Question 1

How does artificial-intelligence property price prediction work?

Accepted Answer

We offer two engines. The statistical model (our reference) starts from the real trend of your area and then converges it towards the long-term pace of the French market, with scenarios anchored to interest-rate trajectories (BCE), recalibrated every month. The AI model (beta) is a machine-learning algorithm trained on 12 years of DVF notarial transactions — 57,000 observations — that learns which local characteristics herald rises or falls. You compare the two with one click.

Question 2

Is the AI prediction reliable?

Accepted Answer

We publish our figures: over the test year, the AI model shows a mean absolute error of 6.5 % at 1 year and 2.8 % per year at 5 years — better than naive extrapolations. No prediction is a guarantee: that is why each projection is framed by three scenarios and the statistical model remains our reference.

Question 3

What data feeds the prediction?

Accepted Answer

Only official sources: real sales recorded at notaries' offices since 2014 (DVF/Etalab, 9.2 million transactions), our Case-Shiller-type repeat-sales index, and the mortgage rates, inflation and 10-year OAT published by the European Central Bank — updated every month.

Question 4

Is MyHomePrice the first French AI property-prediction tool?

Accepted Answer

We are one of the first mainstream French services to combine machine-learning prediction, a departmental-scale repeat-sales index and scenarios anchored to real rates — all 100 % online, with no agency or cold-calling, with a published and verifiable methodology.

Price prediction: your property in 5, 10 and 15 years

DPE & renovation: keep as is or renovate?

Frequently asked questions about the prediction

What your results look like