Measure a marketing agency's ROI by comparing the additional contribution your marketing produces with the complete cost of producing it. Include the agency fee, media, production, relevant tools and internal work. Use sales and cost records to check the result, and distinguish revenue attributed to marketing from sales that would not have happened without it.
For a Dubai business, a useful calculation is:
Marketing ROI = (incremental contribution − marketing costs) ÷ marketing costs × 100%.
Incremental contribution is the additional revenue generated, less the direct costs of delivering those sales, before deducting marketing costs. Define the period, cost basis and evidence for “additional” before calculating the percentage. A strong advertising return can coexist with a weak whole-programme return.
Move up the evidence ladder
- Work delivered: the agreed campaigns, content or implementation are complete.
- Relevant traffic or responses: the work reaches people who could become customers.
- Qualified opportunities: suitable prospects have a real buying need.
- Sales: business records show purchases, with cancellations and refunds accounted for.
- Contribution: revenue is reduced by the direct costs of fulfilling those sales.
- Incremental contribution: estimate how much contribution would not have arisen without the programme, where a credible comparison is possible.
Evidence at one level does not automatically prove the next. Many SMEs cannot estimate incrementality precisely; show the uncertainty rather than presenting attributed revenue as certain additional profit.
Decide exactly what you are measuring
Start with a written measurement agreement. Identify the programme, included channels, acquisition period, outcome period and costs. If you are assessing a new service campaign, separate it from the business's entire revenue.
There are three different questions:
- Did the agency deliver the agreed work? Check implementation, quality and deadlines.
- Did the marketing programme create a useful financial return? Compare its additional contribution with its complete costs.
- Did hiring this agency improve on the realistic alternative? Compare with what internal delivery, a previous supplier or a different approach could reasonably have achieved.
Positive programme ROI does not prove all the return was created by the agency. Existing demand, your sales team and assets built previously may contribute. Likewise, an agency can complete its work while the commercial outcome remains disappointing. Discuss both delivery and economics.
Build a cost total that matches the scope
Include costs necessary to run the measured programme: agency management, media, commissioned content, landing-page work, tracking or reporting tools, and the internal time devoted to it. Avoid counting production twice when it is already inside a retainer.
Separate initial work from recurring expenses. If a website or content library supports several future campaigns, explain the allocation method rather than charging its entire cost to one small campaign or excluding it completely.
Keep cash expenditure and allocated resources visible. Existing employees' time can belong in an economic assessment even when it creates no new payroll payment. It should not be presented as new cash leaving the bank.
Use a consistent currency and VAT/tax basis for revenue and costs, with finance confirming any necessary conversion or tax treatment.
Follow enquiries through to business outcomes
Record the journey that matters to your business. For a Dubai service company, it could be:
Enquiry received → relevant and contactable → consultation or quotation → accepted sale → delivery → payment and any refund.
A click, a received message, a qualified enquiry and a customer are different events. If enquiries arrive through WhatsApp, calls and forms, reconcile them in your business records so repeated contacts do not become several customers.
Keep the relevant dates, source where known, qualification result, sale value and outcome under an internal record ID. Use customer information only through appropriate access, privacy and consent arrangements. Do not place names, email addresses or telephone numbers in ordinary analytics event fields or campaign URLs; Google expressly restricts personally identifiable information in Analytics. Google Analytics privacy guidance.
Track genuine “unknown source” records rather than inventing attribution. A customer saying they found you online can be useful evidence without identifying a particular advertisement.
Separate attribution from additional sales
Attribution assigns credit to marketing interactions. Google Analytics describes models that allocate that credit across the path to an important action. That is useful for reporting, but an attributed sale alone does not demonstrate that the customer would otherwise never have purchased. Google's attribution explanation.
Do not add every platform's claimed revenue together as though each reports different customers. Reconcile against sales records and investigate differences in definitions, dates and credit windows.
Where feasible, use a properly designed comparison to estimate additional impact. Google's Conversion Lift uses treatment and control groups to measure incremental conversions, but it is not available to every account. Google Conversion Lift.
If a formal experiment is not practical, report the limitation. Compare against a reasoned baseline, record changes in prices, capacity and other marketing, and show how different assumptions change the conclusion. A simple before-and-after increase is an indication to investigate, not automatic proof of causation. Dubai branch openings, changed service areas and seasonal promotions can also affect what is being compared.
An AED example with the assumptions kept visible
This is a hypothetical programme, not a Lunasol quote, client result or Dubai benchmark. Assume a completed acquisition cohort has been followed through its normal sales cycle. The business estimates AED 120,000 in additional net revenue after discounts and refunds, excluding tax. That estimate is an assumption for the example, not something an advertising dashboard proves.
Assume a 40% contribution margin after direct fulfilment costs and before marketing costs. The programme uses these resources:
| Cost for the measured programme | Hypothetical amount |
|---|---|
| Agency fee | AED 9,000 |
| Advertising media | AED 20,000 |
| Separately commissioned production | AED 4,000 |
| Programme tools | AED 1,000 |
| Allocated existing staff time | AED 2,000 |
| Total marketing resources | AED 36,000 |
Assume the first four rows are new cash payments. They total AED 34,000; the remaining AED 2,000 values existing staff time. The calculation below uses the full AED 36,000 resource cost and assumes no other necessary programme costs are omitted.
Additional contribution is AED 120,000 × 40% = AED 48,000. After the AED 36,000 marketing cost, the remaining contribution is AED 12,000. The estimated marketing ROI is therefore AED 12,000 ÷ AED 36,000 × 100% = 33.3%.
This is a contribution-based programme return, not company net profit or a cash payback calculation. Collection dates and wider fixed overhead still matter.
Test the assumption that changes the decision
If only AED 80,000 of net revenue is genuinely additional, contribution falls to AED 32,000. With the same AED 36,000 marketing resources, ROI becomes approximately −11.1%. The conclusion reverses without changing the agency's invoice.
At the assumed 40% contribution margin, the programme needs AED 90,000 in additional net revenue to cover AED 36,000 of marketing resources: AED 36,000 ÷ 40%. This break-even figure applies only while the margin and cost assumptions hold. It is not a revenue forecast.
If the advertising platform separately credits AED 120,000 in revenue to AED 20,000 of media, its revenue ROAS is 6×. That ratio does not deduct production, agency costs or fulfilment costs, and does not prove all the credited sales are incremental. ROAS and programme ROI answer different questions.
Give the sales cycle time without losing delivery accountability
For a business selling substantial projects, leads acquired in one month may close much later. Keep those prospects together as an acquisition cohort and update their outcomes. Comparing this month's marketing cost only with this month's sales can mix old opportunities with new spending.
Report open pipeline separately from completed revenue. If you use expected future value, label it as a forecast and state the conversion and margin assumptions. Do not count an unsigned quotation as a completed sale or assume every new customer will repeat indefinitely.
Meanwhile, review what the agency actually delivered. Pending sales do not excuse missing agreed work, and complete delivery does not guarantee future ROI.
Turn the review into a specific decision
Ask the agency to reconcile the financial result with the customer journey. If many enquiries are unsuitable, examine targeting, offer clarity and qualification. If relevant prospects fail to progress, investigate response, pricing, availability and the sales process. Treat these as possible explanations to test.
A practical review should leave you with the measured result, its main uncertainty, the next change, the owner of that change and the evidence that will determine whether to continue. Avoid expanding spend simply because a platform ratio looks attractive when the business cannot explain the underlying contribution.
What Lunasol's case evidence can tell you
Lunasol's DP Business Solutions case study reports 6.2× ROAS for a specific e-learning business, within an integrated system spanning advertising, content, funnels, email, CRM and sales processes. ROAS compares credited revenue with ad spend; it is not whole-programme ROI. This is a first-party account of that engagement, not an independently audited Dubai benchmark, a typical Lunasol result or a promised future return.
The relevant starting point is the connection between marketing activity and sales records. When discussing work with Lunasol, bring the costs, outcome definitions and sales-cycle information available to your business. Agree how the commissioned work will be measured and which additional tracking or integration requires separate scope.