How good are your site forecasts, really?

Your existing store list and the historical performance data for the stores you want kept back from the model; we'll confirm the exact format on the kickoff call. The sprint is scoped for a single market or region, so if you operate across several countries, we'll help you pick the highest-value one to start with. Most estates need one sprint; a larger or multi-format estate may need a second.

Then you have spent €4,000 and you keep what it found: the verdict on your data and the specific fixes where there are any. You never commit to step two until step one has cleared, so there is nothing to claim back and no open-ended day rate. Step two can also end in a no, and that finding is worth having too: building on your data and measuring that the signal is not strong enough is the work itself, and it is the cheapest honest way to learn it.

Then you pay half. The Kick Start is €2,000 rather than €4,000 for estates under 15 locations, and the same half rate goes to not-for-profits and social enterprises at any size, €6,000 all-in instead of €12,000. Same rubric, same report. What we will not sell below about 15 comparable mature stores is the model check, at any price: there are not enough stores to hold some back and still learn from the rest, so we could not honestly report an error figure. You get the data verdict and the fixes, not an accuracy estimate, because nobody could give you an honest one at that size.

No. A demo shows you the platform on generic data. The Validation Sprint measures your own forecast baseline and estimates how well IRIS would do on stores of your own the model was never shown, and gives you a real report to keep.