Ad Creatives That Convert: Matan Guy's Framework for Hooks, UGC and Testing

By Matan Guy · November 11, 2026 · 7 min read

When people ask me, Matan Guy, why their ads aren't working, they usually expect an answer about targeting or bidding. Most of the time the answer is the creative. On Facebook and Instagram today, the ad itself — the image, the video, the first line of text — does more of the work than any setting you can tweak. This is the framework I use to plan, make and test ad creatives without guessing.

Why ad creatives are the main lever

Meta's delivery system has become very good at finding people once it knows what to look for. What it can't do is make a boring ad interesting. Your creative decides three things:

Whether people stop scrolling. If they don't stop, nothing else matters.
Who stops. A message aimed at a specific problem attracts people who have that problem. In that sense, the creative is your targeting.
Whether they act. A clear promise and a clear next step turn attention into clicks.

If you've already set up your campaign structure and tracking (I cover that in my Facebook ads guide for small businesses), creative testing is where most of your improvement will come from.

Hook types for ad creatives

The hook is the first one to three seconds of a video, or the first thing the eye lands on in an image. I keep a list of hook types and rotate through them when I brief new ads.

Problem hook — name the frustration directly. "Tired of your plants dying every summer?"
Curiosity hook — open a loop the viewer wants closed. "The one thing I changed in my packaging…"
Direct callout — speak to a specific person. "If you run a small bakery, watch this."
Visual hook — something unexpected on screen in the first frame: a demonstration, a before-and-after, an unusual angle.
Contrarian hook — challenge a common belief. "You don't need more followers to sell."
Demonstration hook — show the product doing the thing, immediately, with no intro.

A strong hook is specific and honest. It promises what the rest of the ad actually delivers. Clickbait hooks may get attention, but they bring the wrong people and hurt trust.

UGC-style video: why it works

UGC means user-generated content: video that looks like it was filmed by a regular person on their phone rather than produced in a studio. This style often works well in feeds because it looks like the content people already scroll through.

What makes a good UGC-style ad:

Filmed vertically on a phone, in natural light.
A real person talking to the camera like they're telling a friend.
The product in use, in a real setting.
Captions on screen, because many people watch without sound.

A practical point: if you work with creators, use real people who actually tried the product, and be transparent. Don't script fake reviews or present paid creators as random customers. Follow Meta's advertising policies and the disclosure rules where you sell.

Static ads still matter

Video gets most of the attention, but static images are fast to make, cheap to test and often perform well. I use them to test messages quickly before investing in video.

Static formats worth testing:

A clean product photo with one bold line of text.
A simple comparison: "this vs that."
A list of three benefits.
A screenshot-style image, such as a note or a question someone might ask.
A real customer review, used with permission and quoted accurately.

Keep text on the image short. If you need a paragraph to explain the idea, it isn't clear enough yet.

A simple video ad structure

Most of my video ads follow the same five-part structure. It's not a magic formula — it's a checklist that keeps me from skipping steps.

Hook — stop the scroll in the first seconds.
Problem — show you understand what the viewer is dealing with.
Solution — introduce the product as the way through.
Proof — a demonstration, real reviews, how it's made, or how it works. Only use proof that's real.
CTA (call to action) — tell people exactly what to do next: "Shop now," "Book a call," "See the full collection."

Short ads can compress this into a few seconds. Longer ads can spend more time on proof. The order stays the same.

Aspect ratios and placements

Different placements show ads in different shapes. Making the wrong size means your ad gets cropped or looks out of place.

Placement	Common ratio	Notes
Stories and Reels	9:16 (vertical, full screen)	Keep text and faces away from the top and bottom edges
Feed	4:5 or 1:1	4:5 takes up more screen on mobile
Right column and some others	1:1 or 1.91:1	Less common for most small businesses

My default is to create a 9:16 version and a 4:5 version of every ad. Ads Manager lets you assign different versions to different placements. Check Meta's current specs before you produce, because recommended sizes and safe zones get updated.

One-variable testing for ad creatives

The biggest testing mistake I see is changing everything at once. If a new ad has a different hook, a different person, a different offer and different music, and it wins, you've learned almost nothing about why.

How I test instead:

Change one thing at a time. Same video body, three different hooks. Or same hook, two different offers.
Start with the biggest variables. Test the concept and angle first (what you say), then the format (how you show it), then small details like colors or button text.
Give each test a fair chance. Similar budget, similar time, same audience.
Judge on the metric that matters. Hook rate and CTR show attention; cost per result shows whether it actually works for the business.
Write down what you learned. A test isn't finished until you've noted the result and the reason you think it happened.

Creative fatigue and when to refresh

Every ad wears out eventually. The same people see it again and again, and they stop responding. Signs of creative fatigue:

Frequency keeps rising.
CTR slowly declines.
Cost per result climbs while nothing else in the campaign changed.

When this happens, you don't always need a completely new idea. Often you can refresh a winning ad with a new hook, a new person presenting it, a different opening shot or a new format. Keep the core message that worked and change the packaging. I plan to add new creatives to active campaigns regularly, before fatigue shows up, rather than waiting for results to drop.

Build a creative library

Over time, the most valuable asset in your ad account isn't any single ad. It's what you've learned. I keep a simple creative library — a spreadsheet or folder works fine — with:

Every ad you've launched, with a thumbnail or link.
The hook type, angle, format and length.
The main metrics and a short verdict: winner, loser, inconclusive.
A one-line lesson.
Ideas for future variations.

This turns random experiments into a system. When you need new ads, you start from patterns that have worked instead of a blank page. It's the same principle I describe in daily habits that compound: small, consistent records add up.

FAQ

How many ad creatives should I test at once?

For a small budget, three to five ads per ad set is a reasonable starting point. Too many ads split the budget so much that none of them gets enough data to judge.

Do I need professional video equipment?

No. Most UGC-style ads are filmed on a phone. Good lighting, clear audio and on-screen captions matter far more than an expensive camera.

How often should I refresh my ad creatives?

There's no fixed schedule. Watch frequency, CTR and cost per result. When they start moving in the wrong direction, refresh. Many advertisers find it easier to add new variations steadily so they're never caught with nothing ready.