Property Comparables: How Estate Agents Choose Comparables for a Valuation

Property Comparables: How Estate Agents Choose Comparables for a Valuation

Your recommended asking price is only as strong as the comparables behind it. The Property Ombudsman's Code of Practice expects any recommended asking price or possible selling price to be given in good faith, to reflect current market conditions and to be "supported by comparable evidence" (paragraph 4b).

Vendors are told to ask for that evidence. The government's How to sell a home guide tells sellers to "ask the estate agent for evidence to support their valuation; for example, similar properties they have recently sold".

So which comparable properties belong in that evidence, and which should you leave out?

What makes a good comparable?

A good comparable is a recent sale of a similar home, close by, that a buyer for your vendor's property would also have considered. Six tests decide it.

Property type matters more than many vendors expect. In the North West in Q3 2026, flats took 96 days from first listing to sold subject to contract, against 54 days for terraced houses (Propalt, Q3 2026). In Greater London the gap was 82 days for flats against 60 for terraced homes. A flat priced off house comparables will sit.

Should you use sold prices, asking prices or sold STC prices?

Use completed sold prices as your evidence, asking prices to show the competition, and sold STC listings as a signal of where the market is moving.

• Completed sold prices. HM Land Registry's Price Paid Data records residential sales for value in England and Wales. It is the firmest evidence you have. It is also the oldest.
• Sold STC. These show what is selling now, but the agreed price is not public, so you cannot quote it as fact.
• Asking prices. These show what the vendor is up against, not what buyers will pay.

The Propalt data shows why asking prices make weak comparables. In Q3 2026, 17.3% of new Rightmove listings nationally in the quarter (83,212 of 480,766, counted per listing) were reduced below their first listed price. In Greater London the figure was 19.1% (11,650 of 60,840); in the North West, 14.0% (7,190 of 51,417).

An asking price is a hope. A completed sale is evidence. Present them as different things.

How do you adjust comparables for differences?

Adjust for each material difference between the comparable and the subject property, in pounds, and write down why. The aim is to show your workings so the vendor can follow them.

Common adjustments:

• Time. Bring an older sale forward using the local UK House Price Index movement since its completion date.
• Size. An extra bedroom or a larger plot. Use the difference in local sales of each size, not a rule of thumb.
• Condition. A new kitchen, or a property needing full modernisation.
• Position. Main road frontage, a corner plot, a south-facing garden.
• Tenure and lease. Leases of very different lengths are not like for like. Adjust, or exclude the sale.

Use the index for the same property type, because types can move in opposite directions. In Manchester, the UK House Price Index for semi-detached homes rose 2.9% in the 12 months to July 2026, while the index for flats fell 0.9% (HM Land Registry UK House Price Index, via Propalt). A year-old flat sale and a year-old semi sale need adjusting in different directions.

If a comparable needs large adjustments on three or four counts, it is a weak comparable. Drop it and find a closer one.

How many comparables should you use?

There is no fixed number in the TPO Code or in law. Three to five completed sales that pass the tests above, plus the two or three listings a buyer will view alongside your vendor's home, gives a vendor enough to understand the price without burying the point.

More is not better. Ten loosely matched sales invite the vendor to fix on the highest, which may be the very figure a competitor used to overvalue.

Our tutorial on researching a property before a market appraisal shows how to combine current and past comparables in one report.

Is £ per sq ft a reliable way to value a house?

It is a useful cross-check and a poor method on its own. £ per sq ft works best on similar homes in a uniform stock, such as a development of near-identical flats.

It breaks down in three places:

• Size differences. The rate per square foot rarely holds steady across sizes in the same street, so one rate applied to a two-bed and a four-bed will misprice at least one of them.
• Mixed types. A bungalow and a three-storey townhouse of the same floor area are not the same product.
• Inconsistent measurements. EPC floor area, agent floor plans and vendor estimates often disagree. Use one source for every comparable.

How should you present comparables at the market appraisal?

Lead with the three best sold comparables, show the adjustments, then show the current competition. Keep the format consistent so the vendor can compare like with like.

1. A map. Where each comparable sits in relation to the vendor's home.
2. A table. Address, type, bedrooms, floor area, sale date, sale price, adjustments, adjusted value.
3. The competition. What is live now, at what price, and how long it has been on.
4. Market speed. Days to sold STC locally. Nationally, homes in Q3 2026 took 69 days from first listing to sold STC (Propalt, Q3 2026).
5. Your range. The figure the evidence supports, and where you would launch.

Leave the property report with the vendor. If a competitor quotes 10% more, the vendor now has your evidence to test theirs against.

In Property Reports, comparables and planning data fill in automatically, and you can personalise each section before you print the report or share it as a link. Our tutorial on how to create a seller report walks through it. For where comparables sit in the full document, see our market appraisal template.

What are the most common comparable pitfalls?

Most mistakes come from data that looks like a match and is not.

• New build premiums. Price Paid Data flags new builds ("Y = a newly built property"). First sales on a development often carry a premium that resales do not repeat. Do not use them for a resale without adjusting.
• Off-market sales. A sale that never hit the portals will appear in Price Paid Data with no listing history, photos or asking price, so you cannot judge its condition. Use it only if you know the property.
• Unusual sales. Gifts and other transfers not for value are excluded from Price Paid Data. Repossessions, buy-to-let purchases and transfers to non-private individuals are recorded as category B, separate from standard category A sales. Check which you are using.
• The Price Paid lag. HMLR says registration "typically ranges between 2 weeks and 2 months" after a sale, so data for the two most recent months is incomplete. Your best comparable may not appear yet. Check what your own office and local agents have completed.
• Your own sales only. Using only your own completions looks selective. Include the best evidence regardless of who sold it.

The wait is longer when you count from the day a sale is agreed. In Q3 2026, the median time nationally from sold STC to appearing in Land Registry data was 115 days, and 136 days for flats (Propalt, Q3 2026). It counts only homes that sold below their asking price, and the latest quarter reads low while registrations catch up. A sale agreed in the summer may not be a usable comparable until the winter.

📌 KEY TAKEAWAY

Comparables should be recent completed sales of the same type, size, condition and tenure, as close as possible to the subject property, with written adjustments for every difference. Use asking prices to show the competition and never as evidence: in Q3 2026, 17.3% of new Rightmove listings nationally were below their first listed price. Allow for the Price Paid lag, typically two weeks to two months after completion, strip out new build premiums and category B sales, and present three to five strong comparables in a consistent format the vendor can keep. Property Reports auto-fill comparables into a branded seller report.

For the wider pipeline that gets you to the market appraisal in the first place, see our guide to estate agent lead generation.

SOURCES & DATA REFERENCES

1. Propalt (October 2026). 'Quarterly market data, Q3 2026: national, Greater London and North West.' Share of new Rightmove listings in the quarter (counted per listing, not deduplicated) reduced below their first listed price and days from first listing to SSTC by property type (Rightmove). propalt.co.uk
2. Propalt (October 2026). Quarterly market data, national, Q3 2026: median 115 days from SSTC to Land Registry registration, counting only properties whose asking price was above the sold price; flats 136 days. The latest quarter is biased low by registration lag.
3. HM Land Registry UK House Price Index, via Propalt (October 2026). Manchester (local authority), 12 months to July 2026: semi-detached index up 2.9%, flats and maisonettes index down 0.9%.
4. HM Land Registry (accessed October 2026). 'About the Price Paid Data.' Scope, Old/New flag, category A and B definitions, and the 2 week to 2 month registration lag. gov.uk/about-the-price-paid-data
5. The Property Ombudsman (March 2026). 'Code of Practice for Residential Estate Agents (TPOE27-9).' Paragraph 4b: any recommended asking price or possible selling price must be supported by comparable evidence. tpos.co.uk/wp-content/uploads/2026/03/TPOE27-9-Code-of-Practice-for-Residential-Estate-Agents-A4-March-2026.pdf
6. Ministry of Housing, Communities and Local Government (updated September 2019). 'How to sell a home.' Sellers told to ask agents for comparable evidence. gov.uk/government/publications/how-to-sell-a-home/how-to-sell-a-home

Contains HM Land Registry data © Crown copyright and database right 2026. This data is licensed under the Open Government Licence v3.0.