E236: Radoslav Chernaev - How Fixing Fitment Data Cut Auto Parts Returns From 10% to 3%

Richard Hill sits down with Radoslav (Rado) Chernaev, CEO of Autofixa, to get to the root of why automotive eCommerce has spent years trailing fashion and tech. 

Rado’s answer comes down to one word: fitment. 

The pair first met at Autofixa Live in Milton Keynes, in a room of 350 people from across the UK automotive parts industry. This conversation picks up where that day left off.

Selling car parts online is a different game. A fashion retailer deals with a product, a size and a colour. An auto parts retailer is managing 90,000 to 200,000 parts, legacy systems, patchy product data and few or no images, and every buyer is asking the same question: will it fit my car? 

Rado shares the story of one retailer with 500,000 listings and a 10% return rate. A large share of those listings had no fitment data at all. What did exist was around four years out of date and the customer service team was answering compatibility questions by hand. 

Autofixa focused on the 20% of products driving 80% of the returns, and within a year the return rate had dropped to around 3%, with stronger marketplace rankings on top.

This isn’t a polished software pitch. It’s a frank look at how bad data compounds at scale. One error repeated across hundreds of thousands of listings turns into wasted staff hours, lost margin, poor reviews and weaker visibility in Google and in AI tools like ChatGPT and Claude. 

Rado is just as honest about AI itself. Autofixa uses it to scale fitment work, but every result is still checked in-house, because a hallucinated fitment means a returned part and an unhappy customer.

The conversation goes well beyond fitment. Rado explains why buyers are turning to aftermarket and green (refurbished) parts as household budgets tighten. His own hunt for Kia Sorento wiper blades turned up prices anywhere from £35 down to £2. He also argues that OEM brands gain control by selling on marketplaces directly, because resellers are already listing their products, usually with poor data. 

Richard and Rado unpack what a one to two year OEM tender process really involves, from security and GDPR to disaster recovery and financial due diligence. They finish on where automotive eCommerce goes next, as AutoDoc, B&Q and Temu move into parts and more brands go direct to consumers.

If you sell automotive parts or accessories online, you’re weighing up a move onto eBay or Amazon, or you’re dealing with returns you can’t explain, this one’s for you.

Listen to the full episode now and don’t forget to hit subscribe.

Topics Covered: 

00:00 - Introduction: Richard and Rado reconnect after Autofixa Live

03:12 - Why automotive lags so far behind fashion and tech online

09:22 - How Autofixa solves fitment at scale with data providers, automation and AI

12:45 - Case study: 500,000 listings and a 10% return rate cut to 3%

15:37 - How wrong fitment hits reviews, search visibility and margin

18:43 - Tighter budgets, an ageing vehicle parc and £35 vs £2 wiper blades

29:22 - Marketplaces: are brands losing control by selling on Amazon and eBay?

35:43 - Inside a two-year OEM tender process

39:24 - Where automotive eCommerce goes next: DTC, AutoDoc, B&Q and Temu

42:17 - Book recommendation and where to find Rado

Thank you to DataFeedWatch for sponsoring the eCom@One podcast.

DataFeedWatch keeps our clients' data fresh and optimised and lets us group products around what’s selling, what’s profitable, and what we actually want to spend budget on. Easily. No technical hurdles. Get 2 months free here.