Product demo and problem, then fix: two formats that bring cheaper leads

We have run over 18,920 Meta ads, and when we labeled 2,126 of them by how they look, two formats sat near the cheap end of the list without being rare or hard to make. One shows the thing working. The other names a problem and then shows the fix. Neither needs a studio, a model or a big budget. Both need a script, and most owners skip that part.

This playbook covers what the numbers say, how far you can lean on them, and how to script each format so the first one you make has a fair chance.

What the numbers say

Across the 2,126 labeled ads, the average cost per lead was $26.62. Product demo ads came in at $20.58. Problem, then fix came in at $21.73. Before and after was cheaper still at $18.73, and headline only ads were the most expensive look at $34.35.

Source: Creative Benchmarks, benchmark.sneeze.it

Now the honest part. The product demo group is 69 ads. The problem, then fix group is 115. Those are small groups next to the 640 offer-first banners or the 510 cinematic videos. A handful of strong or weak ads can move a small group's average a long way. So treat these two formats as the place to test first, not as a rule that they will beat everything else in your account.

The two formats also behave differently on a second measure. We call an ad a winner when it spent at least 10 times its own account's median ad and at least $500. That is Meta choosing to put money behind it. Of the product demo ads, 11.1% became winners. Of the problem, then fix ads, 6.7% did. Across all 18,920 ads, 6.2% did.

So product demo was cheap per lead and Meta tended to back it. Problem, then fix was cheap per lead, but Meta backed it at roughly the overall rate. That difference matters for how you judge them, which we come back to below. If you want the longer version of why the two measures disagree, read winner rate vs cost per lead.

Cheap leads and Meta's attention are two different grades. Product demo tended to earn both. Problem, then fix mostly earned the first.

How to script a product demo

A product demo shows the thing a new customer actually gets, happening, in the order they would experience it. Not a montage. Not a list of features over stock footage. One clear sequence.

The script has four beats:

1. The moment before. Two or three seconds of the starting point: the door, the chair, the first screen of the app, the empty room before a session.
2. The thing happening. The core of the ad. Show the service or product doing its job, close enough that a stranger understands what they would be paying for.
3. The result in the frame. End on what it looks like when it is done. No claims, just the picture.
4. The next step. One line on screen and in the caption: what to tap, what it costs or what is free, and where.

Keep it to one person, one location and one sequence. If you find yourself cutting between five different services, you are making a feature list, and feature and benefit lists averaged $32.22 per lead in our data.

A phone is fine. The production label with the cheapest leads in our data was phone-shot UGC at $22.21, a group of 64 ads, so it carries the same small-group warning. The point is that you do not need to wait for a shoot. See the phone-shot UGC playbook for how to film it.

How to script problem, then fix

Problem, then fix opens on something the viewer already feels, then shows how you solve it. The risk with this format is the opening. Done badly, it reads as a lecture or a scare.

The script also has four beats:

1. Name the problem in their words. One short line, the way a customer would say it at your front desk. Specific beats general: a real complaint, not a category.
2. Show it, briefly. A second or two of the problem on screen. Keep it light. You are showing you understand, not making anyone feel bad.
3. Show the fix. This is where a short demo slots in. The best problem, then fix ads are a product demo with a better opening.
4. The next step. Same as before: one action, one offer, one place.

Two rules keep this format honest. First, only name a problem you actually fix, and only show a fix a new customer would actually get. Second, do not promise an outcome you cannot back up. The ad has to match what happens when the lead walks in.

The playbook for this month

Here is how to put both formats into your account without guessing.

1. Write one script of each

   Use the four beats above. One product demo, one problem, then fix, built around the same offer so the only real difference is the format.

2. Film both in one session

   Same person, same location, same phone. Shoot each script two or three times so you have options in the edit.

3. Launch them next to what you already run

   Put them in the campaign that already gets leads, alongside your current ads, so they compete for the same budget and audience.

4. Give them a fair first week

   Most ads never get a real chance: 2 of 3 ads in our data spent under $100 in their whole life. Do not switch either one off in the first few days because Meta spent little on it.

5. Grade each one twice

   After two to four weeks, check whether Meta chose to spend on it, and whether its cost per lead beat your account average. Write both down.

6. Make the next version of whichever earned it

   If the demo earned both grades, make two more demos of other services. If problem, then fix brought cheap leads but little spend, give it its own small budget before deciding it lost.

Step 6 is where the second measure earns its place. In our data, 47 in 100 winners brought leads cheaper than their own account's average and 53 did not. Meta's favorite ad is not always your cheapest. A problem, then fix ad that Meta is slow to back can still be your best source of leads if it gets the budget to show it. Grade every ad twice walks through that check in detail.

A worked example, purely hypothetical. Say you spend $1,500 a month. At the $26.62 average from our labeled ads, that would buy about 56 leads. At the product demo group's $20.58, the same $1,500 would buy about 73. That is arithmetic on our group averages, not a forecast for your account, and with only 69 ads behind the demo figure the real gap in your account could be bigger, smaller or the other way round. The only way to know is to run your own version and read your own numbers.

For the full cost per lead picture across every label, start with what 18,920 Meta ads say about cost per lead.