Home / Blog / Strategy
Strategy

Your ROAS looks good, so why is profit not growing?

A commercial audit of the gap between advertising efficiency, customer economics and retained profit.

Sima Damar Beygu
Sima Damar Beygu
Founder, ConversionNest · · 7 min read
Revenue tokens are allocated across cost cards before a smaller retained-value group remains.
Key takeaways
  • Judge scaling decisions using net contribution and payback alongside reported ROAS.
  • Validate product costs, refunds and variable operating expenses before choosing an efficiency target.
  • Use causal evidence to assess additional return rather than adding platform revenue claims.

Strong reported ROAS can coexist with weak profit growth when advertising claims existing demand, revenue excludes later refunds, product margins fall or acquisition costs absorb the remaining contribution. Audit net revenue and variable costs before changing budgets. The commercial test is whether the next spending decision is likely to create acceptable additional contribution.

A high ratio can also accompany shrinking sales. Cutting acquisition and concentrating on people already close to buying may improve reported efficiency while reducing the business's capacity to grow.

Your audit needs to explain the gap between the campaign metric and the business result. It should then identify which commercial lever deserves a controlled change.

Build the bridge from sales to contribution

Start with reconciled revenue under finance's definition. Remove the relevant returns, discounts and variable costs before judging what remains to support advertising and overhead.

A parcel, return arrow, shipping envelope and retained value token form a contribution sequence.
Reconcile the value retained after relevant commercial costs.

A practical management calculation is:

Contribution after media = net sales − product costs − variable fulfilment and transaction costs − media spend.

Agree where each cost belongs. If net sales already reflects discounts and refunds, do not deduct them again. Fixed payroll and overhead still matter to operating profit, even when excluded from the contribution calculation.

Layer

What to inspect

Common mistake

Net sales

Refunds, discounts and cancellations

Treating initial order value as retained revenue

Product cost

Cost by actual product mix

Applying a historical average to a different mix

Variable operations

Delivery subsidy, packaging and payment costs

Ignoring expenses outside the ad account

Media

Relevant spend and currency

Comparing revenue with only part of acquisition cost

Customer mix

New and existing customers

Counting repeat demand as new acquisition

Timing

Mature cohorts and cash collection

Comparing immediate spend with unfinished outcomes

ConversionNest's view: a scaling decision should include the contribution bridge beside platform ROAS. Finance and acquisition should agree the cost definition before debating the target.

Check whether the revenue is economically comparable

Reconcile the reported advertising value with the commercial amount you intend to optimise. Gross order value, net product sales and contribution are different inputs.

A campaign selling a heavily discounted bundle can generate substantial revenue while retaining little value. Another campaign can report less revenue but sell products with healthier economics.

Shopify's profit report documentation explains that its gross-profit reporting depends on product costs recorded at the time of sale. Missing cost data can therefore affect which products appear.

Check coverage before relying on a margin report. A report that excludes products lacking cost records can make the retained sample look healthier than the whole business.

Gross profit is also narrower than your complete contribution calculation. Add the relevant variable operating costs in a separate, agreed layer rather than assuming every expense is already included.

Calculate an economic threshold

A revenue ROAS threshold should follow the contribution available before media, plus the business's payback and profit requirements. It should not be copied from another advertiser's account.

Consider a hypothetical ecommerce business. These figures are hypothetical, not a client result.

For each €100 of net sales, the business spends €45 on products and €20 on variable delivery, packaging and transaction costs. It retains €35 before media, a contribution rate of 35%.

Under those assumptions, revenue ROAS of approximately 2.86 would cover media from that contribution because 1 / 0.35 = 2.86. It would leave no additional contribution towards fixed costs or profit.

This is a simplified management calculation. It assumes the revenue and costs are comparable and does not prove advertising caused those sales. Tax treatment, customer retention and cash constraints may require additional adjustments.

Separate credited revenue from additional revenue

A platform can assign credit to a purchase that would have happened without the advertising intervention. Strong attributed performance is therefore insufficient evidence of the channel's incremental return.

This matters when budgets concentrate on existing customers, branded searches or people already near purchase. Those audiences may still benefit from advertising, but their measured response needs a suitable comparison.

Google's geo Conversion Lift reporting documentation defines iROAS using incremental conversion value divided by incremental cost and includes uncertainty around the result.

A lift estimate is relevant to the tested intervention and conditions. It should not become a permanent adjustment factor applied to every campaign.

Our Conversion Lift analysis explains why causal evidence and attributed campaign feedback support different decisions.

Audit customer mix and retention

Separate new-customer acquisition from existing-customer activity using a documented business definition. Decide how to treat guest purchases, subscriptions, reactivations and incomplete identity.

Then compare acquisition cohorts at similar ages. A recently acquired cohort has had less time to repeat purchase than a mature cohort.

Evaluate whether expected repeat value is supported by observed behaviour. Do not use an optimistic lifetime value estimate to excuse immediate losses without a clear payback plan.

Repeat purchases also carry costs. Include fulfilment, discounts and further marketing required to generate them when estimating future contribution.

A higher returning-customer share can reflect healthy retention, weak new acquisition or both. The ratio alone cannot distinguish those explanations.

Check attribution overlap before adding revenue

If several platforms claim the same order, adding their attributed revenue overstates the total business result. Keep channel attribution views separate from the reconciled sales total.

This does not mean every shared claim is a technical duplicate. Platforms can assign credit under different observation and attribution rules.

Use order and financial records to confirm the commercial outcome, then investigate which evidence supports the spending decision. Avoid dividing each platform's revenue by a generic overlap factor and calling the result incremental.

For reporting discipline, use our guide to metrics that change a decision.

Inspect what happens when you scale

Average historical return does not describe the return from the next increase in spend. More budget may reach less responsive demand, require a different offer or expose operational constraints.

Review changes in net sales, contribution and customer mix alongside the added spend. A comparison between two periods is observational unless the intervention has a credible causal design.

Do not use a simultaneous promotion, price change and budget increase to estimate the effect of advertising alone. The commercial result may be useful, but the contribution of each change remains uncertain.

When a controlled experiment is feasible, define the intervention, business outcome, uncertainty and decision threshold before launch. Confirm that the business can fulfil the resulting demand.

Align the value signal with the decision

Google's conversion value guidance recommends values aligned with business objectives. The implementation still needs consistent amounts, timely delivery and a clearly understood outcome.

If the business changes from revenue values to another economic measure, document the value scale and review the campaign targets accordingly. A historical revenue ROAS target is not directly comparable with a contribution-based value ratio.

Avoid importing speculative customer values as if they were realised revenue. Label predicted values, validate the assumptions and keep financial reporting independent.

Better value inputs can improve the relevance of the campaign objective. They do not guarantee profitable growth or remove the need for commercial validation.

Think we should work together?
Tell us where growth is stuck. If we are the right fit and a slot is open, we start.
Request a consultation

Run a profit-focused media review

ROAS and profit audit

  1. Reconcile the outcome
    Use the agreed net-sales or realised-value definition. Separate mature periods from recent activity still affected by refunds or sales delay.
  2. Rebuild contribution
    Include product costs and the variable operating costs relevant to the decision. Check cost coverage and avoid double deductions.
  3. Segment the economics
    Compare product, market, offer and customer cohorts. Look for changes in mix concealed by the account average.
  4. Inspect the measurement
    Check credited revenue, overlapping platform claims and conversion values. Distinguish reporting changes from real changes in sales.
  5. Choose one commercial intervention
    Test the most plausible constraint, such as an offer, delivery subsidy, product mix or budget allocation. Preserve a useful comparison.
  6. Review the result against the threshold
    Judge net contribution, payback and operational capacity. State whether the evidence is experimental or observational before deciding to expand the change.

What should you change first?

Correct a broken economic calculation before making a media decision. If product costs are missing or refunds are excluded, the apparent target may be wrong.

Fix a weak offer or expensive delivery model when unit economics fail even before media. More precise targeting cannot reliably compensate for every unprofitable product.

If the economics are healthy but new-customer contribution is weak, investigate acquisition quality, the buying journey and incremental return. If demand exceeds capacity, scale in a way the operation can support.

A lower platform ROAS can sometimes accompany a better commercial outcome, but that possibility is not permission to ignore efficiency. Require evidence that the trade-off improves the agreed business result.

Leave the meeting with a decision rule

Record the metric, threshold, review period and owner for the next change. Include the conditions that would stop the test or limit spending.

A business prioritising near-term cash needs a different payback rule from one intentionally funding longer-term acquisition. Make that choice explicit rather than hiding it inside a universal ROAS target.

Keep reported ROAS as an execution measure. Pair it with the contribution and causal evidence needed for the budget decision.

Frequently asked questions

Can a high ROAS campaign lose money?

Yes. Low margins, refunds and variable costs can consume the reported revenue. Advertising may also receive credit for purchases that would have occurred anyway.

Is gross profit the same as contribution after advertising?

No. Your contribution calculation should specify which variable costs and media expenses it includes. Gross-profit reports may exclude several of those costs.

Should I optimise every campaign to profit values?

Only use a value definition you can calculate, deliver and validate consistently. A change in value scale also requires a review of targets and reporting.

Need a hand with this?Digital StrategyWhere you stand, who you serve, and what winning looks like, decided before the first budget goes out. We set the KPIs that every later decision answers to.
Free audit, powered by Nestor
See what your page is costing you.
Nestor is the AI review engine we built, and the one we use on client work. Enter your URL and email, and in a few minutes read the report: scores, the findings that matter, and what to fix first.
Free. No account, no card, no sales call.
Sima Damar Beygu

Sima Damar Beygu

Founder of ConversionNest. 10+ years in growth marketing, managing 300K euro monthly media budgets and scaling acquisition across 15+ markets. MSc in Digital Marketing from University College Dublin. Google and Meta certified.

Sima on LinkedIn
Keep reading