Why a blanket ROAS target loses money
The most common target-setting error, and the first of the ecommerce paid ads mistakes we look for in an audit, is a single ROAS number applied across an entire account. A brand decides it wants 4x and holds every campaign, every product, every margin tier to it. It feels disciplined. It is actually arbitrary.
ROAS is only meaningful next to margin. A 3.0 ROAS is comfortably profitable at 45 percent gross margin and loses money at 20 percent. Applying one number across products whose margins run from 15 to 60 percent guarantees you are simultaneously overpaying for the high-margin winners and starving the low-margin lines you could afford to push. The target has to move with the margin, or it is not a target, it is a superstition.
If your ROAS target is the same across products with different margins, it is wrong for all but one of them.
Step 1: find break-even ROAS from margin
Before you can set a profit target, you need the line where the campaign stops losing money. Break-even ROAS is the point where the gross profit from a sale exactly equals the ad cost that produced it.
The formula is simple: break-even ROAS equals 100 divided by your gross margin percent. At 25 percent margin, break-even is 4.0. At 40 percent, it is 2.5. At 50 percent, it is 2.0. Below break-even you are paying more in ad cost than the sale contributes in margin, so every conversion loses money before you have paid a single overhead. This is the floor. Your target lives above it, never on it.
Step 2: add the contribution cushion
Break-even keeps the ad account from losing money on the media alone, but a business is not just media cost. Rent, salaries, software, returns, payment fees, and the profit the business actually needs all sit on top. The contribution cushion is how you build those into the target.
iClick sets the target above break-even by a cushion that reflects overhead and the contribution margin the business needs to hit. As a working default we add roughly a 25 percent contribution cushion on top of break-even, then adjust for the real overhead load. At 40 percent margin, break-even is 2.5 and the target lands near 3.5. At 20 percent margin, break-even is 5.0 and the target climbs toward 6.5. The cushion is where your finance reality, not a benchmark, sets the number.
The margin-tier table iClick runs
This is the table we set on every new account in the first two weeks, then refine with the client's actual overhead. It is a starting framework, not a law of physics, but it beats a blanket number every time.
At 20 percent gross margin: break-even 5.0, target ROAS around 6.5. At 30 percent: break-even 3.3, target around 4.3. At 40 percent: break-even 2.5, target around 3.5. At 50 percent: break-even 2.0, target around 2.8. At 60 percent: break-even 1.7, target around 2.5.
Notice the shape: low-margin products need aggressive ROAS just to survive, so they get less room to spend. High-margin products can profitably tolerate a much lower ROAS, which means they can buy more growth. Segmenting products into these tiers, and bidding each toward its own target, is usually worth more than any bid tweak inside a single blended campaign.
Step 3: blended target vs channel target
The tier table gives you a channel-level ROAS target. But channel ROAS is self-reported and often inflated by attribution, so it should not be the only number you steer by. The business-level counterpart is MER, your total revenue over total marketing spend, and it is the one that reconciles with the P and L.
The practical approach: set channel ROAS targets from the margin tiers to guide bidding inside each platform, and set a blended MER target from the same margin math to check whether the whole engine is profitable. When channel ROAS looks healthy but MER does not, the platforms are over-crediting themselves and the channel targets are lying to you. On accounts above roughly 100,000 dollars a month, iClick's e commerce ppc services team optimises to the MER target first and treats channel ROAS as the diagnostic beneath it.
Step 4: hand the target to Smart Bidding without stalling it
Once you have a target ROAS per tier, target ROAS bidding lets Google chase it automatically. The mistake is handing the algorithm the ambitious number on day one.
A tROAS set well above what a campaign has ever achieved does not produce efficiency, it produces silence: the model throttles spend until only the surest auctions remain, and volume collapses. Anchor the first tROAS at or slightly below the ROAS the campaign already delivers under maximise conversion value, then raise it toward your margin-derived target in 10 to 15 percent steps, giving each change a full conversion cycle to settle. The margin math tells you where the target should end up. The ramp is how you get there without breaking the campaign on the way.
Derive the target from margin. Approach it from trailing performance. Never set the ambitious number cold.
When to deliberately run below target
A margin-derived target is the default, not a cage. There are times to knowingly run below it. New customer acquisition is the big one: if a customer's lifetime value is far above their first order, you can profitably accept a first-order ROAS beneath your product target because the repeat revenue clears the gap. This is where bidding to lifetime value, through value rules or offline conversion imports, earns its keep.
Product launches, inventory clearance, and competitive defence are others. The discipline is that breaking the rule must be a decision with a number attached, a stated payback window or LTV assumption, not a slow drift where the target quietly erodes because someone wanted more volume. Deliberate exceptions are strategy. Undocumented ones are how accounts stop being profitable without anyone noticing.
The five mistakes that break the math
1. Using revenue margin instead of gross margin, which ignores cost of goods and overstates what you can spend.
2. Setting one blanket ROAS across products whose margins genuinely differ, which is wrong for all but one tier.
3. Trusting platform-reported ROAS without validating against Shopify or GA4, the reconciliation step our ecommerce PPC guide covers, so the target is measured against inflated revenue.
4. Handing Smart Bidding the ambitious target cold, which stalls spend instead of lifting efficiency.
5. Never revisiting the target when margins move. A cost-of-goods increase or a pricing change shifts break-even, and a target set last year against last year's margin is quietly out of date. Recompute the tiers whenever the underlying economics move.
