How to Read Video Ad Metrics on Meta: Stop Rate, Hold Rate, Outbound CTR and Conversion Rate

Short answer: Read every video ad in four stages: attention (stop rate, 3 second video views divided by impressions), interest (hold rate, how far viewers get into the video), desire (outbound CTR, using the most conservative version such as unique outbound CTR) and action (conversion rate). Find the one stage that is breaking and fix that stage only. High stop with low hold means shorten or reorder; low stop with a strong middle means new hooks; strong stop and hold with weak clicks means rework the offer or the middle; strong everything but conversion means look at the landing page, price or offer.

The four stages of a video ad: attention, interest, want and action

The four stages: attention, interest, want and action.

Most teams look at a video ad and see one number: cost per purchase. That number tells you whether the ad works. It does not tell you why, and without the why, the next ad is a guess. Jake's four stage method turns one verdict into a diagnosis, and it works for every video format in our creative playbook. The brands in the examples are made up.

What is the four stage method?

The four stage method is a way of reading a video ad as a sequence of four jobs (stop, hold, click, convert), each with its own metric, so you can see exactly where viewers drop out and fix that point.

Stage Job Metric
Attention Stop the scroll Stop rate: 3 second video views divided by impressions
Interest Keep them watching Hold rate: how far viewers get into the video
Desire Make them want to click Outbound CTR, the most conservative version (such as unique outbound CTR)
Action Turn the visit into a sale or lead Conversion rate

Outbound CTR is the share of people who clicked through to your site, not every click on the ad. We use the unique version because it counts people, not repeat clicks, so it is harder to flatter.

Why does it matter for paid social?

Because you rarely hit a home run with your first ad, and a near miss holds enough data to fix it, if you can read it.

The problem: an ad with a poor cost per purchase gets turned off, and the team starts a new concept from scratch. But that ad might have had the best hook in the account and one weak middle section. Killing it throws away the part that worked.

The solution: split the ad into stages. If one stage is broken and the others are strong, you fix one stage. Improving an ad that is close is easier than starting over. That is why iteration is the core skill, and the four stages are how you know what to iterate.

What is the step by step process?

  1. Label every ad with its length. Put the length in seconds in the ad name or your tracking sheet. Then you can turn "viewers dropped off around halfway" into "viewers dropped off at this exact line."
  2. Pull the four metrics for each ad. Set up a column view in Ads Manager with impressions, 3 second video views, video watch data, unique outbound CTR and your conversion rate.
  3. Compare against your own best ads. "Strong" and "weak" mean compared with your current winners in the same account, not an industry average.
  4. Find the first stage that breaks. Work in order. A broken stop rate makes every later stage unreliable because so few people saw the rest.
  5. Apply the matching fix: * High stop rate, low hold: shorten the video, move the selling points earlier, or reorder blocks. * Low stop rate, strong hold or conversion: the middle works; test new hooks. * Strong stop and hold, weak clicks: rework the offer or swap the middle blocks. * Strong everything except conversion: look at the landing page, price or offer.
  6. Change one thing, then test again. If you change the hook and the middle together, you will not know which fix worked.

What does a worked example look like?

Nightfold sells a weighted sleep mask ($38) with gentle even pressure and total blackout that blocks the light leaks at the nose. It launches three new video ads. Compared with its current best ad:

Ad A has a strong stop rate, but viewers fall away early, well before the product appears. Looking at the length label, the product does not show up until late in the video. Fix: move the nose gap demo to the front and cut the long intro. Same footage, new order.

Ad B has a weak stop rate, but the few people who watch stay to the end, and its conversion rate is the best of the three. Fix: the middle and close work. Keep them, and test new hooks on top. One option: open on the light leak itself, "This tiny gap by your nose is why you wake up at 5."

Ad C stops people and holds them, but almost nobody clicks. The story is enjoyable and the viewer never gets a reason to act. Fix: rework the offer line or swap in a middle block that shows the blackout test, so the desire is built before the end.

A fourth case to know: an ad that is strong on stop, hold and clicks but weak on conversion is usually not an ad problem at all. Check the landing page, the price and the offer. If the page does not continue the promise the ad made, buyers leave.

Notice that none of these fixes needs a new concept. Each one keeps what worked. This is much easier when footage is recorded as modular content.

What tools do you need on a small budget?

  • Meta Ads Manager. Every metric here is available in a saved column view. Build it once and reuse it.
  • A naming rule that includes ad length. Free, and the most useful habit on this list.
  • A simple sheet with one row per ad and a column for each stage, so you can see at a glance which stage breaks most often.
  • A video editor to make the fix. Hooks and reorders are quick jobs when footage is modular.

How do you measure whether the method is working?

The method is working when your team's fixes start to match the stage that broke, and when near misses turn into winners instead of being thrown away.

Track two things over time:

  • Which stage breaks most often in your account. If most failures are at the stop, your hook writing needs work. If most are at conversion, the problem lives after the click. Our hook library helps with the first case.
  • How many winners came from a fix rather than a new concept. If iterations are producing winners, keep feeding them. This connects to your overall creative hit rate.

What mistakes do teams make?

  • Judging only on cost per purchase. It tells you whether, never why.
  • Using the most generous click metric. All clicks and link clicks flatter weak ads. Use unique outbound CTR.
  • Comparing against industry averages. Compare against your own best ads.
  • Fixing the wrong stage. A new hook will not save an ad whose problem is the landing page.
  • Skipping the length label. Without it, a hold rate drop is a percentage you cannot place in the video.
  • Changing everything at once. One fix per test, or you learn nothing.

What to change at each stage

What to iterate based on stop rate, hold rate, CTR and conversion rate
  • High stop rate and low hold rate: look at seconds 3 to 10.
  • Low hook rate with a strong hold or conversion rate: new hook.
  • Strong stop and hold rate with a low CTR: work on text overlays, urgency and the offer.
  • High CTR with low conversion: check the landing page and offer.

Every formula is in creative metrics formulas, and 20 ways to make the change are in the creative iterations checklist.

Key takeaways

  • Every video ad has four jobs: stop, hold, click and convert.
  • Stop rate is 3 second video views divided by impressions; use unique outbound CTR for clicks.
  • Find the first stage that breaks and fix only that stage.
  • High stop and low hold: shorten or reorder. Low stop and strong middle: new hooks. Weak clicks: offer or middle. Weak conversion: landing page, price or offer.
  • Label every ad with its length so drop offs become exact moments.
  • A near miss is data. Fix it before you replace it.

FAQ

What is a good stop rate for Facebook video ads? The useful comparison is your own account. Measure stop rate as 3 second video views divided by impressions and compare each new ad with your current best ads. A high stop rate does not guarantee sales, but if nobody stops, nobody sees the rest of the ad.

What is hold rate? Hold rate is how far viewers get into your video after they stop. Paired with the ad's length, it shows the exact moment people leave, which tells you which section to cut, move or replace.

Why use unique outbound CTR instead of link CTR? Because it is the most conservative version. It counts people who actually left for your site, once each, so it is harder for a weak ad to look good.

What if an ad is strong on every stage except conversion? Then the ad is probably doing its job and the problem is after the click. Check the landing page, the price and the offer before you change the creative. For deciding whether to keep an ad at all, see how a creative testing ad gets judged, and for building the next version, how to iterate a winning video ad.

Want a second set of eyes on how your video ads are performing at each stage, and who they are reaching? We offer a free Meta audit: a full account audit plus net new reach and cumulative reach reports. You can request yours on our contact page.

Check what you learned

Tap an answer to see if you got it right.

1. What are the four stages of a video ad?

Attention, interest, desire, action.

2. High stop rate but low hold rate. What do you do?

The hook works. The middle loses people.

3. What's the rule for fixing?

A near miss is data. Fix it before you replace it.

Ready for the full test? Take the Creative exam and get certified. Want this run on your account? Get my free audit.

About the author

Jake Mehani, founder of Creative Analytics

Jake has spent 15 years running Meta and paid social ads for consumer brands. Meta published his RBX Active Advantage+ case study, he taught the DMAnc Meta and TikTok ads masterclass, and he spoke at Motion's 2024 Creative Strategy Summit. His senior team runs media buying, creative and measurement like an in house team.

Jake on LinkedIn  |  Case studies  |  Free Meta audit

Previous
Previous

What Are Edutainment Ads? How to Use Them on Facebook and Instagram

Next
Next

How to Iterate a Winning Video Ad: Footage and Voiceover as Two Assets