The Science Behind the SHARP Score

Here's a question that should keep every media buyer up at night: if creative quality is the single largest driver of whether your ad works, why do most teams spend 90% of their optimization energy on targeting and bidding?

The data is unambiguous. Nielsen's 2017 meta-analysis of ~500 CPG campaigns found that creative quality drives 47% of sales contribution from advertising — more than reach (22%), brand (15%), targeting (9%), and recency (7%) combined. For digital campaigns specifically, NCS (now Circana) put creative's share even higher: 56% of sales ROI. Kantar's CrossMedia database attributes ~50% of the variation in ad impact to creative quality — surpassing reach, frequency, and media synergy.

And yet. Walk into most agencies or growth teams and ask how they evaluate creative before it goes live. You'll hear some version of: “we review it in a meeting,” “the CD signs off on it,” or simply “we test it and see what happens.” The single most important variable in the entire advertising system is evaluated by committee intuition and post-hoc performance data.

Crevaluate exists because we believe there's a better way. Not to replace creative judgment — but to augment it with the same rigor we apply to every other part of the marketing stack.

The SHARP Score is built on a simple premise: decades of advertising effectiveness research have identified five things that predict whether a creative will work. Not sometimes, not in some categories — consistently, across platforms, formats, and markets.

Each maps to a fundamental question. Each is scored 0–100. Together, they form a profile that tells you not just how good a creative is, but where it's strong and where it's weak — which is far more useful.

Every creative starts with the same challenge: earning the right to be seen. In a feed environment — which is now the dominant context for advertising — you have somewhere between half a second and three seconds before the viewer decides you don't exist.

Nielsen's research on video ad recall shows significant lift within the first 3 seconds of exposure, with 65% of viewers who watch the first 3 seconds continuing to at least 10. TikTok's own data shows that mobile-shot, creator-style creative has a 63% higher chance of outperforming studio-shot content on conversions. On YouTube, the first 5 seconds determine whether viewers hit skip. On Snapchat, you have about a second.

We evaluate Stop by looking for pattern interrupts, curiosity gaps, visual impact in the first frame, text hooks, direct address, and thumbnail appeal — each calibrated to the platform where the creative will run.

Getting someone to stop is necessary but not sufficient. Karen Nelson-Field's landmark attention research at Amplified Intelligence established that roughly 2.5 seconds of active attention is the minimum threshold for memory encoding. Below that, the ad has negligible brand impact — regardless of how many times it was “viewed.”

The implications are stark. On Facebook Feed, where average active attention is ~1.7 seconds, most ads never cross the encoding threshold. The creative must work harder — and differently — than on YouTube, where you start with ~8 seconds of runway.

System1's FaceTrace research adds another dimension: ads with a clear narrative arc generate dramatically stronger emotional response than rapid-cut montages. Story structure isn't a nice-to-have; it's the mechanism that sustains attention. We evaluate pacing, narrative tension, information revelation, energy arcs, and duration fit — because Hook is where most creative silently fails.

This is the big one. If you remember one thing from this page, make it this: emotion is the single strongest predictor of long-term commercial effectiveness.

Les Binet and Peter Field's analysis of the IPA Effectiveness Databank — the largest database of advertising case studies in the world — found that emotional campaigns deliver twice the profit of rational campaigns. Not twice the awareness, not twice the recall — twice the actual profit. Their recommended split is 60:40 brand-building (emotional) to activation (rational).

Daniel Kahneman's dual-process framework explains why. Most advertising works through System 1 — fast, intuitive, emotional processing. Robert Heath's research on Low Attention Processing shows that even ads viewed at low attention can build emotional associations that influence brand preference. You don't need people to actively engage with your message. You need them to feel something.

Orlando Wood at System1 took this further in his book Lemon, showing that “right-brain” creative features — characters, humor, story arcs, melody, metaphor — consistently correlate with effectiveness, while “left-brain” features — text on screen, voiceover monologue, rapid cuts — correlate with declining performance. System1's Star Rating system, independently validated by IPA analysis across 4,000+ ads, nearly doubles the accuracy of predicting market share growth when combined with media spend data. Realeyes' attention AI achieves 75–78% accuracy in distinguishing high-performing creative from underperformers.

We score Affect by evaluating emotional intensity, arc structure (Kahneman's peak-end rule applies directly to ad creative), authenticity, brand-emotion connection, and the presence of right-brain features that research consistently links to effectiveness.

An ad can stop the scroll, hold attention, and make you feel something — and still fail if nobody remembers which brand it was for. This is more common than you'd think, and it's the reason Byron Sharp's How Brands Grow remains the most important book in modern marketing science.

Sharp's core insight is that distinctive brand assets — colors, characters, sounds, logos, taglines — are the primary mechanism of brand growth. Kantar's data backs this up: winning ads consistently rank in the top 20% for brand integration, while poor performers cluster in the bottom half. Google's ABCD framework research shows that ABCD-compliant YouTube ads deliver up to +30% lift in short-term sales likelihood. Veritonic and Audacy's joint study found that sonic branding boosts radio ad recall by +17% and podcast ad recall by +14%.

The key insight is that brand integration needs to feel organic, not forced. An end card with a logo is the weakest form of branding. Weaving the brand into the narrative — making it part of the story rather than a footnote — is what separates creative that builds memory structures from creative that entertains anonymously.

The final dimension is the most tactical — and the most often botched. HubSpot's analysis of 330,000+ CTAs found that personalized calls to action convert 202% better than generic ones. “Get 30% off today” vs. “Learn more” isn't a style choice; it's a 2-3x performance gap.

Platform-specific data reinforces this. Snap Ads achieve a swipe-up rate 5x higher than the average CTR across other social platforms. Amazon's research shows that including reviews or ratings in creative increases CTR by 15-25%. Cialdini's principles of influence — scarcity, social proof, authority — aren't academic abstractions; they're levers you can see (and measure) in the creative itself.

We evaluate CTA presence, specificity, urgency, visual prominence, timing, social proof, and problem-solution framing. Importantly, Push is weighted differently depending on the campaign objective — a brand awareness campaign shouldn't be penalized for a soft CTA, and a conversion campaign shouldn't get a pass for missing one.

A composite score is useful for benchmarking, but the real value is in the shape. Two creatives can both score 71 overall — but one might be all Stop and Push with weak Affect, while the other is strong on Affect and Recall but fails to hook past the first two seconds. They need completely different fixes.

A TikTok ad optimized for awareness and a Google Ads asset optimized for conversion should not be judged by the same criteria. This seems obvious, but most creative evaluation — human or automated — ignores it entirely.

The SHARP composite score adjusts its dimension weights based on two inputs: your campaign objective and your target platform.

An awareness campaign upweights Stop, Affect, and Recall — the dimensions that build brand salience. A conversion campaign upweights Push and deprioritizes everything except Stop. Platform norms further refine these weights: TikTok's extreme scroll speed means Stop and Hook matter more there than on YouTube, where you have more runway but must sustain attention for longer.

This is also why the same creative can — and often does — score differently across platforms. That's not a bug; it's the whole point. A 9:16 creator-style video might score 85 on TikTok and 52 on YouTube pre-roll, not because the creative is inconsistent, but because the platforms are.

We didn't invent any of the science behind SHARP. We synthesized it. The framework draws on five bodies of research, each with decades of evidence behind it: