The Ad Building Blocks

Stop thinking "make an ad." Start thinking "assemble and test blocks." Every high-performing ad you have ever scrolled past is a sequence of modular parts, and the parts repeat from brand to brand and from category to category. Once you can name them, you can do two things most teams cannot. You can deconstruct any winner into its components, and you can diagnose a loser at the block level instead of scrapping the whole thing and starting again.

That second point is where the money is. When an ad underperforms, the instinct is to bin it and shoot something new. That is slow, expensive, and it throws away the parts that were working. A block-level operator does the opposite. They find the one block that is leaking, swap it, and keep everything else intact. A weak ad is rarely weak everywhere. It is usually a strong body with a dead hook, or a great hook attached to proof nobody believes.

The blocks fall into three groups. The structure blocks are the spine that every ad needs. The product blocks are how you show the thing you sell. The person blocks are how you carry the human the ad is for. A finished ad is a structure spine with product and person blocks slotted into the body in a deliberate order.

The three structure blocks

Every ad has all three. They are not optional and they are not interchangeable. Each one has a single job, and an ad fails the moment one of them tries to do another one's work.

Hook (0 to 3 seconds)

The hook signals relevance and opens a curiosity loop. Its only job is to stop the scroll and earn the next three seconds. It is not the place to sell, to name the brand, or to list features. It is a filter that pulls in the right person and lets everyone else keep scrolling. The hook is not just the first spoken line. It is what the viewer reads, hears, and sees in frame zero, plus the overall vibe of the creative, all at once.

Example. A close-up of a hand scraping thick limescale off a shower screen while a voice says, "If your glass looks like this, you are cleaning it wrong." The visual, the line, and the on-screen text all point at one person: someone who scrubs their shower and loses.

Body

The body turns interest into desire. This is where the product and person blocks live. It demonstrates, it pre-empts objections, it builds credibility, and it moves the viewer through the stages of awareness toward the product. A good body does not lurch straight from the hook to the buy button. It walks the viewer from "I have this problem" to "this is the thing that fixes it" one step at a time.

Example. After the limescale hook, the body shows the failed alternative (a cupboard of supermarket sprays that never worked), then the product being wiped on once and the glass going clear, then a row of five-star reviews from people with the same shower.

CTA

The CTA is the natural conclusion of the story, not a bolted-on instruction. It reinforces the transformation the viewer just watched, removes the last scrap of risk, and gives one clear next step. A good CTA feels like the obvious thing to do next, because the body earned it. A weak CTA reads as "buy now" with nothing behind it.

Example. "Get the clear glass back this weekend. Free shipping, and if it does not work on your shower, send it back." One action, the outcome restated, the risk removed.

The five product blocks

These are the ways you put the product itself on screen. You will rarely use all five in one ad. You pick the ones that move the specific buyer at the specific awareness stage you are targeting.

• Product introduction. The reveal, in context. The job is to show what the thing is, the moment it earns the right to appear. Example: the bottle comes out of the cupboard mid-demo, after the problem has been felt, not in the first second.
• Demonstration. The product working, in use, in real time. The job is to collapse doubt by showing rather than claiming. Example: one wipe across the glass, no cut, and the limescale is gone.
• Features and benefits. What it does, and what that does for them. The job is to translate a spec into a consequence the buyer cares about. Example: "Plant-based, so you can spray it with the kids in the bathroom."
• Buying experience. How easy it is to get, set up, and use. The job is to remove the friction a buyer imagines between paying and benefiting. Example: "Arrives in two days, no subscription, twist the cap and go."
• Unboxing. The first-impression moment of receiving it. The job is to make the purchase feel real and well made before it is even used. Example: the box opens, the bottle sits in moulded card, the insert card reads "made in small batches."

The six person blocks

These carry the human. They are what make an ad feel like it is about the viewer and not about the company. Performance creative leans on these harder than brand creative does, because they are what create the emotional pull that the product blocks then justify.

• Problem statement. Name the pain in the buyer's own words. The job is recognition: the viewer should think "that is me." Example: "I scrub the shower every week and it still looks dirty."
• Failed alternative. What they already tried that did not work. The job is to clear the field of the buyer's current options so yours has room to land. Example: "I have tried every spray in the supermarket and a magic eraser. Nothing."
• Desired outcome. The after state they actually want. The job is to make the payoff concrete and worth acting on. Example: "Glass so clear it looks like there is no screen there at all."
• Before and after. The visible gap your product closes, shown side by side. The job is undeniable contrast in a single frame. Example: filthy panel on the left, clear panel on the right, same shower, same lighting.
• Social proof. Other real people, real results. The job is to borrow the credibility of the crowd so the viewer does not have to take the risk first. Example: a stack of genuine reviews, or a creator filming her own shower going clear.
• Storytelling. A narrative that carries all of the above in one arc. The job is to hold attention and lower the viewer's guard so the message lands without feeling like a pitch. Example: a founder's two-minute account of why she made the product after years of renting flats with grimy bathrooms.

The core formulas

A formula is just a deliberate order of blocks. There is no single correct sequence. There are proven sequences that match different awareness stages and different angles. Here are four starting templates we reach for. Treat them as scaffolds, not scripts. The whole point of thinking in blocks is that you can swap any single one and test it on its own.

1. The problem-led direct response ad

Hook → Problem statement → Failed alternative → Demonstration → Social proof → CTA with a reason to act now.

The workhorse. It opens on the pain, clears the buyer's existing options, proves the product works on screen, backs it with the crowd, then closes. Best for problem-aware and solution-aware buyers who already feel the friction and need a reason to believe yours is the fix.

2. The desire-led launch ad

Hook → Desired outcome → Product introduction → Features and benefits → Social proof → CTA.

Leads with the after state rather than the pain. It paints the outcome the buyer wants, reveals the product as the route to it, then earns belief with features and proof. Best for new releases, aspirational categories, and warmer audiences who already know the problem and want the upgrade.

3. The founder story ad

Hook → Storytelling (founder) → Problem statement → Demonstration → CTA.

Carries the whole ad on a narrative. The founder's story is the vehicle that delivers the problem, the reason the product exists, and the proof, all without feeling like an ad. Best for building trust at the top of the funnel and for brands whose origin is a genuine differentiator.

4. The objection-handler retargeting ad

Hook → Problem statement → Before and after → Social proof → Buying experience → CTA with risk reversal.

Built for people who already know the product and did not buy. It names the specific doubt that stopped them, shows the result, stacks proof, removes the friction of buying, and closes with a guarantee. Best for the warmest audience, where the work is handling the last objection rather than introducing anything new.

The seven-layer teardown method

This is how you turn someone else's winner into your own playbook. You do not watch a winning ad once and "get the gist." You watch it seven times, and each pass you focus on one layer only. Trying to absorb everything at once means you absorb nothing. One lens per pass, and by the end you have the full schematic.

1. Pass 1: the hook. Watch only the first three seconds. What did you read, hear, and see in frame zero? Which block opened the ad, and which exact person did it filter for?
2. Pass 2: the script structure. Ignore the visuals. List the blocks in order as they appear. Write the sequence out. This is the formula you are stealing.
3. Pass 3: visuals and transitions. Mute it. Watch what is on screen and how one shot becomes the next. Note the cuts, the camera moves, the settings, the props.
4. Pass 4: on-screen text and supers. Read every caption and overlay. Note when text appears, what it emphasises, and where it is doing work the voiceover is not.
5. Pass 5: pacing. Watch the rhythm. In a strong ad something changes roughly every one and a half seconds: a cut, a new line, a text change, a zoom. Mark where the pace quickens and where it deliberately slows.
6. Pass 6: congruency with the landing page. Click through. Does the page match the ad's promise, look, and language? A great ad pointed at a mismatched page leaks every click it earned.
7. Pass 7: conversion objective and funnel stage. Step back. What is this ad actually for, and who is it for? Is it a cold top-of-funnel asset raising a problem, or a warm retargeting asset closing a sale? The job changes how you judge everything above.

Seven passes on one ad takes fifteen minutes and teaches you more than reading a hundred threads about creative. Do it on three winners in your category before you brief anything.

How to build a swipe file by block

Most people build a swipe file by brand: a folder of ads they liked. That is close to useless when you sit down to write, because you cannot search "I need a hook" in a pile of whole ads. Build your swipe file by block instead. Tear ads down with the seven-layer method, then file each block where you can find it again.

• Hooks, filed by type. Problem-led, curiosity-gap, contrarian, demo-first, founder-to-camera, and so on. When you need a new opening, you open the type that fits your angle.
• Bodies, filed by demonstration style. One-shot demo, before and after, listicle, green-screen explainer, side-by-side comparison. When a body is failing, you reach for a different style.
• CTAs, filed by mechanism. Urgency, scarcity, risk reversal, value anchor, social nudge. When a closer is soft, you swap the mechanism.

Then rate every block out of ten, and record the timestamp it occupies in the source ad. The timestamp matters: "the hook" is the first three seconds, but "the proof block that made me believe it" might be at 0:34, and you want to find that exact moment again. A swipe file of rated, timestamped blocks is a parts bin you can build from. A folder of whole ads is a museum.

Block-level diagnosis

This is the payoff of the whole system. When an ad underperforms, you do not start over. You read the symptom, trace it to the one block responsible, and fix that block while leaving the rest intact. Match the symptom to the block, change the block, relaunch.

Notice that most fixes are a single block swap. A new hook on a proven body is a new ad to the platform, costs a fraction of a fresh shoot, and isolates the variable so the result actually teaches you something. That is the difference between a team that ships one ad a week and a team that ships ten.

The best practices that sit on top

The blocks are the parts. These rules govern how you assemble them. They hold regardless of category, format, or angle.

The three-second rule

Deliver your core promise in the first three seconds. Never save your best material for the end, because most viewers never reach the end. The single strongest thing you have, the clearest result, the boldest claim, the most arresting visual, belongs up front where it earns the rest of the watch. Front-load, do not build to a climax.

Authenticity beats production value

Believable content out-converts polished content, especially in the demonstration and social proof blocks. A shaky phone clip of a real person getting a real result will usually beat a glossy studio version of the same claim, because the polish reads as an ad and the rawness reads as the truth. Spend your effort on what is being shown, not on the gloss around it.

The emotional sandwich

Open on emotion, support with logic and proof, close on emotion. People buy on feeling and justify with logic, so you give them both in the right order. The hook and the desired outcome pull the feeling. The demonstration and features carry the logic. The CTA returns to the feeling and the transformation. Emotion, then logic, then emotion.

One message per ad

Every ad makes one promise to one person. If you find yourself stacking two angles or selling two outcomes in a single asset, split it into two ads. A clear single message is readable in the data and learnable. A muddled one teaches you nothing and converts no one, because the viewer cannot tell what it is for.

Prepare your replacement blocks before launch

Do not wait for an ad to fatigue before you think about the next hook. Have the next hook and the next proof block written and shot before the first version goes live. When the data comes in, you want to hot-swap the leaking block the same day, not start a new production cycle. The teams that win are the ones with the next block already in the parts bin.

Where this leaves you

Naming the blocks changes the job. You are no longer making ads from a blank page. You are tearing down what already works, filing the parts, assembling them into proven sequences, and swapping single blocks when the data points at one. It is faster, it is cheaper, and it compounds, because every teardown and every test adds another rated block to a library you keep drawing from.

If you want a team that runs this system on your account, tears down the winners in your category, and ships the volume of block-level tests it takes to keep finding winners, book a free creative audit and we will show you which blocks in your current creative are leaking.