Marketing

Hook Rate and Hold Rate: Video Ad Metrics That Predict ROAS

Hook rate and hold rate predict ROAS days before conversions land. Get the formulas, cold traffic benchmarks, and a fix for failing hooks, bodies, and CTAs.

Hunter Ezra

Most operators judge video ads on cost per booked call, and that is the right final scoreboard. The problem is timing. On a high-ticket offer, one ad can spend $1,500 and produce one or two sales, which is not enough data to tell a good creative from a lucky one. Hook rate and hold rate fix that. They tell you within 48 hours whether an ad is earning attention, and attention is the first domino behind every conversion that follows.

Ignoring them is expensive. I have seen accounts let a weak ad burn three times its target CPA because "it needs more time," and I have seen teams kill an ad with a great body because the first three seconds were wrong. Both mistakes come from the same blind spot: treating the ad as one unit instead of a funnel with stages.

What Hook Rate and Hold Rate Actually Measure

Every video ad is a small funnel. Each stage maps to a specific part of the creative, and each has its own metric.

  • Hook rate = 3-second video plays divided by impressions. It measures the first frame, the first spoken line, and the on-screen text. Also called thumb-stop ratio.
  • Hold rate = ThruPlays divided by 3-second video plays. A ThruPlay is a view of 15 seconds, or the full video if shorter. It measures whether the body earns attention after the hook.
  • Outbound CTR = outbound clicks divided by impressions. It measures the call to action and the offer.
  • Landing page conversion rate = leads divided by landing page views. It measures the page, not the ad.

When an ad underperforms, it is almost always one stage failing, not the whole ad. Know which stage, and you can fix it without throwing away the parts that work.

How to Calculate Hook Rate and Hold Rate in Ads Manager

Meta does not show either metric by default, so build them once and save the view.

  1. In Ads Manager, open Columns, then Customize Columns, then Create Custom Metric.
  2. Name the first one Hook Rate. Formula: 3-second video plays divided by Impressions. Format: percentage.
  3. Name the second one Hold Rate. Formula: ThruPlays divided by 3-second video plays. Format: percentage.
  4. Add CPM, outbound CTR, cost per lead, and booked calls, then save the preset as Creative Diagnostics.

Read these at the ad level, never the campaign level, where your best and worst creative average into a number that describes neither. For more resolution, add Video plays at 25%, 50%, and 75%. A sharp cliff between two points shows exactly where in the script people leave.

Hook Rate Benchmarks for Cold Traffic

These are rules of thumb from running cold traffic for service businesses and high-ticket offers on Meta: Feed and Reels mixed, broad audiences, videos 30 seconds and longer. Retargeting runs higher, so judge warm traffic separately.

Hook rate

  • Under 20%: weak. The opening is not stopping the scroll.
  • 20 to 29%: workable if hold rate and CTR are strong.
  • 30 to 39%: strong.
  • 40% and up: elite, but check lead quality. Very high hook rates sometimes come from curiosity bait.

Hold rate

  • Under 20%: the body loses people right after the hook.
  • 20 to 34%: average.
  • 35% and up: strong.

Outbound CTR

  • Under 0.6%: the CTA or the offer is not landing.
  • 0.8 to 1.2%: healthy for cold high-ticket traffic.
  • 1.5% and up: strong.

Your own account is the real benchmark: use the median hook rate, hold rate, and CTR of the last 90 days of video ads that hit your cost per booked call target.

How Hook Rate Predicts CPA and ROAS Before Conversions Land

Cost per lead is the product of the stages: CPL = CPM divided by (1,000 x CTR x landing page conversion rate). Hook rate and hold rate sit upstream of CTR. People who never watch past second three rarely click, so a real lift in hook rate usually shows up as a lift in CTR, which divides straight into your cost per lead.

In my experience, ads with stronger hook and hold rates also buy cheaper CPMs on the same audience, because Meta's auction rewards engagement. The gains compound.

The real advantage is speed. Hook rate stabilizes after roughly 3,000 impressions and hold rate after a few hundred 3-second views. CPA on a high-ticket offer needs 20 or more conversions to mean anything, which at $200 to $400 per booked call is $4,000 to $8,000 per ad.

Here is the 48-hour read I use on every new video ad:

  1. Hook rate under 20% after 3,000 impressions: kill it, or rehook it if the hold rate is strong.
  2. Hook rate 25% or higher but hold rate under 20%: re-edit the body before spending more.
  3. Hook and hold healthy, CTR under 0.6% after 10,000 impressions: rewrite the CTA or sharpen the offer.
  4. All three healthy: let it run to 1.5 to 2 times your target cost per booked call before judging it on bookings.
  5. Spend reaches 2 times target cost per booked call with zero bookings: kill it, regardless of attention metrics.

Rule 5 matters. Attention metrics predict, they do not override. If you have not set your target yet, work it out from your break-even cost per call and minimum ROAS math before you launch.

Diagnosing the Problem: Hook, Body, or CTA

Once an ad has enough data, its metrics fall into one of four patterns, each with a different fix.

Low hook rate, healthy hold rate: a hook problem

The people who stay like what they see, but too few stay. The body is an asset, so do not touch it. Replace the first three seconds: a new opening visual, first line, and text overlay. This is the cheapest fix in advertising because the expensive part of the ad is already proven.

High hook rate, low hold rate: a body problem

The hook writes a check the body does not cash. Usually the hook promises a specific payoff and the next ten seconds are a slow intro, a logo, or "so, a little about me." Cut the first five to ten seconds of the body, deliver the promised payoff by second 8, and change the visual every three to five seconds. If longer ads leak here consistently, rebuild them on a proven VSL script structure for high-ticket offers.

Strong hook and hold, weak CTR: a CTA or offer problem

People watch and do not act. Is there a CTA before the 30 second mark, or only at the end after most viewers have gone? Is the next step specific, such as "book a 20 minute call to see if this fits your business," instead of "learn more"? Could a viewer repeat the offer back to you? Fix the CTA first because it is a single edit.

Strong across the board, expensive leads: a page or quality problem

If hook, hold, and CTR all clear the bar and cost per lead is still high, the ad is doing its job and the page is not. Compare landing page conversion against your account average and work through landing page optimization for call funnels. If leads are cheap but calls do not qualify, the hook is attracting the wrong person.

A Worked Example: Two Ads, $1,500 Each

Take a service offer with $4,000 cash collected per sale, a 65% show rate, and a 20% close rate on shown calls. Each booked call is worth $4,000 x 0.65 x 0.20 = $520 in expected cash. Two ads each spend $1,500 at a $50 CPM, buying 30,000 impressions apiece. Both send traffic to the same page with a 12% opt-in rate, and 30% of leads book a call.

Ad A: 20% hook rate gives 6,000 3-second views. 22% hold rate gives 1,320 ThruPlays. 0.6% CTR gives 180 clicks, about 22 leads, and 6.5 booked calls. That is roughly $69 per lead and $231 per booked call, a predicted ROAS of about 2.2x.

Ad B: 35% hook rate gives 10,500 3-second views. 38% hold rate gives 3,990 ThruPlays. 1.2% CTR gives 360 clicks, about 43 leads, and 13 booked calls. That is roughly $35 per lead and $116 per booked call, a predicted ROAS of about 4.5x.

Now look at the sales data. Ad A expects 0.8 sales and Ad B expects 1.7, so each shows zero, one, or two closed deals, which tells you nothing. But the 15 point hook rate gap was visible on day one, two weeks before closed deals could confirm it.

The example also shows where to work on Ad A: five new hooks plus one tighter body edit. If hook and hold both reach 30%, CTR usually rises with them. At a 0.9% CTR, the same math puts Ad A at about $154 per booked call, a predicted ROAS of 3.4x.

Iterating Hooks on Winners: The Rehook Protocol

A proven body with a new hook is far more likely to win than a brand new concept, and it extends the life of a winner when ad fatigue starts raising your costs.

  1. Pick the source ad. A winner that has held your cost per booked call target for 7 or more days, or a near-winner with strong hold and CTR but a weak hook.
  2. Freeze the body. Everything from second 3 or 4 onward stays identical, so the hook is the only variable.
  3. Write 5 to 8 new hooks across categories. Call-out: "If you run an agency doing $50k to $150k a month and clients keep churning at month four, watch this." Contrarian: "Stop hiring more closers. Your close rate is not the problem." Pattern interrupt: open mid-action, with an unexpected prop, or in an unexpected location.
  4. Change the layer that matters most. The opening visual and first spoken line move hook rate more than text overlays.
  5. Test with forced distribution. One rehook per ad set with equal budgets, using the forced distribution creative testing method, so every hook gets a fair read.
  6. Graduate on two conditions. The new hook beats the original's hook rate by 5 or more points, and CTR and qualified booking rate hold steady or improve.

That second condition is the guardrail. Hook rate is not the goal, qualified attention is. A tight call-out hook at 24% that filters out the wrong buyers will often beat a 40% curiosity hook on cost per qualified call.

Common Mistakes With Video Ad Metrics

  • Comparing across contexts. Reels, Feed, and Stories autoplay differently, and warm audiences hook higher than cold. Compare within the same placement mix and audience temperature.
  • Optimizing for the metric. Hook rate is a diagnostic, not a campaign objective. Keep optimization on leads or a deeper event, the same logic behind excluding cheap views in the Thunder Dome testing strategy.
  • Reading hold rate on short videos. Under 15 seconds, a ThruPlay is a completion, so hold rate inflates. Benchmark short and long videos separately.
  • Changing hook and body together. If both change, you cannot know which one moved the result.
  • Reading too early. Under 1,000 impressions, hook rate swings 10 points on noise alone.

Frequently Asked Questions

What is a good hook rate for Facebook ads?

For cold traffic on service and high-ticket offers, 20 to 29% is workable, 30% and up is strong, and under 20% needs a new opening. Warm audiences run higher, so judge them separately. The most reliable bar is the median hook rate of your own winning ads.

How do you calculate hold rate on Meta?

Divide ThruPlays by 3-second video plays, built as a custom metric in Ads Manager. Some buyers use 15-second views or video plays at 50% instead, which works as long as the whole team uses one definition.

Does a higher hook rate mean a lower CPA?

Usually, but not always. A higher hook rate tends to raise CTR and lower CPM, both of which reduce cost per lead. If the hook attracts the wrong audience, cost per qualified call rises, so always confirm with booked call data.

Next Steps: Read Your Ads Before They Convert

Build the Hook Rate and Hold Rate custom metrics today. Pull every video ad from the last 90 days, set your benchmarks from the winners, and label each live ad as a hook, body, CTA, or page problem.

Then take your two best ads, freeze their bodies, and write five new hooks for each. Run them through a direct response creative audit, test with forced distribution, and graduate only the hooks that lift hook rate without hurting lead quality. Do that every two weeks and you will stop waiting on closed deals to tell you what the first 48 hours already did.