Would this location make money — before you sign the lease?

We're Assaria: we predict how a restaurant, gym, or store will perform at a given location, using catchment population, foot traffic, and competition — before a single lease is signed. Below is a live version of that prediction model, scoring any point in Sweden as you click. The data shown is synthetic and illustrative; the method itself is validated on real client portfolios.

01 · The portfolio

A portfolio, mapped

Every dot is a site — click one to see what drives its revenue. Click anywhere else in Sweden to score a brand-new location.

This portfolio is synthetic and illustrative, built to demonstrate the method — not your company's real data.

02 · Model accuracy

How much does location explain?

Out-of-sample, location alone — with no brand, marketing, or pricing data — explains a strong share of the revenue variation across this portfolio.

Strong Out-of-sample accuracy

What matters most

03 · Over- and underperformers

Sites that beat — or miss — what location predicts

Each site's actual revenue compared to what its location alone would predict. The model is trusted to say what location alone should produce, so a site's gap from that prediction points to something specific to that site — its offering, operations, or service — worth investigating, not noise.

Performing above prediction

  • Helsingborg 066 (Helsingborg)+48.2%
  • Borlange 089 (Borlange)+43.5%
  • Gothenburg 039 (Gothenburg)+43.0%
  • Gothenburg 026 (Gothenburg)+42.5%
  • Norrkoping 071 (Norrkoping)+39.9%

Performing below prediction

  • Boras 078 (Boras)-34.9%
  • Gavle 077 (Gavle)-31.6%
  • Vasteras 058 (Vasteras)-28.0%
  • Gothenburg 043 (Gothenburg)-27.8%
  • Stockholm 007 (Stockholm)-26.8%