How me, a nerd, buys a car

There tends to be a lot of factors to consider when you buy a car: style, price, speed, economy, etc. How do you decide?

Narrow down the features you need

First off, if you live in London, you don’t need a car. If you still insist that you do need a car, you don’t really live in London.

We were moving out of London to start living in different places every couple of weeks. We needed to fit all our belongings into this car. Plus my bike. Plus Emma’s bike. It needed to be reliable, and comfortable. We had a budget of max £5,000.

We set to work scouring the internet for models that would fit these space requirements. We eventually narrowed it down to these:

Skoda Yeti, Skoda Roomster, Citroen Picasso C3, Hyundai ix20, Kia Venga

Figure out if you can afford the insurance

Then find a few models that fit the bill

In the UK at least, the brand and model of the car you have is one of the large multipliers on the insurance cost. In other words, though car A may be cheaper than B, car A could cost you more once you add insurance.

I checked this by getting a quote for each model I was looking at:

Works well for the Kia, less well for the Yeti.

Scrape some data

Use a chrome extension to get data across multiple car dealership websites

This will save you hours in setup and is a fantastic tool! I simply need to click the extension, click the elements on the page I wish to capture, and it will find those elements across multiple pages of the same site.

I went through both Autotrader and Facebook Marketplace, with the shortlisted cars, looking for cars available within the London area.

Example of the output from scraping

Analyse the data

As wonderful as data is, if you can’t easily understand it, it is useless.

Insight 1: At a certain age, age no longer affects value

This was a consistent trend I saw across all the models I was looking at.

Yes depreciation hits but it does bottom out.

Insight 2: People really like round numbers

So often, we price things less on value and we gravitate to round numbers. This is the most extreme example but banding was a common effect. I cannot explain why there were no cars in the £5k — £7.5k band though I do expect an error on my part.

Insight 3: After depreciation, mileage is king.

Which is probably unfair, as this doesn’t account for well maintained cars that also have a high mileage — though I expect these are in the minority.

Mileage was the best predictor of cost when looking at all cars in this price range.

Use your newly found knowledge

1. Haggling

Firstly, it helped me haggle as I didn’t need to rely on some vague guesstimation of what the car was worth.

2. Finding gems!

The car we went for in the end was below the average line i.e. cheaper than the average for its age, and we only spotted it as an outlier because of this research and analysis.

3. Prediction of value later

It gave us some confidence as to what sort of price we could sell the car for later.

PS: Still go and see the car in person - and preferably go see a ton of cars. It will help you to get a feel for the size of the cars, as well as notice any unusual elements of the car which may be the reason why it is unusually priced. P.P.S: Always look under the car for excessive rust, and make sure you test drive - to check the clutch length, brakes, and ride. When driving, turn the radio off and just listen. 

Founder of and Product Owner @ dunnhumby; just genuinely interested in a lot of things. Built racecars, built electronics, now building software

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store