What happened when I measured BG’s asking price accuracy?

Published: 25/11/2021

By Tom Bloomfield.

In recent years I’ve noticed more and more estate agents publishing data within their marketing material that draws attention to the sale prices they achieve as a proportion of their asking prices.

For example, it’s common to see firms who claim to achieve 99% or even more than 100% of their asking prices (over sustained periods and on average), and these figures are presented as reassurance to potential vendors that any valuation of their own home will be accurate.

I like to see agents using data in their marketing rather than flowery language, so I decided to join in by measuring and publishing BG’s performance figures. To do this, I compared the sale prices of the last 30 houses and building plots we sold against their asking prices at the time that the successful offer was accepted.

The results were reasonable, with BG achieving an average of 99.17% of our asking prices. For context the highest value example was 117.08%, the lowest was 91.82% and our average sale price was £619,827. (Data is available on request).

However, as I collected the data, I noticed a number of problems which make the 99.17% figure a misleading measure of the general accuracy of our valuation advice:

1. Significant tranches of critical data are missing
The data excludes those houses which we marketed but which we did not sell, and these houses are most likely to be those where the asking prices were too high, and where our valuation advice was the least accurate.

2. Price reductions
Two of the houses in our data set required asking price reductions before we found a buyer. When these reductions are taken into account our figure drops to 98.86%.

3. Differences between asking prices and our valuation advice
Some of the houses were marketed above or below our estimated valuations. This was to take advantage of price thresholds on property portals or to accommodate the expectations of our clients. In some instances, the adjusted asking prices turned out to be more accurate than our valuations.

4. Price is not the only measure of value
The data includes examples where our clients selected a buyer who was not the highest bidder but who brought other advantages to the table, such as their ability to move swiftly.

5. The market conditions changed during marketing
Due to recent market conditions, every property included in our data set will have increased in value to some extent during their time for sale, some by as much as 10%.

Whilst points two to five above all play their part, point one is the most significant as it clouds our valuation accuracy data sufficiently so that our “percentage of asking price achieved” figure is useless as a guide to our future performance.

Although we go on to sell around 75% of the houses we market, inevitably the remainder includes a disproportionately high number of houses that turned out to be priced too high. These were either removed from being for sale or went on to sell with other agents, normally following price reductions.

The same problem will apply to any estate agent who tries to take the similar measurement, which is why I believe figures showing the percentages of asking prices achieved should be taken with a pinch of salt.

I’m reminded of a marketing quote from a famous Will Ferrell movie which some will recognise:

“60% of the time, it works every time”