How to Know If a Creative Is Performing: The Metrics That Matter

You've launched three new creatives. One has 40,000 views and a ton of likes, another has 8,000 views and barely any reactions. Which one is performing better? If you answered "the first one," there's a good chance you're looking at the wrong metrics. Views and likes don't pay the rent: they tell you something caught attention, not that it's bringing in customers.

Judging a creative means reading a handful of numbers, in the right order, to understand not just whether it works but where it breaks. A creative is a small multi-stage machine: it stops the scroll, holds attention, convinces people to click, and drives them to convert. Each stage has its own metric. When the machine jams, the metric tells you at which stage, and from there you know what to fix instead of redoing everything at random.

In this guide we cover the metrics that actually matter (hook rate, hold rate, thumbstop, CTR, CPA), how to read them in sequence, and which "vanity metrics" to ignore when making a decision. It's part of our broader series on performance ad creative, which we recommend for the full picture.

The problem with vanity metrics

A vanity metric is a number that makes you feel good but doesn't change a decision. Likes, follower gains, total views, "reach": they go up, they look positive, but they don't tell you whether the creative is generating sales or leads at a sustainable cost.

The classic case: a "viral" video with 200,000 views and hundreds of comments that brings in 4 conversions. Did it go well? No. It entertained an audience that had no intention of buying. Conversely, a "boring" creative with 5,000 views that brings in 30 purchases at a good cost is a success. The big number is seductive, but the decision rests on the small number that actually counts.

The practical distinction is simple. Ask yourself: if this number doubles, would I change anything in the campaign? If the answer is no, it's a vanity metric. Views double but CPA stays high? You haven't learned anything useful. CTR doubles at the same spend? That tells you the audience is reacting more to the message, and that's actionable.

Careful though: "vanity" doesn't mean "always useless." Reach matters when you're measuring audience saturation or incrementality. Comments matter as a qualitative signal (what people actually say about your offer). But when it's time to decide whether to scale, pause, or iterate on a creative, those aren't the numbers to lean on.

The metrics that matter, stage by stage

You can break a video creative down into four moments, each with its own question and its own metric. Reading them in this order is the heart of the method.

1. Hook rate (thumbstop): does the creative stop the scroll?

A creative's first job is to interrupt the scroll. On feed and Reels people scroll at breakneck speed: if nothing happens in the first couple of seconds, you're invisible. The metric that measures this is called hook rate (or thumbstop ratio).

Practical definition: the percentage of people who, after the creative appears, watch past the first 3 seconds. In Meta data terms, it's the ratio between 3-second video plays and impressions.

• Formula: 3-second video plays ÷ impressions × 100
• What it tells you: how strong the opening hook is (first frame, first second, motion, opening text)
• Reference range: below 20-25% the hook is weak. A good hook rate on Reels sits roughly above 30%; the best creatives clear 40-45%. These are orders of magnitude, not hard thresholds — they depend on industry, format and placement.

If the hook rate is low, the problem is in the first 3 seconds, full stop. You don't need to redo the whole video: change the opening. A stronger opening frame, a direct question, motion in the first second, a visual contrast. We go deeper into concrete opening examples in our articles on Reels that stop the scroll and on Meta ad creatives that stop the scroll.

2. Hold rate: does the creative retain attention?

Stopping the scroll isn't enough: if people bail after 4 seconds, the message never lands. Hold rate measures retention, i.e. how long people stick around after the hook.

There are two common ways to read it. One is the average percentage watched (on average, how much of the video gets watched). The other, more precise, is the percentage that reaches key checkpoints: 25%, 50%, 75%, 100% of the video. The retention curve (exactly where people drop off) is the most useful information you have.

• What it tells you: whether the body of the video holds up after the opening, and at what point people get bored or lost
• The gold signal: the "cliff," the point on the curve where retention crashes. If 60% drop off at second 6, you have a precise problem at second 6 (slow pacing, a weak transition, an unclear message).
• Reference range: for short videos (under 15 seconds), aiming for a 40-60% average completion rate is a good target. Longer videos naturally drop lower.

High hook but low hold is one of the most common patterns: the hook "grabs" but then the video promises something it doesn't deliver, or the pacing sags. The fix is surgical: look at where the curve crashes and fix that piece (trim the dead time, add captions for people watching on mute, front-load the benefit instead of making people wait for it).

3. CTR: does the creative drive action?

Do the people who watched click? CTR (Click-Through Rate) is the percentage of people who click, out of everyone who was shown the ad. Here two clarifications matter — they're the difference between reading it right and drawing the wrong conclusions.

First: there are two CTRs. CTR (all) counts every kind of click (including likes, expansions, profile clicks). Link CTR only counts clicks that lead to the destination (site, landing page). To judge purchase intent, look at link CTR: it's the one that reflects real interest in the offer, not simple curiosity.

• Formula: link clicks ÷ impressions × 100
• What it tells you: how compelling the offer and the call-to-action are for people who saw it
• Reference range: a link CTR between 1% and 2% is often considered healthy on Meta; below 0.5% is a warning sign. Again, it depends heavily on industry and audience temperature. We cover this in detail in our dedicated article on what CTR means on Facebook Ads.

A low CTR alongside good hook and hold numbers almost always points to a problem with the offer or the call-to-action, not attention: people watch the whole thing but aren't convinced to click. A high CTR but a poor CPA (more on that shortly) instead points to a mismatch between the creative and the landing page: the promise made in the creative isn't backed up on the destination page.

4. CPA: does the creative bring in customers at a sustainable cost?

Here's the metric that decides everything: CPA (Cost Per Acquisition), the cost of getting one conversion (purchase, lead, booking). All the previous metrics are diagnostic; CPA is the verdict. A creative with great hook, hold and CTR but a CPA that's off target is still a creative that needs revisiting.

• Formula: total spend ÷ number of conversions
• What it tells you: how much a customer (or lead) generated by that creative costs you
• The yardstick: CPA always has to be weighed against your margin and customer value. A €25 CPA is great if a customer is worth €300 over time, terrible if the product only nets €20.

CPA should never be read in isolation. It has to be weighed against customer lifetime value (LTV) and the threshold beyond which acquisition stops being profitable. To learn how to set up these calculations, you'll find useful reading in how to calculate customer lifetime value and in unit economics KPIs (CAC, CPL, LTV). The right creative is the one that keeps your CPA under that threshold reliably, repeatedly.

How to read them together: the diagnostic chain

The real value isn't in any single metric, it's in the sequence. Reading them in order (hook, then hold, then CTR, then CPA) lets you spot at a glance which stage a creative is losing at. Here's a quick-reference table.

Stage	Metric	Question	If it's low, what to fix
1. Attention	Hook rate / thumbstop	Does it stop the scroll?	The first 3 seconds: opening frame, opening, motion
2. Retention	Hold rate / completion	Does it hold attention?	The body of the video: pacing, the point where the curve crashes
3. Interest	Link CTR	Does it drive clicks?	Offer and call-to-action: the message isn't convincing
4. Conversion	CPA	Does it bring in customers at the right cost?	Creative-to-landing consistency, targeting, the actual offer

A few recurring combinations that will save you hours of blind testing:

• Low hook, everything else unreachable: nobody makes it past 3 seconds. Don't touch the rest, rewrite only the opening.
• High hook, low hold: the hook works but the video doesn't deliver. Fix the pacing at the "cliff" point.
• Good hook and hold, low CTR: they watch it all but don't click. Offer or CTA problem, not a visual creative problem.
• High CTR, poor CPA: they click but don't convert. Look at the mismatch between the creative and the landing page (unmet promise, different price, slow page).

This is the heart of the reasoning: a creative doesn't "fail" as a whole, it breaks at one precise stage, and each metric tells you which one. That's how you stop redoing entire videos and start making targeted fixes. If an entire creative is weak across the board, the problem is probably further upstream (concept or offer): you'll find an overview of the costliest mistakes in our article on the mistakes that kill creative performance.