5 Signs It's Time to Fire Your Ad Agency
February 12, 2026
Your ROAS is 4x. Your marketing team is celebrating. Your agency just sent a report showing strong month-over-month improvement. Everything looks great.
Except you checked your bank account, and the profit is not there.
This is one of the most common and most dangerous disconnects in paid advertising. The dashboard says you are winning. Your bank account tells a different story. And the gap between those two realities can quietly drain a business for months before anyone notices.
ROAS (Return on Ad Spend) is the most popular metric in paid advertising. It is also one of the most misleading. Not because it is wrong, but because it tells a partial story that many businesses mistake for the full picture.
This guide will show you why ROAS alone is not enough, what metrics you should be tracking alongside it, and how to build a measurement system that tells you the truth about whether your advertising is actually generating profit.
Let us be clear: ROAS is not a bad metric. It is useful for comparing campaigns, evaluating creative performance, and making daily optimization decisions. The problem is when it becomes the only metric, or worse, when business decisions are made based on ROAS alone.
ROAS Ignores Your Costs
A 4x ROAS means you generate $4 in revenue for every $1 in ad spend. That sounds profitable. But is it?
If your product costs $50 and you sell it for $100, your gross margin is 50%. At a 4x ROAS, you are spending $25 to generate $100 in revenue. After the $50 cost of goods, you have $25 left. Subtract the $25 ad spend and you break even. A 4x ROAS, and you made zero profit.
Now factor in shipping, transaction fees, overhead, and the agency fee or media buyer cost. You are actually losing money at 4x ROAS.
ROAS tells you nothing about margins, cost of goods, overhead, or any of the other expenses that determine whether revenue translates to profit. Two businesses with identical ROAS can have completely different profitability depending on their cost structure.
ROAS Double-Counts Returning Customers
Here's where it gets worse. Your 4x ROAS likely includes revenue from returning customers who were going to buy anyway.
Facebook's attribution model gives credit to ads that were seen or clicked before a purchase. If a loyal customer who buys from you every month happened to see a retargeting ad before their latest purchase, that sale gets attributed to the ad. Your ROAS goes up, but the ad didn't "really" create that sale. It just took credit for it.
We've audited accounts where 40% to 60% of attributed revenue came from returning customers. The headline ROAS was 5x. The new customer ROAS was closer to 1.5x. The business thought their ads were printing money. In reality, the ads were barely acquiring new customers at a sustainable cost, while the retargeting campaigns were getting credit for organic repeat purchases.
ROAS Does Not Account for Attribution Gaps
Post-iOS 14, every ad platform under-reports conversions to some degree. Facebook might only track 60% to 80% of actual conversions, depending on your setup. This means your real ROAS might be higher than reported.
But it cuts both ways. Some conversions Facebook takes credit for were actually driven by other channels, by email, organic search, word of mouth, or simply by the customer deciding on their own. Without robust measurement beyond the platform, you can't know the true incremental impact of your ads.
If ROAS is not enough, what should you track? Here are the five metrics that will give you a more honest picture of advertising performance.
Marketing Efficiency Ratio (MER)
MER is total revenue divided by total marketing spend. It is the simplest and most honest top-level metric because it does not rely on platform attribution at all.
If your business generated $500,000 in revenue last month and spent $100,000 on all marketing (ads, agency fees, creative production, software), your MER is 5.0. That number tells you the overall efficiency of your marketing investment, regardless of which channel gets the attribution credit.
MER is especially useful for catching situations where platform-reported ROAS looks great but the business is not actually growing. If your Meta ROAS is improving but MER is declining, something is off. You might be cannibalizing organic sales, or your attribution model might be over-crediting certain campaigns.
Track MER weekly. It is the number that keeps everything else honest.
Blended Customer Acquisition Cost (CAC)
Blended CAC is total marketing spend divided by total new customers acquired. It tells you what you are actually paying to bring a new person through the door, regardless of which platform claims credit.
This matters because new customers are the engine of growth. If you are spending heavily on ads but most of the "conversions" are returning customers, your blended CAC will reveal the problem. It is the metric that answers the question: is my marketing actually growing the business, or just making it look like it is?
For a more granular view, calculate CAC by channel. But always anchor to blended CAC as your north star, because it accounts for the messy reality of multi-touch attribution.
Contribution Margin After Advertising
Contribution margin after advertising is revenue minus COGS minus ad spend. It is the amount of money your advertising actually contributes to the business after covering both product costs and the ads that drove the sale.
This is the metric that tells you whether your ads are profitable in absolute terms, not ratios. A campaign with a lower ROAS but higher contribution margin is more valuable than one with a higher ROAS but lower contribution margin. Ratios can mislead. Dollars do not.
Calculate this at the campaign level when possible. Some campaigns might have mediocre ROAS but contribute significant margin because they sell high-margin products. Others might have impressive ROAS but contribute little because the margins are thin.
Payback Period
Payback period is how long it takes to recoup the cost of acquiring a customer. If you spend $100 to acquire a customer who spends $50 on their first purchase (with $25 in gross margin), your initial payback is negative. But if that customer makes a second purchase within 60 days, you have recouped your acquisition cost.
This metric is critical for businesses with repeat purchase dynamics (subscription, consumable products, service businesses). A negative first-purchase ROAS might be completely fine if your payback period is 60 to 90 days and you have the cash flow to support it.
The businesses that scale fastest are often the ones willing to accept a longer payback period because they understand the lifetime value of their customers. But this only works if you actually measure payback and have the data to back the decision.
New Customer ROAS vs. Returning Customer ROAS
Separate your ROAS reporting into new customers and returning customers. This one change will give you more clarity than any other adjustment to your measurement.
New customer ROAS is almost always lower than blended ROAS. That is expected. Acquiring someone who has never heard of you costs more than re-engaging someone who already knows and trusts your brand.
The question is whether your new customer ROAS is sustainable given your margins and payback period. If it is, you have a growth engine. If it is not, your headline ROAS is being propped up by returning customers, and your growth is more fragile than it looks.
You do not need expensive analytics platforms to get started. Here is a practical approach that works for most businesses.
First, get your tracking right. Implement Conversions API alongside the pixel. Make sure your events are firing correctly and your attribution settings match your actual business model. This is foundational, and most businesses have gaps here that distort every metric downstream.
Second, build a weekly dashboard that includes MER, blended CAC, and platform ROAS side by side. When they move in the same direction, you have confidence in your data. When they diverge, you have a signal that something needs investigation.
Third, segment your reporting. Break down ROAS by new versus returning customers. Break down CAC by channel. Break down contribution margin by product line. The aggregates hide important stories that only appear when you look at the components.
Fourth, establish benchmarks for your business. A "good" ROAS depends entirely on your cost structure, your LTV, and your cash flow. Calculate your break-even ROAS (1 divided by your profit margin) and set your target above that. Every business is different, so generic benchmarks are only useful as rough references.
The businesses that scale most effectively are the ones that stop optimizing for a ratio and start optimizing for profit.
This sounds obvious, but it requires genuine discipline. When your agency shows you a beautiful ROAS number, your first question should be: what does that mean for actual profit? When a campaign has a lower ROAS but higher total contribution margin, can you resist the urge to cut it?
The best advertisers have made this shift. They care about ROAS as one signal among many. They make decisions based on the full picture: MER, CAC, contribution margin, payback period, and cash flow. They are willing to accept lower ROAS if it means higher total profit.
This is the difference between managing ads and managing a business. ROAS is a tool. Profitability is the goal.
If you want help building a measurement framework that shows you the truth about your advertising performance, we are here. Our talent has managed hundreds of millions in ad spend and knows how to look past the dashboard to the metrics that actually drive profitable growth.
.png)