Reviewed: 9 October 2026. All figures and business records below are original fictional examples. They illustrate measurement definitions and arithmetic, not a real campaign result, an investment recommendation or evidence that advertising caused sales.
Define the return and cost boundary before calculating
Marketing ROI measurement becomes misleading when the label is chosen before the inputs are defined. Revenue divided by advertising spend and profit divided by total costs are different ratios. A dashboard may also count sign-ups or downloads as conversions, which does not make them realised sales revenue.
Google Ads' ROI guidance discusses profit relative to defined costs; its physical-product example includes production and advertising costs. Its campaign-return guidance connects measurement to campaign goals and explains that tracked actions can include purchases, sign-ups or downloads. These are narrow references, not evidence for this fictional campaign. The calculations below explicitly name their own denominators rather than assuming that accounting cost of goods sold universally includes advertising.
Before calculating, state the period, the revenue basis, the included costs and what the tracking system actually records. If a required figure is missing, identify it as missing. Treating an unknown cost as zero can make a precise-looking percentage answer a different question from the one intended.
Keep preparing
Continue with Sarkari Resume Templates₹299 — coaching के एक महीने से काफ़ी सस्ता / far cheaper than a month of coachingOriginal complete toy packet
The supplied exercise concerns one fictional campaign period. It states that recognised revenue for that period is ₹4,000, and that all costs within the exercise's specified scope are supplied below. No tax, refund or other unlisted adjustment is part of this toy packet. That completeness is an explicit exercise assumption, not something inferred about a real business.
| Item | Supplied value and meaning |
|---|---|
| Revenue | ₹4,000, recognised on the stated fictional period basis. |
| Production cost | ₹2,200 for the same supplied scope. |
| Marketing cost | ₹600 for the same scope; this is counted once in total costs. |
| Other cost | ₹200, also within that scope. |
| Total supplied costs | ₹2,200 + ₹600 + ₹200 = ₹3,000. |
| Residual after supplied costs | ₹4,000 − ₹3,000 = ₹1,000. |
The packet's completeness allows calculation of a residual after those costs. In an actual account with missing overhead, labour, refunds or taxes, the same subtraction would not establish net profit. Use a label that matches the costs actually included, and check whether the accounting basis and period are consistent.
Calculate return on the total supplied costs
For this exercise, define return on total supplied costs as the residual divided by those total costs, expressed as a percentage:
(₹4,000 − ₹3,000) / ₹3,000 × 100 = 33.333…%.
The numerator is ₹1,000. The denominator is ₹3,000, including production, marketing and the supplied other costs. Reporting approximately 33.3% is consistent with rounding that value. The calculation does not say that advertising alone produced the residual or that this percentage is a benchmark for a particular industry.
An original reporting sentence could be: “Within the complete fictional cost scope supplied for this period, revenue of ₹4,000 less total costs of ₹3,000 leaves ₹1,000. The residual is approximately 33.3% of those total costs. This is a return-on-supplied-total-costs calculation, with no claim of causal advertising uplift.”
That sentence makes the input boundary visible. Without the complete-cost assumption, it would need to say which costs remain unknown before claiming the same overall return.
Calculate the separate revenue-to-ad-spend ratio
Using the same packet, revenue divided by the ₹600 marketing spend is:
₹4,000 / ₹600 = 6.666….
This is approximately 6.67 times the supplied marketing spend. It is a revenue-to-ad-spend ratio, often discussed as ROAS when the numerator is defined advertising-attributed revenue and the denominator is corresponding advertising spend. This packet supplies period revenue; it does not establish that all ₹4,000 was attributed to an advertising platform. Therefore, label this result as the packet's revenue-to-marketing-spend ratio rather than falsely reporting a verified platform ROAS.
The 6.67 ratio does not subtract production or other costs. It therefore cannot replace the earlier return-on-total-costs calculation. A large revenue ratio can coexist with a small or negative residual when other costs are large. The two values are not inconsistent; they answer different questions.
Identify two tempting calculation errors
First, dividing ₹1,000 by ₹600 gives approximately 166.7%. That is the supplied residual relative to marketing spend, not the return on the defined ₹3,000 total-cost base. If someone wants that ratio, name it and explain why that denominator answers their question. Do not relabel it as the earlier calculation.
Second, subtracting marketing spend again from the ₹1,000 residual would double-count it, because ₹600 is already included in ₹3,000. The arithmetic should follow a cost ledger where each amount appears once in the intended total. If two columns overlap, reconcile their scope before adding them.
An error investigation should begin with the definitions and included records, not with choosing whichever percentage looks better. A reproducible calculation lets another reader reconstruct the numerator and denominator from the same packet.
Change one supplied cost and recompute
Consider a separate fictional variant where other costs are ₹500 rather than ₹200, with all other inputs unchanged. Total supplied costs become ₹2,200 + ₹600 + ₹500 = ₹3,300. The residual becomes ₹4,000 − ₹3,300 = ₹700.
Return on those total costs becomes ₹700 / ₹3,300 × 100, approximately 21.2%. The revenue-to-marketing-spend ratio remains approximately 6.67 because neither its ₹4,000 numerator nor its ₹600 denominator changed. This contrast shows why a revenue ratio does not reveal the complete cost picture.
The variant is a sensitivity exercise, not an observed change in a campaign. It does not predict what will happen if a real organisation spends more. Its conclusion is confined to the arithmetic when one stipulated cost changes.
Keep conversion counts separate from revenue
Suppose another fictional report supplies ten completed sign-ups but no paid orders, prices or recognised revenue. That report can support a count of ten sign-ups under its stated definition. It cannot support multiplying those sign-ups by an invented customer value and presenting the result as realised revenue.
If a reporting system uses assigned conversion values, name them as assigned values and explain their basis. A purchase event may also need reconciliation with cancellations, duplicates and the chosen recognition period before it represents the revenue figure required for your calculation. The existence of a tracking event does not independently settle those accounting questions.
For the original ₹4,000 packet, revenue is an explicit supplied input. It is not derived from the ten-sign-up alternative. Keep the cases separate so that a reader does not accidentally combine a revenue amount from one scenario with a conversion count from another.
Attribution and causation need different evidence
An attribution rule can allocate credit to a channel. It does not, by itself, establish how many sales would have occurred without that channel. The toy packet has no control group, counterfactual design or evidence of incremental effect. Consequently, the residual cannot be described as profit caused by advertising.
For reconciling order counts and fractional credit, see the marketing analytics guide. When investigating inconsistent reporting definitions, use the research skills guide. Those exercises help inspect the records, but they do not create missing causal evidence.
Use a report that another person can reproduce
Present the period and basis, the included revenue, the nonoverlapping cost components, the formula and the result. State what remains unmeasured. If the revenue basis differs from the cost period, resolve that mismatch before treating the ratio as a campaign result.
For this fictional example, the defensible report contains both the approximately 33.3% return on the supplied total-cost base and the approximately 6.67 revenue-to-marketing-spend ratio, with their separate meanings. It does not recommend increasing spend, judge campaign quality against an invented benchmark or claim verified platform performance.
Frequently asked questions
Can I use revenue divided by spend as ROI?
Explain which ratio you are calculating. Revenue divided by marketing spend does not subtract the other costs in this example and is not the same as its residual divided by total costs.
What if a required cost is unknown?
Keep it unknown. A result using only known costs must be labelled accordingly and does not establish a complete-cost return.
Does a positive result prove the campaign caused profit?
No. The arithmetic describes the supplied records and assumptions. Establishing incremental effect requires evidence this exercise does not provide.