Disagreement is information, not a flaw
When two sources diverge widely on a home, it is almost always because the local market is thin or mixed. Knowing that is worth more than receiving a reassuring average.
The rent
A range, its sample — or an admission of disagreement.
A market rent is a statistic, not a measurement. Depending on the source, the same home can be worth a hundred euros more or less, and the average of the three is no truer than any one of them. Pilot cross-checks, displays the spread, and stays silent when it is too wide.
A single figure without a range is a convenient fiction.
The available rent sources do not measure the same thing. Some observe listings — and therefore asking prices, not transactions. Others aggregate declared tenancies, with a lag of several months. Sample sizes vary enormously by town: plentiful in a major city, almost non-existent for a three-bedroom flat in a town of three thousand people. Displaying a round number without saying where it comes from or how many references it rests on gives an assurance the data does not carry — and it is on that assurance that a landlord sets a rent, turns down a tenant, or decides to sell.
- Several sources cross-checked, not a single one taken as is
- A range, and the size of the sample behind it
- Disagreement between sources is displayed, not silently averaged
- The gap to your current rent is calculated, without concluding that a right exists
In practice.
The property is characterised first
Town, property type, floor area, furnished or not. A market rent without a property type is an average of everything, which is to say information about nothing.
Several sources are queried
Each with its method and its vintage. We do not take the first one that answers: comparing sources is precisely what tells you whether the result is solid.
The spread is measured
If the sources converge, the range is narrow and the estimate is usable. If they diverge beyond a threshold, the result is marked as inconclusive — and the values set aside are shown, not hidden.
The sample accompanies the figure
The number of references the estimate rests on is displayed. A range built on five observations and a range built on four hundred do not read the same way, even if they are the same width.
The gap to your rent is calculated, and nothing more
Above, below, by how much. What you can do with it belongs to the other question — the legal one — and Pilot never mixes the two.
What it changes
You see what the figure is based on
Sample size and spread are displayed with the value. That is what lets you decide whether to use it to adjust a rent, or only to get an order of magnitude.
The estimate feeds the alerts, but cautiously
A rent clearly below the market triggers a signal. When the estimate is unreliable, that signal is explicitly marked as indicative and carries less weight — rather than pushing towards a decision on a weak basis.
The law, and the pitfalls
A landlord who invokes a manifestly undervalued rent at renewal must produce references to rents usually found in the neighbourhood for comparable homes, in a minimum number set by law.
art. 17-2 of Law No. 89-462 of 6 July 1989
A statistical estimate is not a reference within the meaning of this text. References must relate to identified, comparable homes — a market average, however well calculated, does not meet the condition, and the whole increase falls.
In areas applying rent control, the enforceable cap is the upper reference rent set by prefectoral order, not an observed market rent.
art. 140 of Law No. 2018-1021 of 23 November 2018 (ELAN)
A market rent above the cap authorises nothing. Confusing the two notions — one descriptive, the other regulatory — leads to setting an illegal rent while believing you are following the market.
Data published as open data may be reused, including for commercial purposes, provided the source and the date of the last update are stated.
art. L. 321-1 et seq. of the French Code of Relations between the Public and the Administration
The date counts as much as the source. A figure presented as “current” when it relates to a two-year-old vintage is misleading even if the data is genuine — which is why the vintage is displayed.
What Pilot refuses to do
Pilot refuses to put forward a figure its sources contradict
Beyond a divergence threshold, the estimate is marked inconclusive and the values set aside are shown. Choosing the most flattering source, or averaging them, would produce a number with no statistical meaning.
Pilot does not display a rent without a range
A single figure suggests a precision the data does not carry. The width of the range is information in its own right: it tells you whether the local market is readable.
Pilot does not present an estimate as a legal reference
For an upward renewal, the law requires references to identified comparable homes, not a statistic. Suggesting otherwise would lose you the procedure.
art. 17-2 of the Law of 6 July 1989
Availability
Included from the Essentiel tier.
- 7-day trial at launch, card required
- No commitment
Frequently asked questions
Where does the estimate come from?
From several cross-referenced sources, each with its method and its vintage, both displayed. We do not take a single source as it stands: it is the comparison between them that tells you whether the result holds.
What happens if the sources disagree?
Beyond a divergence threshold, Pilot shows that the estimate is inconclusive and displays the values set aside. It is almost always the sign of a thin or uneven local market — useful information in itself.
Can I use this estimate to increase my rent?
It tells you what the property would be worth; it does not tell you whether you are entitled to it. And for an upward renewal, the law requires identified comparable homes as references, not a market statistic.
Why a range and not a figure?
Because a market rent is a statistic, not a measurement. The width of the range, and the size of the sample behind it, tell you whether you can rely on it to decide.
Right next to it.
The rent
The market says one figure. The law says another.
Whether a review is available or not, the date of the next one, the constraints detected with their legal grounds, and what becomes possible on re-letting. It is the distinction missing everywhere else.
Deadlines and compliance
The legal cap, at the right granularity — or a refusal to conclude.
Eight municipalities and inter-municipal authorities, at district × period × rooms × furnished granularity. Where there is ambiguity, Pilot refuses to conclude rather than pick a cap at random.
The property
You enter an address. The property fills itself in.
Enter an address: land registry, DVF price per m², natural hazards and value estimate arrive on their own. No integration work on your part.
Performance
It does not just track: it spots what can be improved.
Rent below market, charges rising over twelve months, yield below the rest of the portfolio, EPC soon banned from letting, loan above market.
Content reviewed on 2026-08-06.
Get started
Add your first property.
At launch: a 7-day trial on your real properties, card required.