A national traffic campaign can produce one attractive total while hiding a basic business question: can either emirate produce the qualified visits you need? A single UAE row blends location signals, landing-page relevance and cost, so a good result in one market can cover a weak result in the other.
A split test gives each emirate its own budget, page and measurement record. The example below uses a hypothetical AED 2,000 test, not a client result. It shows how to compare cost per qualified result without assuming that population size, visitor volume or a supplier estimate predicts performance.
Why a national total hides useful differences
Imagine a service business that wants 20 qualified visits from the UAE test. The report shows 2,000 visits and 24 qualified visits, which appears to beat the target. It does not show whether Dubai supplied 19 qualified visits and Abu Dhabi supplied 5, or the reverse. Those two situations call for different follow-up decisions.
Dubai and Abu Dhabi are also different geographic markets. Official UAE information reports Abu Dhabi's 2024 population at 4,135,985 and describes an emirate that includes Abu Dhabi city, Al Ain and Al Dhafrah. Dubai's official portal reports 3,863,600 residents at the end of 2024.
These figures provide context, but they are not a reason to assign spend by population alone. Your audience, service area, page relevance and qualification rule matter more than a simple resident count.
The practical response is to make location a reporting dimension before the test starts. You may want to use two rows, two destination pages or clearly separated page paths, and two consistent campaign names. Then the business can compare like with like instead of explaining a blended average after the budget is gone.
Build two clear emirate campaign rows
A useful starting point is a 55/45 split: AED 1,100 for Dubai and AED 900 for Abu Dhabi. The split is a test assumption, not a forecast. If Abu Dhabi is the stronger service area, the same method can start at 45/55 or 50/50.
For guidance on the Abu Dhabi portion, see the Abu Dhabi website traffic page. A separate Dubai website traffic page can support the Dubai row when that market needs its own scope.
You can keep the business outcome consistent across both rows. If a qualified result means a visit that reaches the contact section, stays at least 30 seconds, or completes an agreed form event, use the same definition in both emirates. A change in definition would make the cost comparison look precise while measuring different things.
| Emirate | Test allocation | Share | Destination page | Campaign name | Location limit |
|---|---|---|---|---|---|
| Dubai | AED 1,100 | 55% | Dubai service page | uae-test-dubai |
Dubai only |
| Abu Dhabi | AED 900 | 45% | Abu Dhabi service page | uae-test-abu-dhabi |
Abu Dhabi emirate only |
| Total | AED 2,000 | 100% | Two matching local pages | Two separate rows | No blended row |
The allocation adds to AED 2,000 because 1,100 + 900 = 2,000. The location limit should be agreed before launch. If a supplier cannot isolate an emirate precisely, record that limitation instead of presenting a mixed audience as a clean local test.
Match each budget row to a local page and tag
The main problem with a split test is often not the arithmetic. It is a mismatch between the row name and what the visitor sees. If the Dubai row lands on a generic national page while the Abu Dhabi row lands on a specific local page, the page experience becomes a second variable.
A consistent tag makes each destination easier to identify in analytics. For example, a hypothetical destination URL could use this pattern:
https://example.com/dubai-service?utm_source=fastvisits&utm_medium=referral&utm_campaign=uae-test-dubai
https://example.com/abu-dhabi-service?utm_source=fastvisits&utm_medium=referral&utm_campaign=uae-test-abu-dhabi
These are format examples only, not FastVisits destinations. It helps to use your real destination URLs and confirm that your analytics system records the campaign values. Standard campaign parameters commonly identify source, medium and campaign, while a campaign name such as uae-test-dubai separates the emirates. Keep spelling and capitalization consistent so one row does not split into several labels.
Your page check can be simple: confirm that the page names the service area accurately, presents the same offer and qualification event, and loads on mobile. The test should compare emirates, not two unrelated pages or two different conversion rules.
Define qualified results before spending
A traffic count can rise while the useful audience stays flat. For example, a visit that leaves immediately may count as a session, but it should not count as a qualified result if your agreed test requires meaningful engagement. The qualification rule should reflect the next action your business can actually review.
A practical step is to use one rule for both rows. A practical definition might be: “a session from the named emirate that reaches the enquiry section and remains active for at least 30 seconds.” Here, a qualifying result means an engaged session, not a qualified lead or a sale. A form submission should be tracked separately, and accepted inquiries and contribution should guide larger spending decisions once those records are available.
The report can include these fields:
- allocated spend
- delivered visits
- qualified results
- enquiries, if relevant
- cost per visit
- cost per qualified result
- share of total qualified results
- location exceptions or data-quality notes
It is better not to infer quality from cost per visit alone. A low cost per visit may be useful for awareness, but the business decision in this test is based on qualified results and the agreed service-area limit.
Run the arithmetic with one worked example
Suppose the hypothetical review records 1,000 Dubai visits and 750 Abu Dhabi visits. The rows produce 20 Dubai qualified results and 15 Abu Dhabi qualified results. The business target is 35 qualified results from the full AED 2,000 test.
| Emirate | Spend | Visits | Qualified results | Cost per visit | Cost per qualified result |
|---|---|---|---|---|---|
| Dubai | AED 1,100 | 1,000 | 20 | AED 1.10 | AED 55.00 |
| Abu Dhabi | AED 900 | 750 | 15 | AED 1.20 | AED 60.00 |
| Total | AED 2,000 | 1,750 | 35 | AED 1.14 | AED 57.14 |
The calculations are spend divided by visits and spend divided by qualified results. The total cost per qualified result is AED 2,000 / 35 = AED 57.14 when rounded to two decimals. Dubai contributes 20 / 35 = 57.1% of qualified results, while Abu Dhabi contributes 42.9%.
The example meets the 35-result target, but it does not prove that Dubai will always be cheaper. It is a controlled illustration of the report format. Your actual review should also show whether the location data is complete and whether both pages used the same qualification event.
Use a fair review rule to retain, move or stop spend
A test becomes hard to manage when the review rule changes after seeing the results. A pre-agreed rule keeps the next action connected to evidence rather than preference. It should include a minimum observation threshold, because one qualified result can produce an unstable cost figure.
One workable rule for this illustrative budget is:
- Retain: keep an emirate's allocation if it reaches at least 15 qualified results and its cost per qualified result is no more than 10% above the other row.
- Move: if one row reaches at least 15 qualified results and is at least 20% cheaper per qualified result, move up to 20% of the next test budget toward that row, while leaving a smaller comparison allocation in the other emirate.
- Stop and inspect: pause an emirate only if it has received its planned allocation, has a material data-quality issue, or produces zero qualified results after the agreed minimum sample. Inspect the page, location evidence and qualification event before declaring demand absent.
With the worked numbers, Dubai costs AED 55 per qualified result and Abu Dhabi costs AED 60. The difference is 9.1%, calculated as AED 5 / AED 55, so the example falls inside the 10% retain band. No budget move is justified by this rule. The close result also shows why a national average alone would have hidden a useful comparison.
Keep the next report separate from the first one
The last problem is losing the comparison when the next campaign begins. A new period, page or qualification rule can make the old rows difficult to interpret, even if the campaign name looks similar.
You may want to read the Dubai traffic test budget guide for a narrower Dubai planning example, then keep this two-row sheet as the UAE comparison record. If visits appear outside the agreed area, the article on Dubai campaigns reaching other emirates provides a useful diagnostic topic for the review.
The two-emirate budget worksheet contains allocation rows and blank review fields; the formulas above show how to calculate the comparison. You can fill in actual spend, visits and qualified results only after the measurement window closes. A blank field means “not reviewed,” not zero. Record zero explicitly when the system confirms no result.
What the AED 2,000 split should tell you
The purpose of the split is decision clarity. It separates two emirate records, keeps the business outcome consistent, and makes cost per qualified result visible without pretending that a small test forecasts future performance.
It helps to start with an allocation you can explain, match each row to a relevant page, set the location limit, and use the same qualification rule. At review time, check arithmetic and data quality before retaining, moving or stopping spend. A clean comparison is more useful than a large national total that cannot show where the useful results came from.