Independent listing analysis · observed 17 September 2026
What do Moroccan homes actually ask for?
A filterable read of five marketplaces, balanced so no portal dominates the headline. Asking prices are not transaction prices.
Geography
Where price per m² is highest
Size effect
Price intensity by area band
Source effect
Portal medians are not interchangeable
Inventory mix
What the filtered market contains
Readout
What the current filters say
Comparable city table
Medians, spread and sample depth
Use the sample count and portal count before reading any rank as a market fact.
| City | Listings | Portals | Median price | Median MAD/m² | Middle 50% MAD/m² | Read |
|---|
02
The quarter lens
June, July and August are visible — but only where dates are defensible.
The dated cohort contains 728 quality-controlled apartment-sale listings from Agenz, Sakanly and Sarouty. It excludes a Sarouty batch of 445 otherwise eligible apartments dated 7 August because 898 records shared that single creation date while every other in-window day had at most eight.
Mubawab did not expose an original publication date on the inspected national pages. Yakeey exposed dates on detail pages but not in the national result payload used for this capture. They remain in the equal-source snapshot and are not falsely backdated into the quarter.
How to read this
Two datasets, two different questions
Equal-source snapshot
Question: What do current listings look like across portals, cities, size bands, property types and transaction types?
Design: 1,000 validated records per portal, split into 500 sales and 500 rentals. Total: 5,000.
Strength: No marketplace dominates the overall sample.
Limit: It is a 17 September snapshot, not a June–August time series.
June–August dated cohort
Question: What was posted during June, July and August among records with verifiable original dates?
Design: Apartment sales only, 20–500 m² and 2,000–80,000 MAD/m², with the 7 August Sarouty bulk-import anomaly removed.
Strength: Exact quarter boundaries and explicit timestamp rules.
Limit: Three portals, not five. It is not an equal-source sample.
Interpretation rules
Figures are advertised asking prices. They do not reveal negotiated transaction prices, days on market, duplicate cross-posts between portals or final contract terms.
Medians reduce the influence of extreme listings. The middle 50% range shows dispersion. Low-sample rows are marked clearly.
Publish and verify
Everything behind the charts is downloadable.
Use the balanced file for the five-source comparison. Use the dated file only for the quarter analysis. The audit file records the row-count and completeness checks.