> Who this is for: a landlord, family office or management company with 30–600 units in Dubai who has to produce renewal notices unit by unit, not one spreadsheet percentage for the whole building. > > Disclosure: Rent360 sells UAE property management software. We are not neutral about software. We are neutral about the numbers below: every one is dated and sourced, and where we could not verify something we say so instead of guessing.
The change that breaks the portfolio spreadsheet
Until recently you could defend a renewal with one number: the community average. That is no longer how the benchmark is produced.
Public reporting on the 2026 index describes a shift from broad district averages to property-specific benchmarking, with rental values generated from Ejari registrations, executed contracts and transaction data rather than an annual snapshot (dda-realestate.com, RERA Rental Index Update 2026, retrieved 19 Sep 2026 — https://dda-realestate.com/posts/rera-rental-index-update-2025-impact-on-dubai-tenants). Other 2026 write-ups describe the index tracking individual sub-communities and clusters separately, and separate tracking of furnished versus unfurnished stock in some high-density areas (aigentsrealty.com, retrieved 19 Sep 2026 — https://aigentsrealty.com/blog/dubai-smart-rental-index-2026-changes-guide).
What we could not verify from an official DLD source: the exact weighting of the building-classification score, the list of areas where furnished/unfurnished is tracked separately, the specific AED averages quoted by brokerages, and whether a unit's sub-community assignment can be corrected on request. Treat brokerage tables as indicative only. The binding output is whatever the official calculator in the Dubai REST app / DLD website returns for your building on the day you run it. If you have an official circular that settles any of the unverified points, send it to us and we will publish the correction with attribution.
The caps have not changed — the benchmark under them has
Decree 43 of 2013 bands, still the operative maths in 2026 reporting (retrieved 19 Sep 2026):
| Gap between current rent and the indexed market rate | Maximum increase at renewal | |---|---| | Up to 10% below | 0% | | 11–20% below | 5% | | 21–30% below | 10% | | 31–40% below | 15% | | More than 40% below | 20% |
Two consequences most portfolio sheets miss:
- A wrong benchmark moves a unit across a band boundary. A 2% error in the average is the difference between 0% and 5% on a AED 90,000 lease — AED 4,500 a year, per unit.
- The band is computed per unit, so a 100-unit portfolio can legitimately contain units at 0%, 5% and 10% in the same building.
And the procedural rule is unchanged: a rent increase or change of terms requires 90 days' written notice before expiry unless the parties agreed otherwise. A correct number served late is not enforceable.
The mapping method (do this once, then maintain it)
Your goal is one row per unit with the five fields the calculator actually consumes. Everything else is decoration.
Required per unit:
- Area / community name exactly as the calculator spells it — not your internal name, not the marketing name on the brochure.
- Building name exactly as registered — including tower/cluster letter. "Cluster D" and "D Tower" are not interchangeable inputs.
- Property type — apartment or villa.
- Bedrooms — studio is its own category, not "0BR", and studio categories have been added in communities where studio supply grew.
- Current annual rent on the registered contract — the Ejari figure, not the figure in your ledger if the two disagree.
Then three derived fields you maintain yourself:
- Furnished / unfurnished, as stated in the contract.
- Renewal date and notice-due date (renewal date minus 90 days).
- Last benchmark run: date, calculator output range, resulting band, and a screenshot or PDF saved against the unit.
That last field is the whole point. A benchmark you ran in March and applied in September is not evidence; a dated calculator output stored per unit is.
The three places the mapping breaks
1. Internal names drift from registered names. Your portfolio calls it "Marina Tower B"; Ejari says something else; the calculator only recognises the registered string. Fix: keep a registered_building_name column that is only ever copied from the Ejari certificate, never typed from memory.
2. Sub-community is missing entirely. Most landlord spreadsheets carry community only. In a large community where clusters are now benchmarked separately, community-level is no longer a usable key. Fix: add sub_community and populate it from the title deed / Ejari address, then test one unit per cluster in the calculator to confirm the cluster is recognised as its own entry.
3. Furnishing is a habit, not a field. Units get furnished between tenancies and nobody updates the record, so a furnished unit is benchmarked against unfurnished stock. Fix: make furnishing status a required field at contract creation, and re-confirm it at every move-out inspection.
Renewal-season sequence for a 100-unit portfolio
- T-150 days: freeze the data. Reconcile registered building names and sub-communities against Ejari certificates. Flag every unit where furnishing status was changed in the last 12 months.
- T-120 days: run the official calculator per unit. Store date, output range, computed gap, permitted band. Do not batch-estimate from a brokerage table.
- T-110 days: management review of outliers only — units where the permitted band jumped two steps versus last year, and units showing 0% where you budgeted an increase. These are usually mapping errors, not market moves.
- T-90 days: serve written notice for every unit where terms change. Log the service method and the proof of service against the unit.
- T-30 days: re-run the calculator on any unit still under negotiation; the index updates during the window and the stored output is what you will be asked for if the renewal is disputed.
What this looks like in software (our bias, stated)
Any competent system should hold sub-community, registered building name and furnishing as first-class fields, store a dated benchmark result per unit, and drive the 90-day notice clock off the lease record automatically. Rent360 does this and keeps the calculator output in a per-unit evidence trail. Spreadsheets can do it too if you are disciplined about steps 1–8 above.
Not ideal for: single-unit owners (the calculator on its own is enough), holiday-home/short-let operators (different regime entirely), and Abu Dhabi/Sharjah portfolios — those emirates do not use this index, and applying Dubai bands there is simply wrong.
FAQ
Does the Smart Rental Index use community averages in 2026? Reporting from 2026 describes a move away from single community averages to property-specific benchmarking, with sub-communities and clusters tracked separately in large communities (retrieved 19 Sep 2026). Confirm your own unit in the official calculator.
What is the maximum rent increase in Dubai in 2026? Up to 20%, and only where the current rent is more than 40% below the indexed market rate. If the rent is within 10% of the index, no increase is permitted at renewal (Decree 43 of 2013 bands, as applied in 2026).
How much notice must a Dubai landlord give for a rent increase? 90 days in writing before the contract expires, unless both parties agreed otherwise in the contract.
Does furnishing change the benchmark? 2026 reporting describes separate furnished/unfurnished tracking in some high-density areas. We could not verify the official list of those areas, so check the calculator per unit rather than assuming a premium percentage.
Can I apply one increase percentage across a building? Not safely. The band depends on each unit's current rent versus the benchmark, so the same building routinely produces different permitted percentages.
Last reviewed 19 September 2026. Sources are dated and linked above. Corrections welcome — we publish them with attribution.