Social media metrics that matter (and the ones that do not)

Most platforms show you dozens of social media metrics, and the biggest numbers are rarely the most useful. A rising follower count feels like progress, yet it says little about whether anyone understood your message, trusted it or acted on it.

Follower numbers also depend on choices you make. M+R's Benchmarks study notes that nonprofit follower counts are heavily dependent on the strategic choices organisations make and the content they prioritise on each platform.¹ Platform features can matter less than they seem: in the 2025 edition, revenue from Facebook Fundraisers made up 0.2% of participants' online revenue in 2024.² Measure what links to your objectives, not what a dashboard puts at the top.

The UK Government Communication Service's OASIS model asks teams to monitor outputs, outtakes and outcomes throughout a campaign and to evaluate them once it is complete.³ It also says objectives should be measurable and focused on outcomes, not outputs.³ That gives you a simple way to sort social media KPIs:

• Outputs: what you published and who saw it, such as posts, reach and impressions.
• Outtakes: how people responded, such as shares, saves, comments, link clicks and video completion.
• Outcomes: what changed, such as sign-ups, service use, event attendance, petition signatures or questions from journalists and partners.

Pick a small set of KPIs for each objective, and no more than one or two headline figures for senior managers. The CDC's toolkit for health communicators suggests measures such as traffic driven back to your website, influence and reach, and user interactions and engagement, while noting that health impact is harder to measure.⁴ That honesty is useful: social media rarely proves impact on its own, but it can show whether people are paying attention and taking the next step.

In our own practice, these are the measures that most often earn a place in a monthly report. Not every organisation will need all of them, and the right mix depends on whether your objective is awareness, engagement or action:

• Reach: how many accounts saw your content.
• Engagement rate: interactions as a share of reach (see below).
• Shares and saves: signs that content is useful enough to pass on or keep.
• Meaningful comments and messages: questions, requests and stories, read rather than counted.
• Link clicks and website visits from social media.
• Conversions: the action your objective asks for.

Engagement rate is one of the most quoted social media metrics, and one of the easiest to misread. It can be calculated in several ways: interactions divided by followers, by reach or by impressions, with different interactions counted in each. Unless you know which formula sits behind a figure, you cannot compare it with another.

Choose one formula, write it in your measurement plan and use it every month. We usually recommend total interactions on a post divided by that post's reach, because it reflects the people who actually saw it. Compare posts of the same format on the same platform, and look at the median as well as the average, so that one viral post does not hide a weak month.

A view can last a second. YouTube defines watch time as the amount of time viewers have watched your video, and average view duration as the average minutes watched among those who stayed to watch.⁵ Its audience retention report shows how well different moments of a video held attention, which tells an editor exactly where viewers left.⁵

Click-through rate is useful in context. YouTube says half of all channels and videos have an impressions click-through rate between 2% and 10%, and that a high click-through rate combined with low average view duration may indicate clickbait.⁶ It advises comparing rates between videos over the long term.⁶ The lesson applies to every platform: a thumbnail that attracts clicks is only a success if people stay.

Outcomes usually happen off the platform, on your website, at a health centre or at an event. M+R's 2026 study found that mobile users made up 52% of website visits to participating nonprofits, and that 1.6% of visitors from all sources made a donation.⁷ Most visitors do not complete the action you want, so the page people land on matters as much as the post that sent them.

Add campaign tags to the links you share so your web analytics can show which posts and platforms brought people in. Where the outcome is offline, agree a simple question with programme teams, such as asking new participants how they heard about a service. A small, honest sample is more useful than a large number nobody can explain.

Reports work best when they lead to decisions. In our own practice, we keep a monthly social media report to one page: progress against each objective, three or four KPIs with the previous month for comparison, the best and weakest post with a sentence on why, what people asked or said, and one change for next month.

Every quarter, step back and evaluate against your outcomes, as OASIS recommends at the end of a campaign.³ Share the findings with programme colleagues, because they often know why a post worked or why a service saw more visitors that month. If a metric has not changed a decision in six months, stop reporting it.

• Each KPI linked to a written objective
• Outputs, outtakes and outcomes reported separately
• Engagement rate formula written down and used consistently
• Watch time and average view duration tracked for video
• Campaign tags on every link you share
• Comments and messages read for meaning, not just counted
• One-page monthly report with one decision
• Quarterly review against outcomes