What you need to know about estimates
A ranking that includes estimates isn't wrong, but it's much softer than it looks at first glance. To read it properly, you have to look at the estimate flag first, figure out the shape of the postcode second, and only look at the final rank last.
When an estimate genuinely changes things
Taxes are the easiest way to see this in action. 1224 ChĂȘne-Bougeries, 1231 Conches, and 1293 Bellevue all sit at 6.9% for a single person earning 80k. All three have the tax.isEstimate = true flag set. You should read that as "these places are in a very low tax band," not as a precise podium finish where one is mathematically superior.
You see the same thing with 1008 Prilly and 1040 Echallens. Prilly splits exactly 50/50 between the municipalities of Prilly and Jouxtens-MĂ©zery. Echallens spreads equally across three different municipalities. In these cases, our ranking is telling you about the general tax climate of that zone. It is not giving you an exact number for a specific street address.
When an estimate barely matters at all
Not every estimate flag carries the same weight. 6052 Hergiswil NW has both estimate flags turned on, but the neighboring municipality of Horw accounts for only 0.199% of its geographic area. Technically, yes, the postcode is split. Practically, it behaves exactly like a single-municipality postcode.
1000 Lausanne 25 offers another interesting mix: the tax figure is an estimate, but the demographic figure is a direct count. You should treat the tax number with caution, but you can rely completely on the population data.
Where to find solid anchor points
If you want a firm baseline to compare everything else against, look at places like 1468 Cheyres or 4123 Allschwil. Neither their taxes nor their demographics rely on estimates.
This same logic applies to demographic shares. 3929 TĂ€sch reports 68.2% foreign nationals as a direct, unestimated number. 8280 Kreuzlingen reports 57.7%, but has demographics.isEstimate = true. The gap between them is real, but the Kreuzlingen number is still an approximation.
What to check before taking a ranking literally
Check the flag first, the size of the gap second, and the rank itself last. An estimated 6.9% isn't automatically worse than a direct 7.0%, but it absolutely requires more caution. And if you're looking at a deeply split postcode like 1008 Prilly or 1040 Echallens, remember you are looking at a combined regional average, not a guaranteed number.
