How to measure short video results: the metrics that matter

Short videos are easy to measure the wrong way. Platforms show big view counts, and it seems the more the better. But a video with 100,000 views that brought no customers may be worth less to a business than one with 3,000 views after which five people booked a call.

Here is how to set up measurement that helps you decide what to make, in which format and where to publish.

Three levels of metrics

It helps to split metrics into three levels.

Level 1 — attention. How much the video interested viewers: views, retention, completion.

Level 2 — reaction. What viewers did after watching: likes, comments, saves, shares, follows.

Level 3 — business. What changed for the business: site visits, leads, calls, sales, mentions.

The higher the level, the more valuable the metric and the harder it is to measure. A good system looks at all three.

Attention metrics

Views

The most obvious and most misleading metric. Platforms count a “view” differently — sometimes a few seconds, sometimes any autoplay in the feed. So views can’t be compared directly across platforms. Use them to compare videos within one platform and to track the trend: is your typical result growing month to month?

Retention and completion

What share of the video people watch on average, and how many reach the end. This is the main indicator of content quality.

• A sharp drop in the first 2–3 seconds — a weak hook.
• A gradual drop towards the middle — too long, too much filler.
• A drop just before the end — viewers got the point and left; consider giving the conclusion earlier.

Viewed vs swiped away

Some platforms show what share of people who saw the video in the feed started watching versus swiped. A direct measure of the first seconds and the cover.

Reaction metrics

For a business, saves, shares and follows matter most — they signal real value, not a passing like. Count them as a share of views, not in absolute terms: 50 comments on 2,000 views is stronger than 100 on 50,000.

Business metrics

The most important and hardest to measure. Ways to do it:

• Tagged links — different links in the profile or description per platform and campaign.
• Promo codes — a dedicated code for short video.
• Asking “how did you hear about us?” on calls, in forms, at the front desk. Simple and very revealing.
• Branded search — growth in searches for your company name often tracks video growth.
• Period comparison — lead volume before and after starting regular videos.

Don’t expect a direct “video → lead” link for every clip. Short videos work cumulatively: people see you five times and visit your site on the sixth.

Comparing platforms fairly

1. Compare medians, not averages. One viral hit skews the mean.
2. Normalise per platform. Judge a video against the typical result on that platform.
3. Focus on reactions, not reach. Save and comment rates compare better than views.
4. Keep your goal in mind. A platform with less reach but more leads may matter more.

From metrics to decisions

Metrics are useful only if they change what you do. Ask questions about video attributes, not single videos:

• Topic — which topics are consistently above the median?
• Format — livestream clips, infographics, article videos, digital presenter — what works for your audience?
• Duration — which range gives the best retention?
• Opening — question, number or statement?
• Publish time — which slots work on each platform?
• Design — do subtitle style or covers make a difference?

Each answer adjusts your content plan: more of what works, less of what doesn’t. Some services draw these conclusions automatically and feed them into the next plan — even then, check them yourself now and then.

How much data you need

Conclusions from 3–4 videos are hypotheses, not patterns. As a guide: 10–15 videos for first cautious observations, 30–50 to compare formats and topics, 2–3 months of regular publishing for a stable picture. Test hypotheses: if videos opening with a question seem to do better, make the next five that way and compare.

A 15-minute weekly review

1. Look at the weekly summary across platforms.
2. Find the 2 best and 2 worst videos by retention and reaction.
3. Rewatch their first 3 seconds — the difference is almost always there.
4. Note 1–2 takeaways and apply them to next week’s plan.
5. Once a month, compare with business metrics: leads, clicks, mentions.

Common mistakes

• Chasing views. Reach without reaction or business results is an empty number.
• Conclusions from one video. Viral success is often random.
• Comparing views across platforms directly. They are counted differently.
• No link to business outcomes. Without tagged links, codes and “how did you hear about us”, you won’t know whether videos pay off.
• Collecting data without decisions. A metric that changes nothing in your plan isn’t needed.

A worked example

A clinic publishes 6 videos a week on three platforms — 24 a month. The review shows: median YouTube views of 1,800 with a best of 14,000; the “Myths” series consistently above the median and the “Team” series below; videos under 30 seconds completed 20% more often than those over 45; the highest save rate on videos with numbers and procedure steps; and seven patients mentioning the videos at appointments.

Decisions for next month: “Myths” twice a week instead of once, “Team” every other week, target 20–35 seconds, more infographics, and a tagged profile link to count clicks properly. None of this requires advanced analytics — only a regular look at the same numbers.

A weekly report template

Keep one table and fill it in every week:

The last row matters most. If there is nothing to put in it, the report was a waste of time.

FAQ

Why did views drop even though nothing changed? Algorithm changes, seasonality and competition for attention all play a role. Look at the trend over 3–4 weeks, not one, and compare with retention: if retention held, the content is probably not the problem.

Should we delete videos that flopped? Usually not. A weak video barely hurts a channel, and deleting it erases data you could learn from.

How often should we check statistics? Weekly for plan decisions and monthly for business conclusions. Checking daily tends to cause anxiety rather than insight.

What is the single most important metric? For most businesses: retention for attention, saves and follows for reaction, and customer mentions for business impact.

Summary

Measure short video at three levels: attention (retention, completion), reaction (saves, shares, follows as a share of views) and business (clicks, leads, mentions). Compare medians within platforms, analyse attributes rather than single videos, and turn one or two takeaways into plan changes every week.