How to work out what a house is worth: one home, step by step

Updated 7 October 2026 9 min readBy the Perch team

Working out what a house is worth is not one calculation. It is five questions asked in order, each of which can overturn the answer to the last. Below is a real one, worked all the way through: a three-bedroom semi in Didsbury, south Manchester, asking £625,000. The numbers are the ones Perch published for Didsbury on 7 October 2026, and you can check every one of them on the Didsbury market page. The interesting part is not the arithmetic. It is the two places where the obvious conclusion is wrong.

The home

A three-bedroom semi-detached house in Didsbury, about 1,150 sq ft, on at £625,000. Against the area as a whole that looks expensive. Against other semis it looks like a good one priced at the top of its type. Both readings come from the same page of numbers, and the question is which one survives when you look properly.

Step 1: compare it with its own type, not with the area

The first number most people meet is the area median. In Didsbury that is £385,000, across 145 homes for sale. Against that, £625,000 looks 62% too high, and it would be easy to stop there.

It would also be meaningless. Half of what is for sale in Didsbury is flats, and they set that median. Split by type, the picture is completely different:

TypeFor saleMedian asking
Flat72£257,500
Semi-detached36£525,000
Terraced26£437,500
Detached9£875,000
All homes145£385,000

Our home is not 62% above the market. Semis here ask a median of £525,000, and three-bedroom semis ask between £395,000 and £550,000 for the middle half of the 17 on sale. So £625,000 is above most of its direct competition, but it is in the same conversation, and the £385,000 area figure was never a comparison, it was a mixture. The first thing an area median tells you is what the area is made of, not what a home in it is worth.

Step 2: find out what buyers actually paid

HM Land Registry records every completed sale in England and Wales. In Didsbury there were 274 registered sales in the twelve months to February 2026, at a median of £400,000. That figure has the same mixture problem as before, so again it needs splitting, and the cleanest split is per square foot, because it puts homes of different sizes on one scale.

Matching sales to EPC floor areas gives, for the twelve months to August 2026 (the matched series runs on its own pipeline, so its window ends later):

TypeSalesMedian sold £/sq ft
Flat84£366
Terraced40£463
Semi-detached74£469
All homes, detached included217£436

Now put our home on the same scale. £625,000 across about 1,150 sq ft is roughly £543 per square foot. Semis in Didsbury sold at a median of £469. Across all types, the middle half of sales landed between £359 and £500, so £543 is above the top of that range.

That is a real finding, and it is the one the headline price hid: the home is asking about 16% more per square foot than semis in Didsbury have been selling for.

Step 3: do not treat that 16% as a discount

This is where the worked example gets genuinely useful, because the tempting next move is to offer 16% under and call it evidence based. That would be misreading the number in three ways at once.

  1. The two figures describe different moments. A sold price is a completion. It was agreed months before it completed and registered weeks after, so a twelve-month window ending in August 2026 is describing offers accepted across roughly mid-2025 to mid-2026. The £625,000 is today.
  2. They describe different homes. The 74 semis that sold are not the 36 that are for sale. If the ones that sold were smaller, more tired, or further from the village, the gap is partly just a difference in what is being measured.
  3. The floor area may not be the same measurement. The sold figures use EPC floor areas; the 1,150 sq ft came off a floorplan. Those two methods routinely differ by a few per cent, and a few per cent on the denominator moves £543 more than it looks.

So the 16% is not a target. It is a question: is this home better than the ones that sold, or is it priced as though it is? That question is answered by viewing it, and by the next two steps.

Step 4: check which way the market has moved since

A gap against past sales matters much less in a rising market than a flat one, because the past sales are stale in a specific, measurable direction. So look at the same measure over time, for the same type. Median sold £/sq ft for semis in Didsbury, at successive twelve-month windows:

12 months toSemi-detached £/sq ft
December 2020£375
December 2022£427
December 2023£446
December 2024£430
June 2025£433
June 2026£465

Two readings, and both are worth having. Since the end of 2020, semis are up about 24%. Over the last year, they are up about 7%, after a dip in 2024. A market rising 7% a year explains part of the gap between £469 and £543, because the sold figure is describing a market several months behind the asking one.

It does not explain most of it. Several months of a 7% a year rise is worth a few per cent, so well over half of the 16% survives the trend, and that is the number actually worth negotiating about.

Step 5: check whether the market is accepting today’s prices

The last question is the one buyers most often skip, and it is the cheapest to answer. If asking prices in an area are routinely cut, then asking prices in that area are routinely optimistic, and you have a good idea of by how much.

In Didsbury, in the 30 days to 7 October 2026:

  • 24 of 145 listings had their asking price reduced, up from 13 in the previous 30 days. That is about one in six of everything on the market, in a month.
  • The typical reduction was 4%, or £22,500. That is what sellers here actually concede, as opposed to what buyers hope for.
  • 55 homes came off the market and 49 new ones arrived. Stock is shrinking slightly, which is not the pattern of a market where sellers are desperate.

Read together those say something quite specific: prices here do get corrected, but in single-digit steps, and homes are still leaving the market faster than they arrive. That is not a market that will hand you 16%.

Putting it together

Nothing in the five steps produces a valuation, and any tool that claims to produce one from this data is overstating what the data can do. What they produce is a position you can defend:

  • The asking price is at the top end for its type, and the area median was never the right comparison.
  • Per square foot it is at the expensive end, above the top quarter of what sold.
  • A rising market explains a little of that, but most of the gap is still there.
  • Reductions here run at about 4%, so a first offer somewhere in that territory is a number the seller has seen before and will engage with, and one you can explain in a sentence.
  • The single biggest remaining unknown is whether the home is actually better than the semis that sold. That is what the viewing is for, and now you know exactly what to look for.

The thing worth taking from this is the order. Type before area, sold before asking, direction before gap, and local behaviour before any rule of thumb. Skip a step and the number you land on will be confident and wrong.

Do it on your own home

Every figure in this article is published, dated and free to check. The price check runs steps one to three on any home for sale in England or Wales, and the area market pages carry steps four and five for the area around it. If you would rather Perch watched a whole area for you and told you when something worth looking at appeared, set up a search.

See the numbers for an area

Hackney, Lewisham, Manchester, Leeds, Cardiff, or find your area.