Revenue forecasting for site selection

Forecast what your next site will earn, then defend the number to the board.

IRIS forecasts revenue, shows the prediction interval and sales drivers, and validates it on your own stores.

See it in action
  • Refunded first-year licence if we miss the The R2, MAPE and prediction interval set in your order form. IRIS tier only, in markets with 50 or more comparable mature sites. accuracy
  • Validated on your own stores before you sign
  • Published prices, per market. No "contact us" to see them
Map of an Amsterdam catchment around the sample location
Street-level view of the candidate storefront
Amsterdam CentrumSingel

In use at expansion teams across Europe

McDonald’s
SPAR
Zeeman
Sameday
GrandVision
Kwalitaria
McDonald’s
SPAR
Zeeman
Sameday
GrandVision
Kwalitaria
McDonald’s
SPAR
Zeeman
Sameday
GrandVision
Kwalitaria

Know what you are being sold

We won't quote a generic accuracy number.

The usual alternative is a gut-feel call, or a report you cannot re-run. Anyone can quote a percentage before seeing your data, and no one can check it. A single confident number hides the one thing that matters: how wrong it could be. So once we have seen your stores, you get two honest things instead: the accuracy we hit on stores of your own the model never saw, and the range around every forecast, with how often reality lands outside it.

See how we validate How IRIS compares

The box is the nominal 80% interval: we would expect the real number to land inside it about 4 times in 5. The whiskers reach the 99% interval. The dot is the point prediction. Illustrative, and different for every store. USUALLY SOLD“90% accurate”no rangeUSUALLY SOLD · “90% accurate”WITH IRIS · CHECKED ON YOUR STORESWITH IRISchecked on your stores€1.20M€0.80M€1.0M€1.4M€1.60M
80% interval99% intervalpoint prediction

The same figure, with its range, checked on stores you already run.

The platform

Try the interface for a sample location.

01

Catchment analysis

See who a location can reach, by drive time and travel mode.

02

Revenue models

Forecast turnover for new and existing sites, per revenue stream.

03

Cannibalisation

See how a new opening affects the network you already run.

04

Scenario planning

Compare openings, relocations and closures side by side.

The IRIS workspace for a scored location: the map with its catchment, the location pane, and the model-results pane. Illustrative sample.

Ready for the real thing? Two ways to do that.

Start with IRIS Market

Europe

Score any address in Europe from your browser, with catchments and the demand drivers behind each score. Ranking, never a revenue forecast.

  • Unlimited scores on the market you licence
  • €6,000 a year, per market. No sales data needed
See the plans

Book a demo

Worldwide

See your own locations in the platform, for the market you care about most.

  • Your own stores, in your own market
  • Talk to someone who actually builds the models
Book demo

Customer stories

Decisions that held up in the real world.

95% of deployed models still in production. Not a model-accuracy claim.

Sameday parcel lockerLast-mile · 6,000+ points

From intuition to insight for last-mile locations.

Read the Sameday story →
A Kwalitaria restaurantQSR · 130+ locations

Turning a delivery shift into a revenue model.

Read the Kwalitaria story →
A Zeeman store interiorRetail · ~1,400 stores

A critical mirror for a 1,400-store retail network.

Read the Zeeman story →

"The IRIS platform is now an essential part of our decision-making process. It allows us to test scenarios, predict outcomes, and ensure our growth is sustainable and aligned with market demand."

Gheorghe Marius Adrian, Easybox Business Director at Sameday

Who runs on IRIS

Built for the teams whose growth runs on location.

Started with the two verticals where a single opening moves the year: quick-service food and retail chains.

Four freshly made cheeseburgers in kraft-paper trays, held out on a tray by a member of quick-service restaurant staff. Quick-service & food

Quick-service & food

Model catchments, forecast turnover per revenue stream, and see whether the delivery demand is really there before you sign the lease.

Quick-service restaurants →
A retail store checkout: an assistant hands a customer their shopping bag Retail chains

Retail chains

Purchasing power, household composition and footfall for every catchment, so openings, relocations and closures are decisions, not gut calls.

Retail chains →

Also used by subscription businesses. See all industries →

Two teams, one number

Built for the decision-owner. Documented for your data team.

The same forecast has to survive the room where the call gets made and the stack where your engineers work. Neither team has to do the other's job.

The IRIS workspace: a drive-time catchment map with the model-results pane for a locationThe decision team

Run it yourself

The expansion lead owns the workspace and runs scenarios directly. No data scientist in the loop, no ticket queue.

See the workspace →
The IRIS REST API returning a forecast with its 80% interval and a validated-out-of-sample flagThe data team

Wire it in

A documented REST API and MCP server, with full export, so your data team can pull IRIS into their own stack and governance.

For your data team →

Independent and European

The model is yours, and it's built in Europe.

You own the model and its outputs, and your own data stays yours: never resold, never used to train anyone else's without your permission. IRIS is built and run in Europe and deployed for clients worldwide. The market data underneath is aggregated to roughly 120 m hexes, so very little personal data is ever in play.

You own the model and its outputs Built and hosted in Europe Full export, nothing locked in

Common questions

The questions expansion teams ask first.

How do you forecast revenue for a new store?
IRIS starts from the catchment, working out who a site can actually reach by drive time and travel mode, then reads more than 100 local signals across that area to turn reach into demand. It returns the forecast revenue as a range you can defend, with its prediction interval and the demand drivers behind it, validated on your own stores first.
What is catchment analysis?
A location is several catchments, not one: the same site draws drivers, takes delivery by bike and catches passing pedestrians, and each travel mode is its own catchment. IRIS maps who a site can reach by each mode on live street networks, then measures the footfall, spend and household make-up inside that reach. It is the first step before IRIS forecasts a euro, and it holds for both a site you are scouting and the stores you already run.
How is the forecast validated?
Every forecast is validated out-of-sample on your own stores. IRIS predicts a slice of your existing locations that it was never shown, using only what was known before they opened, then shows you the error, so you can see how far a forecast can miss on your own estate before you trust it on a new site. Every forecast is shown with its prediction interval, never a bare number.
Do we need a data scientist or GIS team to use IRIS?
No. IRIS is built for the person who owns the expansion decision, with point-and-click maps, scenarios and reports, and no GIS or data-science background required. Your data team gets a documented API and MCP server with full data export, but nobody has to open a laptop full of code to get an answer.

More on accuracy, data handling and when to ignore a forecast in the full FAQ.

See a live forecast in your market, proven on your own stores.

A store lease is a multi-year bet, signed before the doors open. Book a demo and we will forecast a live location in your market. Then a Validation Sprint reads it on your own stores, before you commit.

Book demo See the plans
Our guarantee
  • Refunded first-year licence if we miss the promised accuracy
  • Validated on your own stores before you sign
  • Your data stays in the EU. No shared-model training without permission