top of page
smL Insights cover: bidding for profit, not revenue

Bidding for profit, not revenue: what changes when ads optimise to margin

Writer: smL
smL
Oct 22, 2024
3 min read

Updated: Nov 3

First published 22nd October 2024. Updated 25th October 2026 with current Google Ads guidance.

Return on ad spend is the most common target in paid search and shopping. It is also one of the easiest to misread. Two products can earn the same revenue from the same spend and leave the business with very different profit. Bidding to revenue treats them as equal.

Why revenue ROAS hides the real result

Consider two products that each generate $10,000 in revenue from $2,000 of advertising, a ROAS of 5. One has a 60% gross margin and returns $4,000 after media. The other has a 20% margin and returns nothing. A revenue-based bid strategy will happily push more budget into the second product if it converts slightly better. The dashboard looks healthy while contribution margin falls. The figures are illustrative, but the pattern is common in retail and ecommerce accounts with a mixed range.

How profit-based bidding works

Google Ads offers two practical routes to bid towards profit:

  • Conversions with cart data plus cost of goods sold. Advertisers pass the products in each order and their cost of goods to Google Ads. Google then reports gross profit and, in eligible Performance Max and Shopping campaigns, can optimise towards it

  • Profit as the conversion value. Instead of sending revenue, the business sends a value that reflects margin, either per order through server-side tracking or by uploading adjusted values after the sale. This works with standard value-based bidding and gives the business full control over the calculation

Either route lets automated bidding pursue the outcome the business actually wants. The second also allows other factors to be included, such as returns, new versus existing customers or predicted lifetime value.

What to share and what to protect

Sharing margin data with an advertising platform is a reasonable thing to think carefully about. The data is governed by the platform's data processing terms and businesses should check those terms against their own confidentiality and privacy obligations in each market they operate in.

It is also worth being clear about what is speculation. Earlier commentary, including our own, raised the possibility that profit data could influence platform pricing or be used for purposes such as tax enforcement. We have seen no evidence that this happens. The practical risks are simpler: data errors that send bidding in the wrong direction and over-reliance on a single platform's view of performance.

Businesses that prefer not to share raw cost data can send margin bands or an index rather than exact costs. Bidding still improves because the relative value of each sale is right, even if the absolute numbers are not disclosed.

Before you switch

  1. Fix the data first. Cost of goods must be accurate and current at product level. Bad margin data is worse than none

  2. Test against a control. Run profit-based bidding in a comparable campaign or market and compare contribution margin, not just ROAS

  3. Expect revenue to change. Optimising for profit can reduce revenue while increasing what the business keeps. Agree in advance which number defines success

  4. Look beyond the platform. Check results in your own finance data, because platform-reported profit is still an estimate

Where it does not fit

Profit bidding is most useful where margins vary widely across a range. For single-product businesses, lead generation or subscription models, the better lever is usually passing qualified-lead or lifetime-value signals rather than gross margin.

Revenue is a vanity metric when it grows at the expense of margin. Bidding to profit puts the commercial objective back at the centre of how media is bought.

Sources: Google Ads Help, About conversions with cart data; Google Ads Help, About value-based bidding

 
 
bottom of page