Our POV
The Agency Incentive Problem.
Why your agency can't tell you the truth - and why switching agencies never fixed it.
What the Agency Incentive Problem is
The Agency Incentive Problem is the gap between what an agency is paid to manage - the inside of your ad accounts - and the numbers that decide whether your brand can scale, which live outside them. It is not a bad-agency problem. It is the agency model working as designed.
They're not hiding the truth from you. They're structurally locked out of it themselves.
The retainer model pays them to manage what's inside your ad accounts. The platforms only surface the metrics the platforms can move. Real CAC by cohort, contribution margin by product, true blended MER - those numbers don't live in Meta Ads Manager. They live in your finance system, your Shopify, the spreadsheet you stopped trusting six months ago.
Why both sides end up steering by platform ROAS
So both sides settle on platform ROAS, because that's the only number the contract can defend and the tool can produce. You've been paying for performance that exists inside an ad account while the business it's supposed to grow stays stuck at the same number.
That's not a bad agency. That's the Agency Incentive Problem - the agency model working exactly as designed.
Why switching agencies never fixed it
It's why you ended up feeding your own agency the ideas, the concepts, the audiences to test - and got back 'okay, great.' That was never supposed to be your job. Which is also why firing your agency and hiring a better one never fixed it. You changed the people and kept the model. If you are asking whether yours is worth keeping at all, take the 9-signal fire-your-agency test first. And if you are weighing what comes after the agency, compare the models: fractional CMO vs growth agency vs advisory, or the Meta ads agency alternative.
What the gap costs every quarter
In the accounts we've audited, Google Ads typically overstates its contribution by 80-90% - and Meta typically understates by 20-80%. Every scaling decision made on those numbers is a guess. The mechanics behind the gap - attribution windows, view-through, brand capture - are in why your ad platforms all show different revenue. At $30K/month - our minimum - that's $90,000 of spend allocated by guesswork every quarter. At $100K/month, it's $300,000. A flat quarter isn't neutral. It's what that guesswork costs.
"We spent an absolute ton of money but the overall profitability of e-commerce absolutely tanked."
"They were just putting out such generic stuff that I felt ashamed to even have my name behind it. I'm a doctor. I shouldn't be better than my graphic designer at creating creative."
"The previous partner was deploying a lot of smoke and mirror tactics. Really focused on returning consumers... just not great tactics if you're looking to actually grow a brand."
"They cannot seem to separate this new customer base with the already aware customers."
What we built instead
The Profit Clarity System.
We stopped competing with agencies and built the thing the model can't: the metrics layer outside the platforms, and a growth practice that runs on it. Three pillars, installed once, yours to keep.
The Outside-Platform Layer
We rebuild attribution, cohort LTV, contribution margin, and true blended MER outside Meta and Google - including how much of your reported ROAS is branded search and retargeting taking credit for customers you already had. Live in your own dashboards. Defensible to your accountant, your investors, your future acquirer.
The 5-Level Constraint Ladder
We diagnose where your real bottleneck lives - Product, Offer, Ads, CRO, or Retention. Most agencies pitch the same fix to every brand. The diagnostic determines which fix applies to yours. The wrong one wastes 6 months.
The 1x-Above-Breakeven Scaling Rule
The operating discipline that stops the most expensive mistake in DTC scaling: spending into negative incremental margin because platform ROAS looks fine. Shows you when to scale, hold, or cut - and how much you can spend, at what CAC, before growth stops being profit.
Fair questions
What founders ask about the incentive problem.
No. They are structurally locked out of them themselves. They are paid to manage the inside of your ad accounts, and the numbers that decide whether your brand can scale live outside them. It is why you ended up feeding your own agency the ideas, the concepts and the audiences to test. You got back 'okay, great.' That was never supposed to be your job. That is not a bad agency. That is the agency model working exactly as designed. The 9-signal fire-your-agency test has a check for whether the ideas only ever run one way.
Platform ROAS is the only number the contract can defend and the tool can produce. Contribution margin by product and true blended MER do not live in Meta Ads Manager. They live in your finance system and your Shopify. So the report covers the part of your business the account can see. Set each platform's claim for last month against what Shopify actually recorded, one platform at a time. Google's will usually run high and Meta's low, so a combined total that happens to reconcile is two errors cancelling out. That reconciliation check is published in the 9-signal fire-your-agency test.
At $30K a month in ad spend - our minimum - that is $90,000 of spend allocated by guesswork every quarter. At $100K a month, it is $300,000. In the accounts we've audited, Google Ads typically overstates its contribution by 80-90% and Meta typically understates by 20-80%. Every scaling decision made on those numbers is a guess. The cost arrives as a flat quarter, not a line item. To see how much of your reported ROAS is warm traffic, pull your prospecting campaigns out on their own - cold audiences only, no retargeting or brand terms - and compare that ROAS with the account average your report leads with. A large gap means warm traffic is carrying the number.
Because you changed the people and kept the model. The next agency is paid for the same thing: managing what is inside your ad accounts. The numbers that decide whether your brand can scale still live outside them. So you get the same platform ROAS reports from a new logo. The execution is fresh. The numbers steering it are not. You keep paying for performance that exists inside an ad account, while the business it is supposed to grow stays stuck at the same number.
In this series.
The same argument from three more angles - each one stands on its own.
Should you fire your Meta ads agency?
Nine signals you can verify yourself, and a hold / investigate / act rubric.
The Meta ads agency alternative
Why the next agency inherits your broken numbers, and the third path.
Growth advisory vs agency vs fractional CMO
What you are actually buying with each model, and when each is the right hire.
Start with the free version of the discipline
The E-Commerce Scaling Scorecard is the 5-metric decision tool we run before every budget move on managed accounts: CM3, nCAC, 90-day LTGP:CAC, payback, and revenue split - with the exact scale/hold/fix/kill thresholds. Free, by email.