A useful comparable set allows another reader to see why each property was included, where the records differ and what the evidence cannot establish. Its purpose is not to collect enough impressive sales to justify a price chosen in advance.
This playbook explains a repeatable research method for Abu Dhabi property records. It is not investment, legal or tax advice and does not constitute a professional valuation.
Define the subject before looking for a price
Write down the exact building or project, dwelling type, relevant phase, size basis, layout and condition. Record whether the home is completed or proposed, occupied or vacant, and any material attributes you can verify.
Keep missing details marked unknown. If the advertised area combines internal and external space but the comparison uses internal area only, the apparent price difference is not yet meaningful.
The strongest comparison starts with the same question for every candidate. Do not change the subject's description after seeing which version produces the preferred result.
Preserve the source and evidence type
ADREC's market interface separates sales and residential lease views and offers property and location filters. Use the relevant view and retain the extraction date and available record definitions.
A recorded transaction, an advertised offer and an agent's estimate are different evidence types. Keep them in separate columns or sections. Do not present a listing as achieved merely because it has disappeared from a portal.
The retrieved public interface did not provide a usable current matched sample for this article. The example below is entirely hypothetical, not an ADREC extract.
Apply the same selection rules throughout
Start with the same building or closely comparable cluster, then match the dwelling format and readiness. Apartments should not be ranked against villas simply because the total price looks attractive.
Use consistent area units and definitions. Check whether a terrace, plot, common area or parking allowance is included before calculating a price per unit of area.
Record floor, outlook and condition where supported. An unknown view is not a sea view. If those details are missing from a record, reduce what you claim rather than filling them in.
Choose a stated date range appropriate to the available evidence. There is no universal three month window that makes a thin sample robust. When older or less closely matched records are necessary, identify them separately and explain why.
Record exclusions as carefully as inclusions
Remove duplicate records and explain exclusions. A repeat advertisement is not another sale. A wrong dwelling type does not become comparable by sharing a district.
Do not automatically discard a low sale merely because it weakens the preferred price. Investigate whether it is relevant and whether any unusual terms are actually documented. Unsupported speculation about a distressed seller is not evidence.
Keep a compact research table:
| Field | What to show |
|---|---|
| Record identity | Source reference and date |
| Evidence type | Recorded sale, asking price or other clearly labelled basis |
| Property identity | Project, building, phase and dwelling type |
| Size | Value, unit and measurement definition |
| Material differences | Supported condition, layout, readiness or view details |
| Decision | Included, excluded or context only, with reason |
| Uncertainty | Missing fields and possible effect on the conclusion |
This makes the selection reproducible and shows when the available evidence is too weak for a confident conclusion.
A hypothetical example of selection and median
Suppose six candidate records are found. One is a duplicate of another, and one describes a villa rather than the apartment being researched. Four distinct apartment records remain.
Assume, solely for this teaching example, that those four are genuinely matched on the relevant attributes and each uses the same 1,000 square foot area basis:
| Illustrative record | Hypothetical sale price | Price per square foot |
|---|---|---|
| A | AED 1,980,000 | AED 1,980 |
| B | AED 2,020,000 | AED 2,020 |
| C | AED 2,060,000 | AED 2,060 |
| D | AED 2,100,000 | AED 2,100 |
The median is the average of the two middle prices: (AED 2,020,000 + AED 2,060,000) / 2 = AED 2,040,000. On the stipulated common area basis, that corresponds to AED 2,040 per square foot.
The observed illustrative range is AED 1,980,000 to AED 2,100,000. It is not a confidence interval, a real market estimate or a guarantee for another home. The arithmetic cannot repair weak matching or missing records.
Do not manufacture precise adjustments
A better view, different condition or later transaction date may be relevant, but its financial effect needs evidence. Do not add a standard percentage for each desirable feature simply to make unlike homes resemble the subject.
If reliable support for an adjustment is unavailable, describe the difference and keep the result appropriately qualified. A candidate requiring several large unsupported changes may be more useful as context than as a primary comparison.
For a formal valuation, refer the assignment to a qualified valuer. A transparent research worksheet is useful preparation; it does not confer professional valuation status.
RICS maintains a professional standard on comparable evidence. Its page reviewed on 2 October 2026 still directs members to the first edition while the second edition is a consultation draft. Do not present a draft as the adopted standard or describe this article as RICS certification.
Present the result without overstating it
Lead with the closest supported matches and show the selection period, size basis, evidence type and exclusions. Keep current asking prices as a separate view of advertised competition.
An asking price above the selected sales does not by itself prove that a listing is overpriced or explain why it has not sold. Condition, terms and unobserved differences may remain unresolved.
Finish with what the set supports and what further evidence is needed. If the data cannot support a narrow conclusion, a wider, clearly limited discussion is better than a precise number with hidden assumptions.