Catchments that follow roads
Drive-time isochrones from a self-hosted routing engine, not radius circles. A 15-minute contour in Riyadh has 156 holes and hugs the arterials, because that is what 15 minutes actually looks like.
Site Selection · Riyadh
Know where to open next,
and what it costs the stores you already have.
Expansion decisions in the Gulf are still made from broker lists and consultant decks. The inputs exist — population, income, competitors, drive times, footfall — but they are scattered, unjoined and stale. This turns them into a score you can defend line by line.
42,682
baseline hexagons
7.03M
residents modelled
209,332
places mapped
under 1s
to profile a site
What it does
A composed city baseline
42,682 hexagons for Riyadh, each carrying population, built form, road access, land value and a footfall index, reconciled against the 2022 census to within 0.35%.
Scores that explain themselves
Every result arrives with its per-factor contributions: what helped, what hurt, and how much of the intended weight was actually available. A score with no derivation is not an answer.
Competitors, de-duplicated
209,332 places across 649 brands, merged from three sources and normalised through Arabic and English spellings, so that a competitor count means something.
White space, not heat maps
Rank the whole city for a given vertical, respect the network you already have, and get back concrete points snapped to the road network.
It refuses when it should
Where a factor is flat across the city or missing for the region, the engine excludes it and says so, or declines to score at all. A confident number built on a constant is the failure that survives review.
What it will not do
Scores are a suitability index, not a revenue forecast, and are never denominated in currency. The footfall layer is a modelled proxy pending calibration against real store performance. Where a factor has no spatial variation, or no data for the region, it is excluded from the score and reported as excluded. The platform would rather return less than return something that looks right and is not.
Read the full methodology and limitations →