Ad Blueprint

The hook. Wins or loses the ad in the first 1–3 seconds — one idea, one visual, one reason not to scroll.

The Opening is the only beat with no second chance. Its entire job is to buy the next three seconds, which means it must carry exactly one idea. Decoded winners overwhelmingly open on a single visual with a single line of on-screen text; the moment a second idea enters the frame, completion rate drops. Note that Opening takes the smallest share of runtime of any front-half beat — the hook is short by design, because its output is attention, not comprehension.

Most common failure: Stacking three ideas into two seconds. Brand logo, product shot and a claim in the same frame reads as an ad and gets scrolled.

Establishes who is talking and why the viewer should care — the credibility and relatability deposit the rest of the ad spends.

Context is the most under-appreciated beat in the study and the second-largest by runtime. It answers two questions the viewer asks silently: who are you, and is this for someone like me? Ads that skip Context to get to the product convert worse across every vertical we decoded, because Delivery has nothing to spend. The strongest Context beats are shown, not stated — a kitchen at 6am does more relatability work than any line of voiceover.

Most common failure: Credential dumping. Reciting qualifications instead of demonstrating them turns trust-building into a résumé.

Opens the emotional gap between where the viewer is and where they want to be — the ad's problem made personal.

Tension converts a general problem into a felt one. It is the shortest beat in the front half, and that brevity is deliberate: dwell too long on pain and the ad becomes unpleasant to watch, which the feed punishes. The best Tension beats name a cost the viewer is already paying but has never articulated — the recognition itself is the persuasion, and it primes Delivery to land as relief rather than as a pitch.

Most common failure: Manufacturing a problem the audience doesn't have. Invented pain reads as manipulation and poisons the rest of the ad.

Presents the product as the mechanism that closes the gap — features shown as felt benefits, never listed.

Delivery is the largest beat in the study, and the one most often executed badly. Winners spend it demonstrating a mechanism — why this works — rather than enumerating features. A single uninterrupted shot of the product doing its job outperforms a montage of specifications in every vertical we decoded, because the viewer's objection at this moment is not 'what does it do' but 'why would that work for me'.

Most common failure: Feature listing. Six benefits in six cuts gives the viewer nothing to believe and nothing to remember.

Proof that it worked for people like the viewer — reviews, results and receipts arriving right when skepticism peaks.

Validation exists because Delivery creates a debt: you have just made a claim, and the viewer's next thought is 'says who'. Timing matters more than volume here. Proof placed immediately after the mechanism converts; the same proof placed at the very end of the ad is usually never reached, because the viewer who needed it left three seconds earlier.

Most common failure: Anonymous proof. '10,000 happy customers' validates nothing — one specific person with a specific result validates everything.

Reframes the decision — from 'should I buy this product' to a bigger identity or goal the purchase now serves.

Shift is what separates ads that convert from ads that merely inform. It moves the frame off the transaction: the purchase stops being an expense and becomes a step toward something the viewer already wants. Ads without a Shift beat close on price; ads with one close on meaning, and meaning survives comparison shopping in a way that a discount does not.

Most common failure: Reframing into something the viewer never wanted. A Shift only works if it lands on a goal they already hold.

The ask. Converts attention into action with one unambiguous next step — soft prompts for cold traffic, hard offers for warm.

The Close is the shortest beat and the one most often over-engineered. One instruction converts; two compete. The decisive variable is temperature matching: cold audiences reward a low-commitment ask (comment, save, watch the next one), warm audiences reward a hard offer with a reason to act now. Running a hard offer at cold traffic is the single most common structural error in the decoded set.

Most common failure: Stacking asks. 'Link in bio, comment YES, and follow for more' splits intent three ways and converts on none.