Target ROAS and CPA aren’t optimization settings. They’re business decisions.
Many accounts inherit a target from a previous agency, finance team, or account manager without questioning whether it still makes sense. Set it too aggressively, and you limit volume and lose auctions. Set it too loosely, and you sacrifice profit.
This four-step framework shows you how to calculate a defensible target, test whether it’s realistic, and confirm that your last advertising dollar is still making you money.
Why your target matters more than you think
Two companies sell the same product. One tells their agency to hold ROAS at 800%. The other is fine with 400% because they’d rather take market share than protect margin. Guess who wins more auctions, shows up more often, and slowly takes the category.
As my high school economics teacher would have said, ceteris paribus, the advertiser with the more aggressive target wins. In other words, all else being equal, including conversion rate and Quality Score. In real life, ceteris is rarely paribus. But for this example, we’ll pretend it is.
The first company isn’t being disciplined. It’s being outbid, and it probably doesn’t know it. Somewhere, a target was set, entered into Smart Bidding, and never questioned again.
That doesn’t make the second company the smart one, by the way. Trading margin for market share can be brilliant or reckless, depending on the business behind it. The difference between these two companies isn’t who has the lower target. It’s whether anyone chose it on purpose.
This is the quiet problem with target ROAS and CPA. Most practitioners treat the number as a given. It came from the client, finance, or whatever the account was doing when they inherited it. It rarely came from a calculation they were involved in.
Now that most campaigns use a bid strategy based on target CPA or ROAS, that target is the main lever you still control. Get it wrong in either direction, and you’ll either leave growth on the table or quietly lose money on every sale.
The good news is that the right target isn’t a matter of opinion. You can calculate it two ways:
- From the inside out, starting with your profit margin and how much of it you’re willing to spend to grow.
- From the outside in, sanity-checking against what the auction and your current performance will actually allow.
Once you have your number, there’s one final check most accounts never run: whether your last dollar is still making you money. None of it requires more than arithmetic, and all of it should feed one conversation you ought to have at least once a year with whoever owns the number.
I’ll share every formula along the way so you can rebuild them in a spreadsheet and run your own numbers before you finish reading.
Uncover the keywords, ads, landing pages, and strategies driving your competitors’ paid search success—and find your next opportunity to outperform them.
1. Start with the floor: Your break-even target
Start with the part you can probably do in your head:
- Break-even ROAS = 1 / profit margin
A 40% margin breaks even at 250% ROAS. Below that, every sale costs more than it makes.
CPA gets its own formula because you’re usually buying a lead, not a sale:
- Break-even CPA = Average profit per customer within your payback period x Lead-to-sale conversion rate
The payback period is a decision in itself. Profit over the full customer lifetime is the most generous window, but lifetime value (LTV) can take years to materialize, and few businesses have the cash to prepay for it. That’s why many advertisers count only the profit that arrives within six or 12 months. Pick the window first. The formula is only as honest as that choice.
Say a customer delivers $1,000 in profit within your chosen window, and one in five leads becomes a customer. Your break-even CPA is $1,000 x 20% = $200. Pay more per lead, and you’re underwater.
Nothing new so far, I know. Here’s why the refresher matters: Most accounts feed these formulas the wrong numbers.
The number that belongs in the formula is your effective margin, not your headline gross margin. Effective margin is what’s left after everything it takes to fulfill the order: shipping you subsidize, payment fees, fulfillment, and, in some categories, the quiet killer: returns.
A fashion retailer with a 40% gross margin and a 25% return rate isn’t operating at 40%. After those deductions, it might be in the low 30% range.
That gap isn’t academic. At a 40% effective margin, your break-even ROAS is 250%. Drop the effective margin to 30%, and it climbs to 333%.
Base your target on the headline number, and you’ll end up with one that looks profitable in a spreadsheet but loses money on every order. Nobody notices until finance asks why a “profitable” account isn’t generating cash.
The CPA version of the same mistake is counting lifetime profit when the business actually needs its money back within six months. Same formula, wrong window, same quiet losses.
So step one isn’t the formula. It’s getting the right numbers into it. Agree on the true effective margin and the payback window with whoever owns the P&L before you calculate anything
Dig deeper: How to optimize for ROAS in Google Ads using LTV insights
2. Inside-out: The target your margin can actually support
Break-even tells you where you stop losing money. It says nothing about how much profit you keep. For that, you need one more input, and it’s a business decision, not a math one:
How much of your margin are you willing to spend to win a customer?
My PPC friends and fellow Dutchmen Bob Meijer (no relation, excellent surname) and Miles McNair of PPC Mastery have a three-letter acronym for this number: PAR, or Profit-to-Acquisition Ratio, the percentage of profit reinvested in Google Ads for acquisition. Credit to them for coining it and for finding three letters PPC hadn’t already claimed.
For this article, I’ll simply call it your acquisition share. Spend none of your margin, and you grow nothing. Spend all of it, and you’re back at break-even. The share lives between those poles, and where you set it is the single most consequential number nobody on the account ever discusses.
For ecommerce, the formula folds margin and acquisition share into one line:
- Target ROAS = 1 / (Profit margin x Acquisition share)
Take a 40% margin and decide to spend half of it on acquisition:
- Target ROAS = 1 / (40% x 50%) = 500% (5.0)
So 500% is your target. At break-even, you were at 250%. Spending half your margin to grow doubles the efficiency you demand.
Want to grow faster? Raise the acquisition share to 70%, and the target drops to 357%. You’re accepting thinner returns per sale to win more of them. Want to protect profit? Drop it to 30%, and the target jumps to 833%, fewer sales, and a fatter margin on each.
For lead gen, fold the acquisition share into the break-even formula from earlier:
- Target CPA = Average profit per customer within your payback period x Acquisition share x Lead-to-sale conversion rate
Run the same numbers: $1,000 in profit per customer x 50% acquisition share x 20% lead-to-sale conversion rate = $100. That’s exactly half of the $200 break-even, as you’d expect when you spend half your margin on acquisition.
That lead-to-sale conversion rate does more work in this formula than most lead-gen accounts give it credit for. Halve it from 20% to 10%, and your target CPA halves with it, from $100 to $50, for the exact same customer. The campaign didn’t change. The sales team did. If your CPA target feels impossible, the problem may not be in the account. It may be two desks over, or at your client’s office, wherever the leads you deliver become customers.
For most companies, the profit-maximizing acquisition share sits between 50% and 70%.
That range comes from George Michie‘s math, not my gut. His square root rule puts the theoretical profit peak at spending exactly half your margin, and his studies of real accounts found the peak slightly higher, around 60% to 70%.
But profit maximization, on average, isn’t a law for your business. The right number depends on
- How aggressive the business wants to be.
- How much competitors are willing to spend.
- Whether this quarter is about growth or profit.
That’s exactly why it shouldn’t be set once and forgotten.
Sit down with your client or manager at least once a year and decide it together. Not the ROAS or CPA target itself, but the thinking behind it: the margin, the share of it you’re betting on acquisition, and the trade-off you’re accepting.
Done right, “we need 8x ROAS” stops being a demand handed down from finance and becomes a conversation about how much growth the business is willing to pay for. That single conversation prevents more poorly set targets than any optimization you’ll do all year.
3. Outside-in: Sanity-check the target against the auction
The inside-out target tells you what your business needs. It says nothing about what the auction and your current performance will allow.
I learned this the awkward way in my agency days: A client handed me a target, I ran the math below, and the number that came out was nowhere near the one they walked in with. The arithmetic takes about 30 seconds, and apparently nobody had ever spent them.
The outside-in math needs two or three metrics you either have or can estimate in minutes.
- Achievable ROAS = (Conversion rate x Average order value) / CPC
For CPA, it’s even simpler. Two metrics:
- Achievable CPA = CPC / Conversion rate
There are no weightings and no hidden variables. Each input moves the result one for one: Double your conversion rate or your average order value, and your achievable ROAS doubles. Halve your CPC, and the same thing happens. Fill in actuals from your account, or estimates if you’re sizing up a new market, and the number rolls out.
Say your account shows a $0.80 average CPC, a 2% conversion rate, and a $120 average order value:
- Achievable ROAS = (2% x $120) / $0.80 = 300%
Now put that next to the 500% inside-out target from earlier. The business wants 500%. Reality offers 300%. That’s the health check: The target isn’t realistic right now. Sure, Smart Bidding can technically deliver 500% by retreating to the handful of auctions where the math works, but that’s hitting the target by giving up volume, which is rarely what anyone meant.
By the way, 300% still clears the 250% break-even point. You’re not losing money. You’re keeping less of your margin than the business planned, which is a different, much calmer conversation.
The CPA version works the same way. A $4.50 CPC and a 2.5% conversion rate yield an achievable CPA of $180 ($4.50 / 0.025). Against the $100 target from earlier, it’s not realistic right now either.
Now the two companies from the intro come back into the picture. Your CPC is only partly your metric. The auction sets the price level based on what your competitors are willing to accept.
Run the formula in reverse to see what a target does to your bidding power. With a 2% conversion rate and a $120 average order value, a 400% target lets you pay up to $0.60 per click. An 800% target caps you at $0.30. The company demanding 800% isn’t being outbid by better marketers. It’s being outbid by its own target.
So what do you do when the desired target fails the check? The same formula tells you exactly what passing would take.
To reach 500% on those metrics, you need one of the following:
- Increase your conversion rate from 2% to 3.33%.
- Decrease your CPC from $0.80 to $0.48.
- Increase your average order value from $120 to $200.
- Make smaller improvements across all three.
Now you have something better than a missed target: a shortlist. Improve your landing pages to increase conversion rate, reduce wasted spend, and improve Quality Score to lower CPC, or increase average order value with bundles or free shipping thresholds.
Something on that list can always improve. Just don’t assume it will improve enough. Whatever gap remains is evidence for the target conversation from the inside-out section.
I won’t pretend the renegotiation is fun. Telling your manager, client, or prospect that the desired target isn’t feasible right now makes you the bearer of bad news, and I never learned to enjoy it. What you’re really asking them to accept is a feasible target today, a plan toward the one they want, and the trade-offs in between. But that conversation happens once.
Accept the fantasy target instead, and you’ll have a worse conversation every month explaining why volume keeps falling short of expectations. Awkward now beats awkward forever.
4. The last-dollar check: Where the profit-maximizing target hides
By now, you have a target your margin can defend and a verdict on whether current reality will allow it. Here’s the catch: Every number we’ve calculated describes your average sale. Profit doesn’t happen on average. It happens one incremental dollar at a time.
You already know the principle behind this. Every campaign picks its cheapest conversions first, and Smart Bidding is ruthless about it. Each extra dollar buys slightly worse auctions than the one before: pricier clicks and vaguer queries. So the last dollar you spend always earns less than your average dollar. Blame the law of diminishing returns, not Google.
How much less? You don’t have to guess, and you don’t need anyone else’s benchmark. Google will show you, campaign by campaign.
How to run the last-dollar check in Google Ads
The bid simulator shows what your campaign would have done at different targets over the past week. For each simulated target, it estimates cost and conversions or conversion value.
Google won’t calculate the incremental numbers for you (I wonder why), but they’re hiding in plain sight between any two rows.
Take two target levels and divide the differences between them. For ROAS, the extra conversion value divided by the extra cost equals the incremental ROAS for that step.
From the screenshot above, changing the target ROAS from 500% to 436% would add €7,218 in spend and €19,493 in conversion value. That step earns an incremental ROAS of 270%, just above the 250% break-even point. So it could still be worthwhile.
Once your incremental ROAS or CPA falls below your break-even point, it’s “like feeding money into a shredder,” as George Michie put it.
For CPA, use the same approach in reverse: Divide the extra cost by the extra conversions to calculate the incremental CPA. If raising your target from $100 to $120 adds $3,000 in cost and 12 conversions, those conversions cost $250 each. Against a $200 break-even point, that’s too expensive.
Walk the rows until the incremental number crosses the bar you’ve set. The last step that reaches the break-even point marks your profit-maximizing target. If you want every extra dollar to earn the full inside-out target instead, use that stricter bar. You’ll stop sooner and keep more margin per sale. Either way, your own data determines what’s acceptable.
That completes the health check: a break-even floor, an inside-out target your margin can defend, an outside-in verdict on what reality allows, and a last-dollar reading of where profit actually peaks, based on your own campaigns.
See where competitors are investing, which keywords drive their results, and how to capture more of the market.
Make the target an annual conversation
The full health check takes four steps, one afternoon, and no math beyond arithmetic:
- Step 1: Calculate the break-even floor
- Agree on the effective margin with whoever owns the P&L, then run your break-even ROAS or CPA. Wrong margin in, wrong everything out.
- Step 2: Set the target inside-out
- Decide together how much of that margin buys growth. The acquisition share is the business decision. The formula simply converts it into a target.
- Step 3: Check it outside-in
- Compare the target against your actual (or estimated) CPC, conversion rate, and, for ROAS, average order value. If they can’t produce the number, you need a plan or a renegotiation, not a pep talk.
- Step 4: Check your last dollar
- Open the bid simulators and look at what your next dollar earns. Below break-even, tighten. Far above it, you’re leaving profit on the table.
The target conversation should happen at least once a year with whoever owns the number, whether that’s a client, CFO, or manager. Margins change, competitors change their appetite, sales teams close better or worse, and a target that made sense when someone set it can slowly turn into a number nobody can explain.
The last-dollar check runs on a faster clock. The bid simulator uses the past seven days of data, so it can show you a new curve every week, and it doesn’t require a meeting.
The formulas are what turn the yearly conversation from “we want more” into an actual discussion about trade-offs: this much growth, at this much margin, given this auction.
Remember the two companies from the beginning? The difference between them was never the number. It was that one of them chose theirs on purpose. Be that company.
