Blog · Finance metrics and formulas
Average marketing ROI is useful only when campaigns share the same profit definition, cost boundary and observation window. Work through a campaign comparison, prove the weighted return and separate attributed contribution from incremental profit.
Average marketing ROI is comparable only when the profit definition, included costs, attribution rule and observation window match. Start with your own completed campaign cohorts, then compare the return on the combined investment. An outside average with different boundaries is a different measure, even if both reports call it ROI.
For this comparison, contribution before marketing means attributed net revenue less the variable product, service and fulfillment costs of earning it. Marketing cost includes campaign delivery spend and the other campaign costs finance agrees to include. Contribution after marketing is what remains after deducting that marketing cost once.
Portfolio marketing ROI = total contribution after marketing / total marketing cost
This is an attribution-based return on marketing investment. It is not a claim that the campaigns caused all the assigned sales. The existing marketing ROI versus ROAS guide owns the detailed formula comparison; the question here is which results belong in the same average.
Cost conventions matter even in official examples. The Google Ads ROI example includes both production and advertising costs in its denominator. That differs from dividing contribution after marketing by marketing cost alone. Neither percentage can be compared fairly without stating the cost base.
Start with the export your marketing team already uses to list campaign IDs, delivery dates, audiences and assigned outcomes. Add invoice IDs, invoice dates, net revenue, variable costs and campaign spending from finance. Keep returns, credits and sales tax treatment consistent across the compared rows.
Use a stable outcome allocation rule so one invoice is not fully credited to several campaigns. Keep unassigned revenue visible. Reconcile the campaign delivery spend to its source export and the additional costs to the marketing ledger; do not silently discard costs that lack a campaign ID.
Record the following alongside the amounts:
| Comparison field | What the reviewer needs |
|---|---|
| Cohort entry | The event that starts observation, such as a first qualifying response |
| Observation cutoff | The same elapsed time after entry for each included outcome |
| Financial outcome | Net invoiced revenue, collected cash or another explicitly named basis |
| Cost boundary | Included variable costs and marketing costs, with no overlap |
| Attribution basis | The rule allocating an outcome and its contribution across campaigns |
| Coverage | Which campaigns and outcomes are included, excluded or still incomplete |
A channel report's conversion setting is also part of the comparison. Google's conversion-window documentation defines the window as the period after an ad interaction during which a conversion is recorded. That platform window and your finance cohort's observation period need to be reconciled, rather than assumed to be identical.
Consider a hypothetical US industrial supplier comparing three completed acquisition campaigns. Each outcome cohort has been observed for 90 days after its first qualifying response. Outcomes are allocated under one agreed attribution rule, with no invoice counted twice; the figures below are invented for the calculation, not external benchmarks.
Delivery spend means media, postage or event execution as applicable. Other marketing costs include the assigned creative, agency and campaign labor costs. Neither marketing column is included in variable product and fulfillment cost.
| Campaign | Net revenue (USD) | Variable cost (USD) | Delivery spend (USD) | Other marketing cost (USD) |
|---|---|---|---|---|
| Search | 200,000 | 120,000 | 40,000 | 10,000 |
| Direct mail | 60,000 | 30,000 | 8,000 | 2,000 |
| Trade event | 100,000 | 65,000 | 20,000 | 5,000 |
| Total | 360,000 | 215,000 | 68,000 | 17,000 |
Subtract variable cost from net revenue, add the marketing cost columns, then calculate the remaining contribution and return:
| Campaign | Contribution before marketing (USD) | Marketing cost (USD) | Contribution after marketing (USD) | ROI |
|---|---|---|---|---|
| Search | 80,000 | 50,000 | 30,000 | 60.0% |
| Direct mail | 30,000 | 10,000 | 20,000 | 200.0% |
| Trade event | 35,000 | 25,000 | 10,000 | 40.0% |
| Portfolio | 145,000 | 85,000 | 60,000 | 70.6% |
For search, $200,000 minus $120,000 leaves $80,000 before marketing. Subtract $50,000 of marketing cost and divide the $30,000 remainder by that investment: 60.0%. Portfolio ROI is $60,000 divided by $85,000, or 70.6%, rounded to one decimal place.
The simple average is (60% + 200% + 40%) / 3 = 100.0%. That describes an equally weighted campaign, allowing the smaller direct-mail investment as much influence as search. It does not describe the return on the dollars invested. The count-weighted and value-weighted guide explains that distinction across business ratios.
The portfolio must reconcile in dollars: $360,000 of revenue minus $215,000 of variable cost minus $85,000 of marketing cost equals $60,000. A second check weights each campaign ROI by its marketing cost:
(60% × 50,000 + 200% × 10,000 + 40% × 25,000) / 85,000 = 70.6%
Use unrounded values in the calculation and round only the displayed result. A campaign with zero marketing cost has no defined ROI under this convention; investigate the cost allocation before including its percentage in an average.
Now change only the search cohort's observation cutoff. Hold its marketing costs constant and show the revenue and variable costs observed at each cutoff:
| Cutoff | Net revenue (USD) | Variable cost (USD) | Contribution before marketing (USD) | Marketing cost (USD) | ROI |
|---|---|---|---|---|---|
| Day 60 | 150,000 | 90,000 | 60,000 | 50,000 | 20.0% |
| Day 90 | 200,000 | 120,000 | 80,000 | 50,000 | 60.0% |
The later cutoff captures $20,000 more contribution before marketing and raises ROI by 40 percentage points. Comparing the earlier search result with fully matured campaigns would mix observation periods. It would not establish a difference in campaign quality.
Finance owns the cost boundary and ledger reconciliation. Marketing operations owns campaign classification, attribution settings and cohort dates. The budget owner decides whether the measured contribution justifies another investment, considering capacity and uncertainty.
Do not turn attributed contribution into incremental profit. Google's controlled-experiment explanation distinguishes outcomes among people exposed to ads from a control group to estimate lift. An assigned invoice alone supplies no such counterfactual. Covirage is not affiliated with Google.
If you use an average marketing ROI by industry, require the publication date, sample, averaging method and matching measurement boundaries. A high return in this example identifies a question for further review; it does not prove that increasing the same campaign's spending would reproduce that return.
Bring a file or Google Sheet containing campaign cohorts, attributed net revenue, variable costs and the full marketing cost boundary. Covirage maps the columns, and its tools compute and check contribution, campaign returns and the investment-weighted total; the external AI model explains the result with every figure traced to its rows. Create a free Covirage account and bring a sample of your own export; creating the account takes about a minute.
See customer profitability for the contribution behind the campaign revenue.
Use the company's required return and a comparable internal baseline. A useful outside benchmark must disclose its sample, measurement period, margin definition, included costs and attribution basis. Without those details, a published average cannot tell you whether your own campaign return is strong.
A simple average describes the average campaign when each campaign gets equal weight. To measure the return on the combined marketing investment, divide total contribution after marketing by total marketing cost. These answer different questions when campaign investments differ.
No. ROAS compares attributed revenue with advertising spend. The marketing ROI convention used here deducts variable product and delivery costs, then all included marketing costs, before dividing the remaining contribution by marketing cost. State that convention whenever you compare results.
Choose a window that fits the buying cycle and the decision being made. Then give compared cohorts the same elapsed time and disclose outcomes still pending. Keep realized invoice contribution separate from estimates of future renewals, repeat purchases or lifetime value.