Creative Strategy
The methodology for turning creative production into a predictable, data-driven system that fuels efficient scale.
The Philosophy
Why a creative operating system beats intuition. The premise that underpins every section that follows.
The brands that find efficient scale don't do so by guessing. They do so by creating the creative operating system that maps the set of ads required to achieve their forecasted spend.
Not a wishlist. Not a mood board. It's the Ad Plan. The data-driven system that tells you exactly what to produce based on your desired financial outcomes.
Like every CTC methodology, this one exists to bridge marketing and finance. The financial plan sets the month's revenue target and the spend required to achieve it. The Ad Plan converts that spend forecast into a defined set of creative deliverables, how many ads, of what type, from which producer, attached to which calendar moment, that the creative team operates against. The forecast and the brief become two views of the same number. Neither side has to guess what the other needs.
The ad plan is broken into two parts:
- Part 1: Quantifying the number of ads needed to hit spend targets.
- Part 2: Determining how to allocate those ads across calendar moments, evergreen, products, creative types, and producers.
Ad Creative Is a Hits Business
A small number of ads drive a disproportionate share of all results. Volume is the lever that bends probability in your favor.
The Pareto Principle applies with brutal clarity to advertising. After analyzing over 10,000 ads for brands like Igloo, Nike Strength, Skullcandy, and GoRuck, we see the same pattern repeat: most ads produce average outcomes. Some produce nothing, and a tiny fraction become outliers that carry the entire account.
Why this matters before we talk about volume. Nielsen Catalina Solutions found that 56% of variance in ad auction outcomes is attributable to creative, not targeting, not budget, not campaign structure. [Nielsen Catalina Solutions, 2017] Once creative is established as the dominant variable, the question becomes how to produce enough of it to make the math work.
Of all ads created in 2026 across the Statlas dataset:
The raw "hit rate" from creation to whale is roughly 1 in 100.
This is not about making better ads. It is about making enough ads that the math works in your favor. With a 1.8% outlier rate, the difference between launching 10 ads and 100 ads is not linear. It is the difference between gambling and investing.
Creative strategy is not about making more ads. It is about increasing your probability of producing the next outlier. And it is about building a system where your average ad is profitable, so that you can afford to take as many shots on goal as possible.
What an outlier actually means for a specific brand
The dataset numbers above describe the typical brand. But a flat dollar threshold breaks at the extremes. A $5K ad is a whale on a $20K/month account and a rounding error on a $1M/month one. So for any specific brand, we define outliers relative to the account's own spend distribution:
- Whale: more than one standard deviation above the account's mean spend per ad.
- Winner: above the mean, within one standard deviation.
- Testing: around the mean.
- Dust: below the mean.
The Creative Audit computes the brand's actual mean and standard deviation, then assigns every ad in the active inventory to a tier against that brand's own math. The dataset benchmarks are the reference distribution. The brand's calculation is what feeds the formula.
Creative strategy is not about making more ads. It is about increasing your probability of producing the next outlier. And it is about building a system where your average ad is profitable, so that you can afford to take as many shots on goal as possible.
Motion's 2026 Creative Benchmarks confirm the same pattern
Across 550,000+ ads and $1.3 billion in Meta spend from 6,000+ advertisers, Motion found that only ~5% of ads become real winners. Roughly half receive minimal or zero spend. Critically, hit rate increases with volume: top-tier advertisers ship 12-19+ new creatives per week and achieve hit rates more than double those of smaller accounts. High hit rates can actually signal insufficient testing, not superior judgment.
The Creative Demand Formula
Every brand on Meta faces the same monthly question: how many ads do I need? The answer is a function of two account-specific variables, not a fixed number.
The Problem
Most brands guess. They look at what they made last month and add a few. They pick a round number. They copy a competitor's cadence. They let vendor capacity dictate volume. None of this is a strategy. All of it produces the same outcome, either too many ads wasting production dollars, or too few ads starving the algorithm and stalling growth.
We manage $3.4 billion in Meta spend across our dataset. We can see exactly what happens to every ad, every month, in every account. And what the data shows is that the answer to "how many ads do I need" is not a fixed number. It's a function of two things: how well your existing creative holds up, and how efficiently your new creative converts into spend.
The Formula
Every input is defined, sourced, and brand-specific.
| Input | Definition | Source |
|---|---|---|
| Target Spend | The brand's spend goal for the planning month. | Media plan or ad demand forecast |
| Current Spend | Estimated end-of-month spend. Primary: MTD spend plus the Statlas calendar projection. Fallback: MTD spend plus daily run rate adjusted for day-of-week pacing and any planned moment spikes. | Statlas prod.metrics_daily + prod.targets_daily
|
| Carry Rate | Computed per ad, then aggregated. For every ad in the brand's trailing 60-day active inventory, we measure whether it survives into next month and how much of its spend it retains. The brand-level Carry Rate is the spend-weighted average of those per-ad outcomes. | Statlas compass.ad_demand_health
|
| Expected $/New Ad | Computed per ad, then aggregated. For every newly introduced ad in the brand's trailing 60 days, we measure the spend it absorbed in its first month. The brand-level Expected $/New Ad is the average of those per-ad outcomes, distributed across Whale, Winner, Testing, and Dust tiers. | Statlas compass.ad_demand_health
|
The formula tells you the gap between where your existing creative will land you and where you need to be, then divides that gap by the productivity of each new ad you introduce.
How Meta accounts actually work
Your ad account is a living inventory. Every month, three things happen simultaneously:
- Some ads keep working. The algorithm keeps feeding them impressions, they keep converting, and spend flows. These are your survivors.
- Some ads die. The algorithm stops serving them, efficiency drops below threshold, or they fatigue out. This is your decay.
- You introduce new ads. Some catch. Most don't. The ones that catch absorb spend, sometimes a lot. The ones that don't are the cost of discovery.
What the data says about Carry Rate
Carry Rate is not a guess. We measured it across the trailing 12 months of ad survival data, grouped by each brand's own spend tiers (relative to that brand's account mean).
- Whales (more than 1σ above account mean) survive at the highest rate, around 82%, but the survivors lose roughly 30% of their spend as the algorithm rebalances.
- Winners (above account mean, within 1σ) survive at around 70% and only lose about 10% of their spend.
- Testing tier ads (around the account mean) often grow. Survivors here can see 118% spend retention as the algorithm scales them up.
- Dust (below the account mean) survives at the lowest rate, around 57%. But the ones that do survive can see 3-4× spend growth as the algorithm promotes them out of testing.
The net effect: roughly 70-75% of your current spend carries forward on its own. The other 25-30% must come from new creative you introduce.
But that's the passive rate, what happens if you don't touch the throttle. Brands that actively push budgets on existing winners can carry 85-100% or more of prior spend forward. The actual observed range is 65% to 122%, with the highest carry rates belonging to brands that scale budgets on proven performers before reaching for new production.
What the data says about Expected Spend Per New Ad
Also not a guess. Across the trailing 12 months of new-ad launches, the distribution is a power law for every account we have data on, with the absolute dollar values scaling to each account's mean.
| Tier | Share of New Ads | % of All New-Ad Spend |
|---|---|---|
| Whales (more than 1σ above mean) | ~1% | ~41% |
| Winners (above mean, within 1σ) | ~5% | ~34% |
| Testing (around the mean) | ~17% | ~19% |
| Dust (below mean) | ~77% | ~6% |
The expected spend per new ad is the probability-weighted sum, calculated from each brand's own trailing 60-day distribution. Dataset benchmarks show this number tends to range from roughly $192 at mid-spend accounts to $397 at higher-spend accounts, but the brand's actual number is what feeds the formula.
Both of these numbers, the Carry Rate and the Expected Spend Per New Ad, change based on the health of the specific account. That's where the Creative Score comes in.
The Five Creative Score Metrics
The Creative Score is not a vanity number. It's a diagnostic that tells you which version of the Carry Rate and Expected Spend Per New Ad to plug into the formula. Five metrics, each mapped to a specific variable in the math.
Three metrics drive Carry Rate
Evergreen Share directly measures survival rate, an ad running 30 days is by definition an ad that survived. A brand at 45% evergreen has far more inventory carrying forward than one at 10%. Higher evergreen = higher Carry Rate = smaller gap for new creative to fill.
Spend Degradation measures spend retention, how much of a surviving ad's spend holds up as it ages. A brand showing −39% spend degradation has surviving ads bleeding almost 40% of their spend each week. Every dollar that bleeds off an existing ad is a dollar that must be replaced.
Ad Concentration determines the spend mix. Whale-heavy accounts carry at 59% per tier while mid-tier distributed accounts carry at 64-92%. High concentration means your Carry Rate is hostage to a handful of ads. Low concentration = structurally more resilient Carry Rate.
Two metrics drive Expected Spend Per New Ad
Zero Revenue Rate is the most direct predictor of how efficiently new creative converts into spend. The dataset baseline is 77%, more than three quarters of all new ads earn under $100. A brand at 30% produces creative the algorithm wants to distribute. A brand at 65% needs two to three times more ads to fill the same spend gap.
ROAS Degradation modifies the expected value of new creative over time. A brand with 5% ROAS degradation has winners that stay efficient as they age, today's winners become tomorrow's carry-forward inventory. A brand with steep ROAS decay finds winners that burn hot and die fast. Over a multi-month horizon, ROAS Degradation determines whether your creative investments compound or constantly reset.
The Creative Score makes the difference between accounts visible. The formula makes it actionable.
How the Score becomes a plan
The five metrics combine into a Creative Score that answers the question the formula leaves open: which version of the math applies to this brand?
- A brand with a Creative Score of 76, high evergreen, low concentration, low zero-revenue rate, mild degradation, has both a high Carry Rate and a high Expected Spend Per New Ad. Their existing creative carries forward 80-90% of current spend. New creative converts at well above the $192 baseline. They can scale 36% with 75-110 new ads, closer to the MSQC case, which grew 27% with just 41 new ads because their creative inventory was healthy.
- A brand with a Creative Score of 35, low evergreen, high concentration, high zero-revenue rate, steep degradation, has a Carry Rate that might be 55% instead of 73%, and an Expected Spend Per New Ad that might be $120 instead of $192. They need 300+ ads to achieve the same growth, and even then the gains are fragile because next month's carry will be low again.
Two brands at identical spend levels can have wildly different creative demands based on the health of their existing inventory and the conversion efficiency of their new production. The Creative Score makes that difference visible. The formula makes it actionable.
The Three Levers
Every creative strategist, every media buyer, every growth lead should think about creative demand in terms of three levers. Everything we do maps to one of them.
Mine historical winners
The work we do before asking the client for anything new. Revive last year's top performers. Re-enable paused winners. Launch existing creative into untouched audiences and international markets. Extract underspending ads with strong ROAS into their own campaigns instead of letting them die. Every dollar a historical asset earns is a dollar new production doesn't have to.
Make better ads, not just more ads
Move the Expected Spend Per New Ad. The single biggest efficiency lever in the formula. This is where messaging diversity, format breadth, and Entity ID expansion live. The work isn't volume for volume's sake, it's clarity about what each ad is actually saying, to whom, and why.
The Creative Score makes that difference visible. The formula makes it actionable.
The Catalog Layer
Catalog ads aren't creative. They're inventory. They sit on a different math, run on a different cadence, and need to be carved out before the formula applies.
Every other section of this canon treats an ad as a creative entity, an Angle expressed through a Treatment, produced by a source, attached to a moment. Catalog ads break that frame. Meta assembles them at run-time from a product feed. The unit of work isn't the ad, it's the feed.
This is not an esoteric distinction. Across the CTC dataset, catalog ads are routinely the highest-ROAS line item in an account, and routinely the most under-discussed in creative planning, because they don't get briefed, designed, or tagged in Motion the way creative ads do. They get configured.
Meta's own published numbers corroborate the directional finding. Meta reports a 39% increase in ROAS for advertisers running Advantage+ catalog ads, a 25% CPA improvement when Advantage+ Catalog runs inside an Advantage+ Sales campaign, and a 4% lower cost-per-purchase for catalogs with 20+ items. [Meta Business · Advantage+ Catalog Ads] Meta's published lifts are for catalog as a whole. The CTC portfolio 1.60× blended ROAS lift sits in the same neighborhood. The more actionable insight is where the lift concentrates (acquisition) and where it does not (retention).
How catalog share varies by archetype
Catalog share of total Meta spend tracks closely to SKU count and revenue distribution. Across the 44 CTC managed Meta accounts, the portfolio splits cleanly into four archetypes. Spend-weighted, trailing 90 days, 7-day click:
| Archetype | Brands | Spend (90d) | Cat ROAS | Crea ROAS | Cat ACQ | Crea ACQ | Cat RTN | Crea RTN |
|---|---|---|---|---|---|---|---|---|
|
Catalog-led ≥30% catalog share |
7 | $4.8M | 3.59 | 2.87 | 2.83 | 2.33 | 7.37 | 7.83 |
|
Hybrid 10-30% catalog share |
16 | $11.4M | 4.10 | 2.90 | 3.97 | 2.57 | 5.06 | 5.26 |
|
Creative-led 5-10% catalog share |
5 | $4.2M | 2.88 | 2.60 | 2.39 | 2.38 | 4.52 | 3.89 |
|
Minimal catalog <5% catalog share |
16 | $14.6M | 2.48 | 1.68 | 1.74 | 1.33 | 3.28 | 5.56 |
Catalog ACQ outperforms creative ACQ in every archetype. Catalog RTN is tied with creative RTN in three of four archetypes and meaningfully under-performs in the fourth, the bucket where the catalog layer is barely being run. The catalog edge is in acquisition, not retention.
Three patterns hold across the portfolio. One: catalog ACQ wins by 1.21-1.55× depending on archetype, with the biggest lift in the Hybrid bucket where the account already trusts the feed for some share of acquisition. Two: catalog RTN and creative RTN are essentially tied, once an algorithm is retargeting a warm audience, asset format is not the lever. Three: the Minimal-catalog bucket represents 42% of portfolio Meta spend and the lowest efficiency of any bucket (1.68 creative ROAS). It is the largest single opportunity in the CTC managed portfolio.
Why catalog ads have different math
The Creative Demand Formula assumes ads are creative entities that fatigue, get replaced, and need monthly production. Catalog ads invert all three assumptions.
Fatigues. Replaces. Compounds slowly.
Carry Rate 70-75% passive. Expected $/New Ad ~$200-$400, power-law distributed. Outlier rate ~1 in 100. Production cost is a brief, a shoot, a designer. Replaces monthly.
Persists. Recombines. Compounds fast.
Carry Rate effectively ~100% as long as the feed and product set persist. The unit is the product set plus dynamic overlay configuration. Outliers come from feed segmentation, not from creative ideation. Production cost is a strategist, a feed audit, a few hours.
This means catalog spend should be carved out before applying the Creative Demand Formula. If a brand spends $1M/month and 30% is catalog, the formula's target spend is $700K. The remaining $300K of catalog spend doesn't need new ads to sustain. It needs feed hygiene, product set design, and overlay testing.
Carve-out rule. Catalog spend is excluded from the Creative Demand calculation and tracked as a separate line. The formula computes how many creative ads the brand needs. Catalog is a layer underneath, governed by feed hygiene rather than monthly production.
The two roles catalog plays
The long-tail retention layer
Lapsed customer DPA, site visitor DPA, abandoned cart DPA. The reason every account runs catalog at BOF is operational, not efficiency. Catalog RTN ROAS and creative RTN ROAS are statistically tied across the portfolio (5.52 vs 5.36). Catalog isn't winning retention: BOF targeting is. What catalog earns at BOF is coverage: feed-assembled retargeting requires zero monthly production and reaches SKUs the creative team will never brief. The configuration questions are feed segmentation and product set design.
The acquisition engine
Broad-audience DABA and Advantage+ with catalog. Across the portfolio, catalog ACQ runs at 3.28 ROAS vs 1.93 for creative ACQ, a 1.70× lift. The lift holds across every archetype, with the biggest gain in the Hybrid bucket (1.55×) and the smallest in Creative-led accounts where creative ACQ is already finely tuned. The diagnostic question is no longer "should this account run catalog ACQ?", it's "why isn't this account running it already?" 16 of 44 CTC managed accounts sit under 5% catalog share and at the bottom of the portfolio efficiency curve.
Why the algorithm prefers it. Meta clusters creatives by Entity ID: "If a creative has the same Entity ID as a previous ad, our system will treat the creatives as the same entity, sharing the learnings and delivery, which will likely limit your ability to reach new audience cohorts." Catalog ACQ natively generates distinct Entity IDs through product × audience × overlay combinations, what would take a creative team weeks to ship as static or video variants, the feed assembles at runtime. [Meta Growth Academy]
Where catalog enters creative-strategy decisions
Four general principles connect the catalog layer to the rest of this canon. Catalog is not a separate-track allocation question. It modifies how the Creative Demand Formula, the Product Matrix, the Creative Score, and the Format Split are computed.
- Carve-out math, quantified. Portfolio median catalog share is ~10%. The Catalog-led archetype runs 30-50%. The Hero-SKU archetype runs 5-15%. Single-SKU and editorial-led brands run under 5%. Accounts below the bottom of their archetype's range are flagged as "under-leveraged catalog" before the Creative Demand Formula is applied, that gap is structural opportunity, not a creative production gap.
- Catalog absorbs the long tail of the Product Matrix. Products that score into Test or Cut tiers (low spend, low performance via creative) may still earn through catalog without dedicated briefs. The Product Matrix decision sharpens to: brief creative or delegate to feed. The audit pulls both the Product Matrix and the catalog feed taxonomy so the long tail is reachable without competing for the creative production budget.
- Feed health is a separate diagnostic from the Creative Score. The five Creative Score metrics measure ad-inventory health, not feed-inventory health. The audit needs a parallel Feed Health check, product set coverage, custom-label depth, dynamic overlay testing, in-stock saturation, before the catalog allocation can be trusted. A healthy Creative Score over a malnourished feed produces the Minimal-catalog bucket outcome.
- Catalog ACQ sits inside Q3 format allocation, not alongside it. Today catalog appears in the format split as a separate line. Given the 1.70× ACQ lift, the audit should emit a recommended catalog ACQ allocation for every brand under 30% catalog share, not just for brands that already run it. Catalog ACQ is treated as a peer to static, video, carousel, and partnership in the Surround Sound coverage grid, with its own Entity ID expansion (product set × audience × overlay).
Evidence base
Four sources triangulate the catalog claims above. Two from inside Meta, one independent industry dataset, and one foundational study on creative as the dominant ad-auction variable.
Advantage+ Catalog Ads documentation
Meta's published lifts for Advantage+ Catalog: +39% ROAS, +25% CPA improvement when nested in Advantage+ Sales, +4% lower CPP at 20+ items. facebook.com/business/ads/advantage-plus-catalog-ads
Entity ID & creative diversification
Meta's documented rationale for why distinct Entity IDs unlock new audience cohorts. Catalog ACQ natively produces them at scale: product × audience × overlay combinations are individually fingerprinted by the algorithm.
550K ads · $1.3B Meta spend · 6,000+ advertisers
Independent dataset confirming the hits-business pattern: ~5% of new ads become real winners; top advertisers ship 12-19+ creatives/week. Catalog inverts the distribution by surfacing every active SKU at runtime. motionapp.com/thumbstop-pulse/creative-benchmarks-2026
Creative drives 56% of auction outcomes
56% of variance in ad-auction outcomes is attributable to creative, not targeting, not budget, not structure. The corollary for catalog: the 88% non-catalog majority of CTC portfolio spend still rides on creative quality. Catalog does not replace creative; it covers the SKUs creative will never brief.
Pull the brand's trailing 90-day Meta spend split by campaign objective and naming convention. CTC uses CatalogYes/No, DPA, DABA in campaign names. Compute catalog share = catalog spend ÷ total spend. Inventory the feed itself (product sets, custom labels, dynamic overlays) via the Meta Ads MCP catalog tools. The Creative Audit's Catalog Layer analysis reports the share, the retention vs acquisition ROAS split, and whether the brand is under-leveraging its catalog given its archetype.
The Creative Audit
Reading the account before building the plan. The formula gives you a number. Before you can turn that number into a plan, you need to know what you're working with.
Before we ask the client to produce anything, we do the diagnostic. The audit is the work we do on the brand's own data before we recommend a single new ad. It is the difference between asking a client for X new pieces of creative and telling them what they already have that can move the business first.
Every ad account has an existing inventory. Ads that are currently running, currently earning spend, currently serving the algorithm's allocation decisions. That inventory has a shape. It's concentrated in certain products and absent from others. It leans on certain formats and ignores others. It repeats certain motivators and has never tested others.
The allocation system in Part 2 makes decisions about products, formats, motivators, and producers. But those decisions are only as good as the diagnosis they're built on. Allocating creative to products without knowing your product revenue concentration is guessing. Splitting formats without knowing your format-level activation rates is guessing. The Creative Audit eliminates the guessing.
Carry-Forward Inventory, the tangible carry rate
The formula's inputs are calculated from the brand's own data, not the dataset benchmark. The audit computes this brand's mean spend per ad, this brand's standard deviation, and assigns every ad in the active inventory to Whale, Winner, Testing, or Dust against those numbers. The Carry Rate the formula uses is the brand's actual trailing-60-day survival math applied to those tiers, not a borrowed benchmark.
The audit is a structured diagnostic that answers six questions about the brand's current state. Each question maps directly to a pull from the Statlas and Motion data systems, and each answer feeds a specific allocation decision in the plan.
The six diagnostic questions
Where does the account stand today?
The Creative Score, five metrics that determine the Carry Rate and Expected Spend Per New Ad for this specific brand.
What is the spend target?
The brand's media plan or ad demand forecast for the planning month.
What does the marketing calendar look like?
Every confirmed moment for the planning month, plus every historical moment scored through the Promo LTV Scorecard. Each historical promo gets a composite score across NC Lift, aMER, CM%, and Post-30d returning-revenue lift. Tiers fall out of the math. Future moments inherit the tier of their closest historical comp.
Which products matter?
The revenue concentration curve, product scores across demand, margin, LTV, and inventory, and the product role assignments that determine which products get creative.
What's already in the account?
The current creative inventory, the best performing ads classified by format, product, motivator, and producer. This builds the Surround Sound coverage grid and reveals where the gaps are.
What production capacity exists?
Current-month scope from each active producer contract: UGC, Street Interview, Dedicated Creator, Branded Ads, and any brand-side creative capacity. Pull the active ongoing scopes for the planning month only. Never use historical or outdated scopes.
The six pulls produce a complete diagnostic of the account. The output is the Creative Audit, a brand-specific evidence base that makes every allocation decision in Part 2 provable rather than assumed. The methodology is universal. The audit is unique to the brand. The plan connects the two.
The Allocation System
From volume to plan. The Creative Demand Model gives you a number. The allocation system tells you where to aim it.
Every ad you produce must answer four questions in sequence. How you answer each one determines whether that ad has a chance of earning spend or whether it's dead before it launches.
- Is this ad serving a calendar moment or building evergreen inventory?
- Which product is it featuring?
- What creative type and format is it?
- Who is producing it?
Each question narrows the field. And each one is answerable from data, not instinct.
Moments vs. Evergreen
The first and most consequential split. It determines the shelf life of every ad you produce.
Moment creative has a defined expiration. It's tied to a promotion, a product drop, a seasonal event, or a cultural window. When the moment ends, the creative dies.
Evergreen creative has no expiration. It runs as long as the algorithm keeps serving it. It compounds. Last month's evergreen winners become this month's carry-forward inventory. Every point of Carry Rate in the formula is sustained by evergreen creative that's still working.
This is the connection to the formula. The Carry Rate is, fundamentally, a measurement of your evergreen portfolio's durability. A brand with a 75% Carry Rate has a strong evergreen base. A brand at 55% has an evergreen problem.
Input 1: The marketing calendar
Pull the brand's marketing events for the planning month. Every confirmed moment, promotions, product launches, seasonal events, brand milestones, gets classified by tier. The CTC tier methodology stays as the structural framework. The Promo LTV Scorecard layers on top, scoring every historical moment against actual revenue impact so the tier assignment is grounded in the brand's own results, not in judgment.
| Tier | Type | Revenue impact (Promo LTV) | Sanity-check range |
|---|---|---|---|
| Tier A · Evergreen Add | Product drops, collection launches, brand moments that compound after the event closes | T1 composite (50+) with positive Post-30d Lift. The moment's creative continued generating returning revenue after it ended. | 30+ ads |
| Tier A · Promo | Time-bound promotions that die on end date | T1 composite (50+) without Post-30d Lift. Strong in-window performance, no compounding tail. | 10-30 ads |
| Tier B | Secondary peaks; moments extending or preceding a Tier A | T2 composite (30-49). Moderate revenue impact, not a flagship. | 8-15 ads |
| Tier C | Calendar fillers; narrative continuity | T3 composite (under 30). Low revenue impact; candidates to retire. | 4-8 ads |
The tier definitions remain the CTC methodology. The Promo LTV columns connect each tier to actual revenue impact from the brand's own historical data. Future moments inherit the tier of their closest historical comp.
Input 2: Ad counts back-calculated from expected incremental spend
The ad count ranges in the table are sanity-check bounds, not the answer. The answer is back-calculated from the expected incremental spend each moment is forecast to drive:
If last year's Memorial Day Sale scored T1 with a MER of 2.4 and this year's MDS is forecast to drive $50K in revenue, that's roughly $21K of incremental spend. Divided by a brand Expected $/New Ad of $400, the moment needs about 52 new ads. The math sets the number. The sanity-check range tells you whether the math is producing something reasonable.
Input 3: Tier A · Promo vs Evergreen Add is measured
The split between Tier A · Promo and Tier A · Evergreen Add is not descriptive. It is determined by the historical Post-30d Lift signal. If a moment's creative continued generating returning revenue in the 30 days after the moment closed, it compounds and earns the full Evergreen Add allocation. If the returning-revenue line flattened or dropped, the creative died with the moment and the allocation is capped to avoid wasted production.
Sum the back-calculated allocations across all moments in the planning month. That's the moment allocation. Everything left over from the total demand number is evergreen.
Input 4: The peak allocation trap
Brands get this wrong. The instinct is to pile all production into the big moments. It feels smart. Why waste ads on evergreen when you have a product launch coming?
The data says the opposite. Brands with the highest Carry Rates and the best Creative Scores protect the evergreen allocation even during peak months. The reason is mechanical: Tier A · Promo creative dies. If you allocate 80% of a month's production to non-compounding moments, you've just built 80% of your inventory with a defined expiration date. Next month's Carry Rate collapses.
Soft guideline: moments that don't compound (Tier A · Promo and below without Post-30d Lift) should generally not exceed 30% of the planning month's total ad production. Tier A · Evergreen Adds (moments with positive Post-30d Lift) are not capped because the creative compounds and contributes to next month's Carry Rate. The back-calculation math will surface when this guideline is being violated.
Pull the brand's marketing events from Statlas for the trailing 12+ months. Run the full historical calendar through the Promo LTV Scorecard, computing baseline (non-promo days), per-promo NC Lift, aMER, CM%, and Post-30d returning-revenue lift. Composite score each promo (0-100) and assign T1/T2/T3. Map the upcoming month's planned moments to their closest historical comp and inherit the tier. The Creative Audit's Marketing Calendar analysis (Pull 3) surfaces this with per-promo and per-type averages, plus calendar recommendations for retire/repeat/rework.
Product Allocation
For moments, the product set is defined by the moment. For evergreen, product allocation is a portfolio decision, and not every product deserves equal creative.
For moments: the product set is defined by the moment
A sitewide sale features the top-performing products by trailing 90-day demand. A product drop features the drop's collection. A gifting moment features curated SKUs that fit the narrative. The canonical source is always the offer's landing page, confirmed against Statlas for in-stock status and recent demand.
For evergreen: product allocation is a portfolio decision
Revenue concentration across products follows the same power law as ad spend concentration. The shape varies enormously by brand:
- Hero-SKU brands, top 5 products drive 60%+ of revenue (Riddle Oil 64%, Denver Modern 58%). Need creative diversity within a narrow product set.
- Distributed-catalog brands, top 5 below 15% (Crop Shop Boutique 9%, Rifle Paper 4%). Need creative breadth across a wider product set.
The Product Matrix
Score each qualifying product across four dimensions:
- Demand, trailing purchase volume and velocity
- Margin, contribution margin per unit
- LTV, average customer lifetime value when this product is the first purchase
- Inventory, in-stock depth, weeks of supply remaining
The weighted composite is the Product Score. Products fall into roles based on Performance × Spend:
Hidden Gem
High performance, under-invested. Boost creative.
Champion
High performance, well-invested. Refresh with steady creative.
Test / Historical Revival
Low spend, low or historical performance. Small allocation to probe.
Cut
High spend, low performance. Reallocate budget.
Default evergreen split by product role: Champions 55% · Hidden Gems 20% · Historical Revival 10% · Test 10% · Reserve 5%. This is a starting position, not a rule. The Product Score adjusts the weights. A brand with strong new-product pipeline shifts toward test and launch. A brand with a stable catalog and high evergreen rate leans heavier into Champions and refresh.
Low-SKU brands
For brands with 5-10 products driving the entire business, the product allocation collapses. You can't get creative diversity from product breadth when you only have three products. Diversity comes from motivator expansion within the same product set, covered in Question 3.
Pull the brand's product sales from Statlas prod.product_sales (top 500 variants per day for Shopify stores) for the trailing 90 days. Pull LTV cohort data from dw.customer_cohorts (first-customer + 30/60/90 day LTV). Rank every product by revenue. Calculate the cumulative revenue curve. Score each product on Demand × Margin × LTV × Inventory. The Creative Audit's Product Matrix (Pull 4) provides the concentration profile, product scores, and role assignments. For hero-SKU brands (top 5 > 50% of revenue), the audit flags that creative diversity must come from motivator and format expansion within a narrow product set. For distributed-catalog brands (top 5 < 15%), it flags that product breadth is the primary diversity lever.
Creative Types & Formats
Entity ID diversity enters the system here. Meta's algorithm treats every unique creative asset as a separate entity. More unique entities = more independent signals = more optimization headroom.
Meta treats every distinct combination of image, video, headline, primary text as a separate entity. Two ads with different images but the same copy are two entities. The same image with different copy is also two entities. Entity ID volume matters independent of the number of "concepts" you're producing.
Meta documents this directly: "If a creative has the same Entity ID as a previous ad, our system will treat the creatives as the same entity, sharing the learnings and delivery, which will likely limit your ability to reach new audience cohorts." [Meta Growth Academy · How Does Entity ID Impact Creative Diversification?] Same Entity ID = shared signal pool, capped reach. Distinct Entity IDs = independent signals, expanded reach.
Meta's own measurement quantifies the lift. Across 2,703 ad sets, visually distinct creatives delivered +32% efficiency vs repetitive Entity IDs and +9% incremental reach from creative diversity. [Meta · Creative Diversity Study, 2,703 ad sets] Diversity is not a soft preference, it is a measured efficiency lever the algorithm explicitly rewards.
Three "different" ads, one signal
Same product shot, same copy, different headline color or aspect ratio. Meta's clustering treats these as one entity. You produced three ads and got one signal in return.
Three ads, three signals
Different format (static vs video), different visual environment (studio vs outdoor), different messaging angle (benefit vs social proof). Three independent learnings the algorithm can optimize against.
The format split
The right ratio of static images, short-form video, long-form video, carousels, and partnership ads is derived from the account's current state. Pull format-level activation and outlier data from Motion, not from Meta's Best Creatives report.
And resist the assumption that production value drives performance. Motion's 2026 benchmark dataset finds that text-only ads, product images with text overlays, and basic GIFs are common among top performers across the 550K-ad dataset. [Motion 2026 Creative Benchmarks] The format that wins is the one that converts efficiently, not the one that looks most expensive. Allocation should track activation and outlier rate by format, not aesthetic preference.
Baseline heuristic: allocate proportionally to activation rate, with a floor on every format to maintain diversity. If static activates at 28% and video at 18%, the split skews static, but you never zero out video because the algorithm needs format diversity to reach different placements.
Outlier divergence exception: when a format's outlier rate diverges from its activation rate, hold the floor. Video often activates less but produces bigger winners when it does. If outlier spend per winner in video is 2× the outlier spend per winner in static, you maintain the video allocation even though activation is lower.
Surround Sound: One Angle, Every Treatment
Within the format split, the Surround Sound framework governs Entity ID expansion at the messaging level. The equation is simple:
Two nouns, two definitions:
- An Angle is a Product (or Moment) plus a Message. The Osprey shoe plus "seamless transition from office to course." Crosswind Polo plus "performance under pressure." It is the what we are saying.
- A Treatment is one position on the brand's Surround Sound spectrum. The spectrum runs from product-focused to brand-focused. Each Treatment carries its own audience, visual cue, copy cue, and ad-type DNA. It is the how we say it.
Each cell of the grid (one Angle treated in one Treatment) equals one ad. The work of planning a month becomes writing a few Angles and treating each one across the full spectrum.
Treatments drive Entity IDs. Angles drive performance inside an Entity ID. Meta's algorithm clusters ads by execution shell (visual style, format, tonal register). The same Angle treated five different ways across the spectrum becomes up to five separate Entity IDs. The same Treatment with five different Angles probably collapses to one. Treatments are pre-engineered to force more than 70% execution distance between cells.
The Surround Sound spectrum
Example: one Angle running across six Treatments
| Treatment | Audience it reaches | Visual cue | Copy cue |
|---|---|---|---|
|
1. Spec Sheet Most product-focused |
Spec-comparison shopper | Isolated product, diagrams, callouts, cross-sections | Numbers-first; units, tolerances, comparative specs |
| 2. In Action | Technique-curious buyer | Hands in frame, real product use, slow-mo on critical moments | How-it-performs, technique callouts |
| 3. Social Proof | The "show me it works" buyer | Real customers, UGC voice, full-look or product-in-use | Quotes, ratings, peer language |
|
4. On the Counter Middle of spectrum |
Design-conscious buyer | Real environment, natural light, minimal styling | "Earns the space," design-restraint language |
| 5. Authority | Insider, category enthusiast | Pros, founders, third-party validators, documentary feel | Insider language, no hand-holding |
|
6. Philosophy Most brand-focused |
Brand believer, gift-giver buying meaning | Cinematic, abstract, product almost implied | Philosophy first, product implied |
Six Treatments × the month's Angles produces the matrix. A brand running 5 Angles across this spectrum produces 30 distinct cells, each with its own designer brief.
How the brand's spectrum gets defined
- Pull the brand's last 90 days of active creative from Motion.
- Auto-tag each ad on Motion's 8 dimensions (visual format, messaging angle, motivator, offer, asset type, hook, hook tactic, CTA).
- Cluster the ads into 5-7 Treatment archetypes for that specific brand. Name each Treatment in the brand's own vocabulary.
- Define each Treatment's audience, visual cue, and copy cue. This is a once-per-brand artifact, refreshed quarterly.
- Each month, write the Angles. Multiply Angles × Treatments. Emit a per-cell brief.
The Creative Score's Zero Revenue Rate tells you how efficiently you're filling the grid. A high zero-revenue rate often means you're overcrowding the same cells (same Angle in the same Treatment) instead of expanding into new ones. A low zero-revenue rate means your grid coverage is working.
Pull the brand's active creative inventory from Motion. The Creative Audit's Surround Sound analysis (Pull 5) tags every active ad across Motion's 8 dimensions, clusters them into the brand's Treatment archetypes, and renders the Angles × Treatments matrix. Cells already filled show what the brand has executed. Empty cells with high-potential combinations (based on the product's revenue rank, the motivator's review frequency, and the format's activation rate) become the allocation targets for new production.
Producer Allocation
The output is a producer allocation recommendation. The math runs in the background.
CTC operates four production sources, Branded Ads, UGC, Dedicated Creator, and Street Interviews, and the brand contributes a fifth. The Audit returns a recommended allocation across these sources for the planning month, calibrated to contracted floors and weighted by recent performance.
Step 1: Current-month scope (the floor)
Pull each producer's current-month scope from the active producer contracts: UGC, Street Interview, Dedicated Creator, Branded Ads, plus any brand-side creative capacity. Use the active ongoing scopes for the planning month only. Sum the scopes across all producers. That is the floor for the planning month.
Step 2: Compare floors to total demand
If the floor total already meets or exceeds the Creative Demand System output, the brand is in maintenance mode. Distribute the contracted volume across the allocation grid and the producer question is answered.
If total demand exceeds the floor, the gap is allocated to the producers with the strongest recent performance. The detailed math runs in the background; the surfaced output is a per-producer ad count.
Hard rule: minimum three production sources active. Brands with fewer than three systematically underperform on Entity ID diversity. Algorithmic diversity requires input diversity. Three sources producing 30 ads each beats one source producing 90.
Pull the current-month scope from the active producer contracts. The Creative Audit's Producer analysis computes the performance weighting and unit economics in the background, then outputs a producer allocation recommendation. The recommendation is what the strategist and client see; the granular activation, outlier, and efficiency data stays internal to the audit's working layer.
The Unified Workflow
Putting the pieces together. The Creative Demand System ties everything above into a single workflow. Five steps, deterministic given the inputs.
Total Demand
Run the Creative Demand formula. Brand-specific 60-day Carry Rate and Expected $/Ad from the Audit.
Moments vs Evergreen
Score the historical calendar through Promo LTV. Back-calculate ad counts from expected incremental spend. Protect Evergreen.
Products
Moments follow the moment. Evergreen: Product Matrix → role assignment → weighted distribution.
Angles × Treatments
Format split from Motion. Write the month's Angles. Multiply across the brand's Treatments. Fill the high-potential empty cells.
Producers
Floors from active producer contracts. Background math distributes the gap. Output is the recommendation.
The output is a production plan: this many ads, featuring these products, in these formats, produced by these sources, launching on these dates. Every number traceable back to the spend plan, the Creative Score, and the brand's performance data.
The Creative Demand formula tells you how many. The allocation system tells you where. And every decision is traceable to a data point, not a preference.
Building This as a Skill
Reference for codifying the methodology into an automated workflow. Each pull feeds a specific step in the allocation sequence; together they produce two outputs.
When this becomes an automated workflow, the analysis for a specific brand follows a defined data-pull sequence using the Statlas MCP or the Motion MCP when more efficient.
Creative Score & five metrics
Source: compass.ad_demand_health. Sets Carry Rate and Expected Spend Per New Ad for the brand.
Spend plan for the planning month
Source: targets table or ad demand forecast.
Marketing events + Promo LTV Scorecard
Planning month plus the brand's full historical promo calendar. Run every historical promo through the Promo LTV Scorecard (NC Lift, aMER, CM%, Post-30d Lift → composite score → T1/T2/T3). Map upcoming moments onto their closest historical comp. Source: Statlas prod.metrics_daily.
Product sales + LTV cohorts (trailing 90 days)
Build the Product Matrix. Score Demand × Margin × LTV × Inventory. Assign product roles. Source: Statlas prod.product_sales (top 500 variants/day, Shopify) and dw.customer_cohorts (30/60/90d LTV).
Active creative inventory from Motion
Pull the brand's active ad inventory from Motion (not Meta's Best Creatives report). Use Motion's 8-tag system. Cluster into the brand's Treatment archetypes. Render the Angles × Treatments matrix and surface the high-potential empty cells.
Current-month producer scope
From the Creator Content dashboard. Filter to Active Ongoing. Toggle across UGC, Street Interview, Dedicated Creator, Branded Ads. Always pull current month. Never historical.
Producer allocation output
Background math: Activation × Outlier × (1/Outlier CPA) per producer for gap weighting; EV/Ad check for unit economics. Surfaced output: a per-producer ad count for the planning month. Granular performance data stays internal to the audit's working layer.
Each pull feeds a specific step. The output is deterministic given the inputs. Different brands get different plans, not because the methodology changes, but because their data changes. The methodology is the same everywhere. The allocation is unique to the brand.
The seven pulls produce two outputs. The first five generate the Creative Audit, a brand-specific HTML diagnostic that shows the evidence behind every allocation decision. The full seven generate the Ad Plan, a production calendar with specific creative assignments by week, product, format, motivator, and producer. The audit is the "why." The plan is the "what, when, and who." Both are derived from the same data, rendered in the same system, and traceable to the same formula.
