Creative Metrics for Paid Social: Hook Rate, Hold Rate, CTR and More

Creative is the biggest lever in paid social. The same audience, same offer and same budget can produce wildly different results depending on the ad. But "the creative is not working" is not actionable. You need to know where it is failing: Is it not stopping the scroll? Is it not holding attention? Is it getting clicks but not conversions?

Creative metrics answer these questions. This guide covers the key metrics for diagnosing ad performance on Meta and other visual platforms: hook rate, hold rate, click-through rate, cost per mille, cost per landing page view and conversion rate. We explain how to calculate each from Meta columns, how to read them together, and how these ideas translate (or do not) to ChatGPT ads where there is no video.

The creative funnel

Before diving into metrics, understand the funnel a paid social ad travels:

1. Impression. The ad appears on screen.
2. Attention. The viewer notices the ad (stops scrolling).
3. Engagement. The viewer watches, reads or interacts.
4. Click. The viewer clicks through to your site.
5. Conversion. The visitor takes the desired action (purchase, signup, lead).

Different metrics measure different stages. A problem at stage 2 (attention) requires a different fix than a problem at stage 5 (conversion). The goal is to diagnose where the funnel breaks.

Hook rate (thumbstop rate)

What it measures

Hook rate measures how well the first moment of your ad grabs attention. On video ads, it is the percentage of impressions where the viewer watched past a certain point, typically 3 seconds.

How to calculate it

Hook rate = 3-second video views / Impressions

In Meta Ads Manager, use:

• 3-second video views: The number of times your video was watched for at least 3 seconds.
• Impressions: Total times the ad was shown.

Hypothetical example: If your ad had 50,000 impressions and 7,500 three-second video views, your hook rate is 7,500 / 50,000 = 15%.

What it tells you

• High hook rate: The opening frame, text overlay or first few seconds are compelling. People stop scrolling.
• Low hook rate: The ad is getting skipped. The hook is not strong enough.

How to improve it

• Change the first frame. Use a face, movement, bold text or an unexpected visual.
• Lead with the most interesting part. Do not build up to the hook; start with it.
• Test different opening lines or text overlays.
• Match the hook to the audience. What stops a new parent scrolling is different from what stops a fitness enthusiast.

Our guide to Facebook ad hooks covers hook writing in depth.

Hold rate (completion-related metrics)

What it measures

Hold rate measures how well your ad keeps attention after the initial hook. Common variations include:

• ThruPlay rate: Percentage of impressions where the video was watched to 15 seconds or completion, whichever is shorter.
• Average watch time: Mean seconds watched per impression.
• Video completion rate: Percentage of impressions where the video played to the end.

How to calculate it

ThruPlay rate = ThruPlays / Impressions

Or:

Average watch time = Total seconds watched / Impressions

In Meta Ads Manager, use the ThruPlays column and Impressions, or add the "Video average watch time" column for average seconds.

Hypothetical example: If your 30-second video had 50,000 impressions and 3,000 ThruPlays, your ThruPlay rate is 3,000 / 50,000 = 6%. If total watch time was 150,000 seconds, average watch time is 150,000 / 50,000 = 3 seconds.

What it tells you

• High hold rate: The ad holds interest throughout. The content is engaging.
• Low hold rate with high hook rate: People stop to watch but leave quickly. The content after the hook is not compelling.
• Low hold rate with low hook rate: The ad is failing at the start. Fix the hook first.

How to improve it

• Keep the pacing fast. Cut anything that drags.
• Deliver value throughout, not just at the end.
• Match video length to content density. A 60-second ad needs 60 seconds of interesting content.
• Use visual variety: scene changes, text overlays, different angles.

Click-through rate (CTR)

What it measures

CTR measures how often viewers click on your ad. Meta reports two versions:

• CTR (link click-through rate): Clicks to your destination divided by impressions.
• CTR (all): All clicks (including profile clicks, reactions, comments) divided by impressions.

For performance campaigns, CTR (link) is usually more relevant.

How to calculate it

CTR (link) = Link clicks / Impressions

In Meta Ads Manager, use the "Link clicks" and "Impressions" columns.

Hypothetical example: If your ad had 50,000 impressions and 750 link clicks, your CTR is 750 / 50,000 = 1.5%.

What it tells you

• High CTR: The ad compels action. The offer, call to action or curiosity created is strong.
• Low CTR with high hold rate: People watch but do not click. The CTA may be weak, the offer unclear, or the landing page destination unappealing.
• Low CTR with low hold rate: People are not engaged enough to consider clicking. Fix engagement first.

How to improve it

• Strengthen the call to action. Be specific about what happens next.
• Clarify the offer. "Shop now" is vague. "See the 3-step routine" is specific.
• Remove friction. If people do not know where they are going, they do not click.
• Test different CTAs on the same creative.

Cost per mille (CPM)

What it measures

CPM is the cost to reach 1,000 people. It reflects how much you pay for impressions, regardless of engagement.

How to calculate it

CPM = (Total spend / Impressions) x 1,000

Meta reports this directly in Ads Manager.

Hypothetical example: If you spent $500 on 50,000 impressions, your CPM is ($500 / 50,000) x 1,000 = $10.

What it tells you

• High CPM: You are paying more to reach people. This could be due to competition for your audience, seasonality, or lower ad relevance.
• Low CPM: Impressions are cheap. This is good if engagement is also strong; less good if you are reaching low-intent audiences cheaply.

What affects it

• Audience competition: More advertisers targeting the same audience raises CPM.
• Seasonality: Q4, Black Friday and holidays typically have higher CPMs.
• Ad quality: Meta rewards engaging ads with lower CPMs.
• Targeting breadth: Broad audiences often have lower CPMs than narrow ones.

CPM is useful context but not directly actionable for creative optimization. It tells you what you paid, not why.

Cost per landing page view

What it measures

Cost per landing page view is what you pay for each person who clicks and actually loads your landing page.

How to calculate it

Cost per landing page view = Total spend / Landing page views

In Meta, "Landing page views" counts only clicks where the destination page loaded. This is lower than link clicks because some people click but leave before the page loads (slow sites, back-button behavior).

Hypothetical example: If you spent $500 and had 450 landing page views, cost per landing page view is $500 / 450 = $1.11.

What it tells you

• High cost per landing page view (relative to CPC): Many clickers are not reaching your page. Site speed or mobile experience may be issues.
• Normal cost per landing page view: The path from click to page load is working.

How to improve it

• Speed up your landing page. Compress images, reduce scripts, use fast hosting.
• Ensure mobile experience is smooth. Most paid social traffic is mobile.
• Check that links are not broken or redirecting unexpectedly.

Conversion rate

What it measures

Conversion rate is the percentage of landing page visitors who complete your desired action (purchase, signup, lead form).

How to calculate it

Conversion rate = Conversions / Landing page views

In Meta, use "Purchases" (or your conversion event) and "Landing page views."

Hypothetical example: If you had 450 landing page views and 18 purchases, conversion rate is 18 / 450 = 4%.

What it tells you

• High conversion rate: The landing page and offer are working. People who arrive are ready to buy.
• Low conversion rate: Something is breaking between arrival and purchase. This could be message mismatch, price objections, checkout friction, or low intent traffic.

How to improve it

• Match the landing page to the ad. Our guide to landing page message match covers this.
• Simplify the path to purchase. Fewer clicks, fewer fields, fewer distractions.
• Address objections. Social proof, guarantees and clear policies reduce hesitation.
• Test different landing pages for different audiences.

The conversion rate calculator can help with quick math.

Reading metrics together

Individual metrics are useful, but the real insight comes from reading them together. Here is a diagnostic framework:

Start at the top of the funnel and work down. Fixing conversion rate does not help if nobody is clicking. Fixing CTR does not help if nobody is watching past the hook.

How these metrics apply to ChatGPT ads

ChatGPT ads are fundamentally different from Meta video ads. There is no video, no scroll to stop, no "3-second view." The ad appears as a text card below an answer. The user reads the title and body, then decides whether to click.

This changes which metrics matter:

On ChatGPT, the "hook" is the ad title and body together. If CTR is low, the copy is not compelling enough to earn a click. You cannot separate "stop" from "engage" the way you can with video.

What transfers from Meta to ChatGPT

• Message clarity. A confusing value proposition fails on both platforms.
• Offer strength. A weak offer gets skipped on both.
• Relevance. Showing the right message to the right person matters everywhere.

What does not transfer

• Visual spectacle. ChatGPT uses a small static image. There is no room for cinematic video.
• Hook-then-hold structure. ChatGPT ads are scanned in one moment, not watched over time.
• Format variety. Carousels, Reels and Stories do not exist in the same way on ChatGPT.

The strongest Meta ads can inform ChatGPT ads, but the translation is about message, not format. Our guide to reusing Meta winners in ChatGPT ads covers how to extract the promise, proof and offer from a video ad and rewrite it as a compact chat card.

Building a metrics dashboard

Track these metrics weekly to spot trends:

A simple spreadsheet tracking these weekly is enough for most brands. More sophisticated setups use dashboards in Looker Studio, Supermetrics or similar tools.

How SecondWin helps with creative performance

SecondWin studies the longest-running ads in your niche in Meta's Ad Library to find the promise, proof and offer that keep working. We use those insights to write original ChatGPT ads around the buyer questions your customers ask. Because ChatGPT ads are text-focused, the creative translation emphasizes message over format.

If your Meta creative is performing well and you want to extend those winning messages to a second channel, SecondWin handles the translation. Campaigns run in your own OpenAI ad account.

Get a free analysis of your brand to see which buyer questions we would target, or compare plans and pricing. SecondWin is independent and not affiliated with OpenAI.

FAQ

What is a good hook rate?

There is no universal benchmark because it varies by niche, audience and creative style. Track your own baseline and watch for trends. If hook rate drops significantly from your average while spend and targeting stay constant, the creative is fatiguing.

How do I improve hold rate without changing the whole video?

Start with pacing. Cut any sections where energy or information density drops. Add text overlays to maintain visual interest. Move the strongest content earlier. Sometimes trimming a 60-second video to 30 seconds improves hold rate without reshooting.

What is the relationship between CTR and conversion rate?

They measure different stages. High CTR with low conversion rate often means the ad over-promises or misaligns with the landing page. High conversion rate with low CTR means the landing page works but the ad is not compelling enough. Optimize both together.

Do hook rate and hold rate matter for static image ads?

Not directly. Hook rate and hold rate are video metrics. For static ads, focus on CTR (is the image and copy compelling enough to earn a click) and conversion rate (is the landing page converting traffic).

How do ChatGPT ads measure creative performance differently?

ChatGPT ads lack video, so there is no hook rate or hold rate. CTR is the primary engagement metric: does the ad title and body earn a click? Beyond CTR, focus on conversion rate. The diagnostic is simpler: low CTR means the copy needs work; low conversion rate means the landing page or offer needs work.