation: easy to give, a weak signal of value.
• Video views across networks: definitions differ so much that comparisons mislead.
• Year-on-year impressions on Meta: the switch to views changed the definition in late 2024.
• Industry averages: they vary by formula, sample and niche. Use your own trend.
• Single viral posts: one outlier distorts monthly averages. Report the median too.

How to report social media KPIs

A good report is short and leads to decisions. A practical monthly structure:

1. Headline: one sentence on whether the goal moved.
2. Three KPIs: this month, last month, three-month baseline, target.
3. Top three posts and why they worked.
4. One thing that did not work and what you learned.
5. Next month's actions: two or three specific changes.

Where to get the data

Each network has its own analytics: Meta Business Suite for Facebook and Instagram, LinkedIn Page and post analytics, YouTube Studio, TikTok analytics, Pinterest analytics and Threads insights. Web outcomes come from your analytics tool, leads from your CRM. The pain is stitching nine sources together every month.

Postprism's analytics bring reach, engagement and growth for the networks you publish to into one place, show top-performing posts, and export to CSV, so the monthly report starts from one file instead of nine logins. Because posts are planned and published from the same content calendar, you can see which pillars and formats drove the numbers. For a wider comparison, see our roundup of the best social media analytics tools.

Common mistakes with social media metrics

1. Tracking everything. Twenty metrics means none gets attention. Fix: three KPIs per goal.
2. Mixing engagement formulas. By-followers this month, by-reach next month. Fix: one formula, written down.
3. Comparing with the wrong benchmark. A by-reach rate against a by-followers benchmark looks amazing and means nothing. Fix: match the formula or use your own baseline.
4. Ignoring definition changes. Meta's views replaced impressions. Fix: note definition changes in your report.
5. Celebrating follower count. It is history, not performance. Fix: report growth rate.
6. No UTM tags. Traffic from social shows up as "direct" or "referral". Fix: tag every link.
7. Inconsistent UTM spelling. Values are case sensitive. Fix: one naming convention.
8. Judging posts after an hour. Metrics settle over days. Fix: compare posts at the same age, for example seven days.
9. Letting data expire. Platform retention is finite. Fix: export monthly.
10. Reports without decisions. Numbers with no actions get ignored. Fix: end every report with next month's changes.
11. Comparing networks on raw views. A view means different things on each. Fix: compare within a network.
12. Averages distorted by one viral post. Fix: report the median alongside the mean.

Frequently Asked Questions

What is the difference between a social media metric and a KPI?

A metric is any number a platform reports. A KPI is a metric you have tied to a business goal, with a target and an action if it misses. Most teams need about three KPIs, not thirty metrics.

What is a good engagement rate on social media?

It depends on the formula. By reach or impressions, Buffer's 2026 medians range from about 2.5% on X to about 6.2% on LinkedIn. By followers, Socialinsider's 2025 figures are far lower, such as 0.48% on Instagram. Compare with your own baseline first.

Which social media metrics matter most?

The ones tied to your goal. For most businesses that means reach, a resonance metric such as saves or shares, and an action metric such as clicks or enquiries. Follower count and likes matter least.

How do I measure social media ROI?

Track outcomes with UTM-tagged links, count every cost including time, and divide net return by cost. Our social media ROI guide walks through the formula with worked examples.

How often should I review social media metrics?

Glance weekly at leading indicators, review KPIs monthly and change strategy quarterly. Compare posts at the same age, such as seven days after publishing.

Why did my impressions change on Facebook and Instagram?

Meta replaced impressions and plays with views from late 2024. Views count repeat appearances, so numbers can be higher than old impression counts. Treat the switch as a definition change in year-on-year reports.

Should I track the same KPIs on every network?

Use the same structure, attention, resonance and action, but pick network-appropriate metrics: saves on Instagram and Pinterest, comments and sends on LinkedIn, watch time on YouTube and TikTok.

Can I see metrics for all my networks in one place?

Yes. Postprism's analytics show reach, engagement and growth for the networks you publish to, highlight top posts, and export to CSV for monthly reporting.