Happy Sunday, Everyone!

I hope you’re all enjoying the first hours of daylight savings time (which, we can all agree, is the best time).

A few weeks ago, I was reviewing a Meta account for a brand spending $40,000 a month. Their agency had told them they needed more creative. It was the same feedback they heard month after month. In the interim, they hired a UGC shop, added a designer and 3x’d the number of assets they were adding to the account. Everyone on their (admittedly small) team was cranking out content, from the owner/founder and his wife to the part-time guy they had helping them with fulfillment.

The brand owner’s frustration when he and I spoke was palpable: “it doesn’t matter how much more creative we make, these guys just say we need more. How much is enough?”

All the while, CAC was climbing – both in the ad account and in their 3rd party attribution tool. He was - understandably - at his wit's end.

Once I opened up the account, the problem was obvious: this wasn’t a creative volume issue. CPMr (cost per 1,000 unique accounts reached) was at $64 over the L30. That’s exceedingly high for most verticals, and this was no exception. Meta had long since stopped finding new people. It was hammering the same audiences at increasing frequency, running up the cost of reaching people who’d already decided whether or not to buy. Every new creative they launched was effectively more meat into the grinder.

Producing more was the equivalent of running faster on a treadmill and wondering why the scenery isn’t changing. To be fair, their agency wasn’t wrong on the importance of creative - but they were wrong about which problem needed solving first.

This is where the advertising industry gets it wrong every time. Everyone has an opinion on how many creatives you should have in an account/ad set. Almost nobody has a framework for figuring out why you need that many, or whether the number they’re giving you is the answer to the question you’re actually asking.

The Industry’s Favorite Non-Answer

If you’ve been at any marketing conference in the past 2-3 years, you’ve heard some version of the same argument: You don’t have enough creatives. You need more creatives. You need more diverse creatives. You need more. More makes everything better.

Stop me if you’ve heard this before.

The more operationally serious version of this argument has started to take shape in some agency circles: demand planning models that connect spend forecasts to monthly creative counts. It’s a meaningful step beyond gut feel. Credit where it's due: moving from 'more' to 'here's how many and why' is objectively the right path forward for our industry.

But even those well-meaning attempts are falling short. In almost every case, the math behind them is proprietary. The mechanism (how it works) is unexplained. The output is a single number regardless of what’s actually constraining your account. I’ve seen these models in action. A brand at $70K monthly spend with a 20% ROAS decline is told they need to produce 32 ads in January. They launched 15. Performance declined. The implied conclusion is that creative volume was the problem.

Maybe it was. Maybe it wasn’t. The model can’t tell you. It just tells you that you should produce a certain number of ads based on your planned investment in the channel.

This is further complicated by the environment in which we are all operating: CPMs are rising (which means CPMr is rising). The broader macro-economic conditions in much of the US and EU are unstable. Hit rates fluctuate wildly across industries + brands, but 10% - 20% is the norm. Put those together, and the prescription “produce more” is as likely to accelerate waste as produce results.

A model that doesn’t have a way to parse through that complexity to arrive at a number you can trust isn’t one I’m interested in using - so I built a formula myself and am sharing it here.

How Creative Volume Actually Works

The first thing to understand is that creative volume is not a single variable. It’s the output of 4 independent constraints, any one of which can be the limiting factor in your account.

The formula:

CV_final = max(F’, P, S_floor, S_ceiling) / H × R_eff

You take the maximum of 4 constraint values (not the average, not a blend). Whichever constraint is tightest is the “bottleneck” - and therefore governs the output. Then you divide by hit rate and adjust for reach. The result is a number you can actually understand and defend, because it’s built from your business + your account, not generated by someone else's model.

Ready? Let’s get to it.

F’ — How Fast Your Creatives Die, Adjusted for How Your Customers Buy

F’ = Active ad slots × (30 / L_adj)

L_adj = Raw lifespan / (Median days to conversion / [baseline consideration cycle = 7]

Every creative has a measurable lifespan. What most people miss is that lifespan is relative, not absolute. It doesn’t matter how long the creative “lasts” - the only thing that matters is how long it lasts relative to your audience’s consideration cycle.

An ad doesn't fatigue in a vacuum; it fatigues relative to the conversion cycle of the business. For organizations with a short consideration cycle, a given creative is seen a small number of times before a purchase decision is made; the audience moves through the funnel quickly and the creative retains its novelty across most of that journey.

Conversely, when the consideration cycle is long, the same creative is being served to the same unconverted prospect across days or weeks of exposure before they ever take action. Frequency accumulates against a decision that hasn't been made yet. By the time they convert (or don't) the creative has already exhausted its relevance to everyone who was in the market during that window.

The raw lifespan tells you how long a creative survives by platform metrics; the adjusted lifespan tells you how long it survives relative to the actual conversion behavior of your audience. Those are not the same number, and in high-consideration categories, the gap between them is where many agencies/media buyers go wrong.

For example: a supplement brand with a 3-day consideration cycle and an 18-day raw creative lifespan gets an adjusted lifespan of 42 days. A senior living operator with a 28-day consideration cycle and a 32-day raw lifespan gets an adjusted lifespan of 8 days. Same creative quality. 5x the creative demand, simply because the business has a radically different sales cycle.

Most operators running senior living accounts at $30K/month think they need modest creative production. But, when you actually calculate this variable, the result is surprising: they’re in one of the most creative-hungry environments in performance marketing. Not because of competition or CPMs, but because of how their customers make decisions.

One important note: the formula uses Meta's default 7-day click attribution window as the baseline. If your CRM shows a meaningfully different median conversion cycle for your specific product or service, substitute that number. The formula is only as honest as the inputs you feed it.

P — The Minimum Floor Required to Learn Anything Useful

P = N × M × T × R

(Personas × Angles × Asset types × Variation multiplier)

3 personas, 3 messaging angles per persona, 2 asset types and a 1.5x variation multiplier: your floor is 27 creatives before you've tested anything. Below that number, your account lacks the coverage to isolate which variable is actually driving performance. You may get results. You won't know why.

One thing to keep in mind before you open up your Excel spreadsheet: the formula treats each variable as independent, but in practice not every angle translates across every asset type. A long-form VSL angle doesn't cleanly translate into a static image. A UGC testimonial doesn't work as a carousel. Roughly 20% of your theoretical N × M × T combinations will be structurally incompatible. The formula gives you the coverage floor. Apply light judgment about which combinations are actually viable before proceeding.

The R multiplier is where most teams undercount and where the formula's output changes most dramatically. Testing one execution of a message angle tells you nothing about whether the angle works. It tells you whether that specific execution, with that specific hook, performed on that specific day in that specific environment.

R of 1.5 is the budget-constrained floor - the minimum required to start separating signal from execution variance. R of 2 is where you actually begin generating a reliable signal, as every angle has at least 2 distinct executions and you can start attributing performance to the angle rather than the execution. The gap between 1.5 and 2 looks small, but is often the difference between data and insight. One of the best things you can do for your account is to ensure that you test at least 2 expressions of each creative concept before ruling it out AND methodically revisit your previous “losers”.

I have Meta accounts where a given creative concept performed horribly 8 months ago, but when we re-launched it last month, it went gangbusters.

S_ceiling — How Many Creatives Your Budget Can Actually Feed

S_ceiling answers a single account-level question: given your budget and CPM environment, how many creatives can this account absorb before signal starts fragmenting? It is not a per-ad-set prescription. How you distribute creatives across ad sets is a separate structural decision: one governed by your campaign structure, audience segmentation and margin priorities. S_ceiling just tells you the outer boundary of what the account can productively support in total.

S_ceiling = Monthly budget / (CPM × 10 × Ad sets)

Every platform needs a minimum impression threshold per creative to accumulate usable data. Call it 10,000 impressions as a working proxy. This is NOT a platform specification, but a practical approximation of the point at which meaningful patterns begin to emerge. At $18 CPM across 8 ad sets on $75K/month, the account-level ceiling is roughly 52 unique creatives. Go above it and you're not producing more winners; you're fragmenting your budget across too many creatives, which means individual ads can’t accumulate the optimization event volume required to move from exploration to exploitation. The account ends up knowing a little about a lot of creatives and a lot about none of them.

One important distinction: S_ceiling and S_floor (below) are related but measuring different things. S_floor asks whether each individual creative is funded to minimum conversion volume. S_ceiling asks whether the total creative count is compatible with the account's budget at the macro level. An account can satisfy S_floor on every individual creative and still violate S_ceiling if the total number of creatives is too high for the budget to support coherently. Both constraints matter. Neither substitutes for the other.

S_floor — The Constraint Most Agencies Will Never Mention

S_floor = Monthly budget / (CPA target × Min conversions / 30)

Each creative needs a minimum budget to generate the conversion volume Meta requires to do something useful. An ad set with a $150 CPA target and a $50/day budget can never exit learning, simply because the data never reaches exit velocity. Think of it like launching a rocket into space: in order for a rocket of a certain size (that’s your CPA) to reach orbit (that’s the number of optimization events for Meta to be effective), you need a certain amount of fuel (that’s your budget). If you don’t have a sufficient quantity of fuel for the size of your rocket, it’s not making it to orbit.

The same logic applies at the creative level. If you’re producing 50 creatives on a budget that can only fund enough optimization events to support 20, then you’ve hindered the performance of every creative in the account (including your winners/hits).

The floor tells you the maximum number your budget can legitimately support. Produce above it and you’re self-sabotaging.

H — Hit Rate a/k/a Variable That Exposes Everything

H = Creatives reaching min. viable performance / Total launched

Hit rate is where the formula stops being abstract and starts being uncomfortable. It’s a direct measure of your creative process.

If your hit rate is 20%, you need to produce 5 creatives for every 1 that earns sustained rotation. At 40%, you need half as many. The agencies telling you to produce more without asking about your hit rate are selling you inputs while ignoring the yield on those inputs. You wouldn’t evaluate a sales team by the number of calls they make without looking at close rate. Same logic.

What drives hit rate? Brief quality. Quality/Quantity of audience insight. The rigor of your pre-production hypothesis. The discipline to craft ads for specific audience segments with a specific emotional state. Accounts with 15%+ hit rates aren’t luckier than accounts at 6%.

H is also where the formula makes the art/science boundary concrete. Science gives you the equation; art determines whether H is 0.05 or 0.25.

R_eff — The Formula’s Way of Telling You When Creative Is the Wrong Solution

R_eff = Target CPMr / Current CPMr

When R_eff is less than 1 (when your current CPMr exceeds target), the formula inflates the required creative volume. That inflation is not a production recommendation. It’s a warning. It’s the formula communicating that adding creative into a degraded reach environment is the wrong intervention sequence.

The brand I opened with had an R_eff of 0.63. Dividing by 0.63 increased their creative required by 59%. That’s not a reason to produce 59% more creative. That’s a reason to ask why CPMr is running 37% above target before spending another dollar on production.

Fix the reach problem first and the creative requirements fall back into an acceptable range.

If you don't have a historical CPMr baseline for your account, a reasonable starting point by vertical: DTC and eComm accounts often run $12–$28; lead gen and B2C services $15–$31; high-consideration and B2B $18–$42. Anything materially above those ranges warrants investigation before you touch creative volume. The brand I opened with was at $64. As you can see, that’s way beyond reasonable for almost any industry/vertical.

Does This Actually Work?

I ran the formula through 8 accounts ($15K to $250K monthly spend) across DTC, SaaS, senior living, home services and legal services to test whether the outputs were coherent across different business types.

The most important finding has nothing to do with any specific account: 2 accounts at $50K/month produced opposite diagnoses.

The SaaS lead-gen account was constrained by S_floor. Its ad sets were structurally underfunded relative to the CPA target. Adding creative would have made things worse. The solution was consolidation: fewer, properly funded ad sets.

The legal account was constrained by H and R_eff simultaneously. CPMr was badly elevated, hit rate was 3%, and the CPA target was $350. The formula’s output was technically accurate, but practically impossible to satisfy. The right answer for that account was to stop producing new creative until reach was repaired and they were able to develop new audience insights/angles.

That’s exactly what the model should do: recommend different interventions for different accounts with the same spend level. A model that gives both accounts the same output (“produce more creative”) is not a strategy. It’s an easily defensible “best practice” thing to say that happens to be right roughly half the time.

The senior living account was the one that surprised me most. At $30K/month, most agencies/operators assume modest creative needs - maybe 3 to 5 new assets per month. But, this industry has a LONG consideration cycle. That has the effect of compressing adjusted lifespan to 8 days, which, in turn produced a recommended creative output well above that assumption.

In this case, the business model itself was creating the demand, regardless of what's happening in the auction or with your competitors. That's something a spend-based model simply can’t identify.

The Allocation Problem

Getting the total creative count right is half the challenge; the other (arguably more important) is the underlying distribution of that creative.

Most brands distribute creative production proportionally to revenue: the product driving 60% of revenue gets 60% of the creative budget. This is the peanut butter problem applied one level up. Revenue share is not the right numerator.

Margin-adjusted revenue is.

Alloc_i = (Revenue share_i × Gross margin_i) / Σ(Revenue_j × Margin_j)

For the direct-to-consumer home goods account in the stress test (3 product lines at 48%, 72%, and 22% margin), margin-weighted allocation pulled production dollars away from the low-margin line and concentrated them in the 72% margin accessories. You can run an account at 3.2x ROAS and lose money because the creative is scaling the wrong products. This formula ensures that doesn’t happen by design.

Here’s what I’ve been thinking about since running this stress test: the formula is not actually about creative volume.

Creative volume is just the output. What the formula is really doing is forcing you to answer 6 questions your account probably can’t answer right now:

1. What is your creative lifespan, adjusted for your consideration cycle?

2. What is your hit rate over the last 90 days?

3. What is your CPMr relative to target/baseline?

4. How many genuinely distinct personas are you running creative against?

5. How much budget is available to fund each creative?

6. What is the margin profile of the products/services your creative is actually scaling?

If you don’t know the answers to those questions, you don’t have a creative volume problem. You have a measurement problem. Producing more creative without those answers is as likely to accelerate waste as it is performance.

The agencies and frameworks telling you to produce more without asking these questions first aren’t wrong in saying creative matters. It does. They’re wrong about the order of operations. Measurement comes first. The formula comes after. And the measurement (uncomfortable, occasionally humbling, always clarifying) is the part everyone skips.

Here's the uncomfortable version of everything I've laid out above: if you run this formula on your account and are surprised at the result, that is the finding. It means something in your account has been misdiagnosed, probably for longer than you'd like to admit and at real cost. The formula didn't create that problem. It just made it visible. What you do with that knowledge is the only part that's actually up to you.

Until next week,

Sam

P.S. Whenever you’re ready, here’s how I can help: if your brand needs a strategic partner that blends performance marketing, analytics, and brand into one integrated team — not five siloed agencies — reach out or reply to this email.

Loving The Digital Download?
Share this Newsletter with a friend by visiting my public feed.

Follow Me on my Socials

​​​

1700 South Road, Baltimore, MD 21209 | 410-367-2700
​Unsubscribe | Manage Preferences​