Automated Vehicle Recommendation Engine

Velvis

Velvis Cars is a small, fast moving team operating in a highly competitive used car market. Their buyers were spending hours each day reviewing up to 500 auction listings, checking every vehicle against specific purchasing criteria, cross referencing details across multiple platforms and manually assessing commercial viability.

The client problem

The process was slow, inconsistent and heavily dependent on individual judgement. With the volume of listings increasing and pressure to act quickly, Velvis needed a way to streamline decision making and improve the accuracy of their buying process without expanding the team.

What we found in discovery

Our discovery phase revealed several opportunities to strengthen and standardise the buying operation.

Key insights included:

The team relied on several different websites and platforms to validate the details of each vehicle. This created unnecessary friction and wasted time.
Purchasing rules were not formally documented. The process depended on the experience, memory and intuition of a small group of buyers, which increased the risk of missed opportunities and inconsistent decisions.
Subconscious bias was influencing buying patterns, leading to certain vehicle types being overlooked even when they met ideal purchasing criteria.
There were recurring inaccuracies in auction listings that could be flagged automatically. Identifying these errors early represented a significant commercial opportunity since other buyers would not have this insight.

These findings highlighted the need for a standardised, automated and market responsive buying system.

Services provided

We designed and delivered an Automated Vehicle Recommendation Engine tailored specifically to Velvis' operational needs.

1. Standardised Purchasing Rules

Analysed historic buying data to identify consistent patterns and decision drivers.
Documented a clear set of purchasing criteria to assess every auction listing objectively and consistently.

2. Automated Listing Assessment

Built an automated process that evaluates each listing against Velvis' purchasing rules.
Added accuracy and validity checks to highlight inconsistencies or incorrect listing data.

3. Weighted Purchasing Algorithm

Created a bespoke algorithm that blends Velvis' preferences, commercial criteria and historical performance.
Enabled ongoing adjustments so the system adapts as the market evolves.

4. Daily Ranked Shortlist Delivery

Automated a daily shortlist of the top 50 cars.
Ranked by strength of match against the algorithm, ensuring only the highest scoring opportunities reached the team.

Outcomes

Velvis now operate with a consistent, scalable and data driven buying process that strengthens decision making, reduces bias and uncovers profitable stock opportunities that competitors overlook. The automated buying system delivered significant operational improvements within the first three months of deployment:

At least 70 hours of manual sifting work saved every month
The automated process has delivered a 40% increase in vehicles purchased
Increased business revenues by 22% in the first 3 months
Added £250,000 to the bottom line

Ready to streamline your buying process?

Book a free audit or send us a message to discuss how we can help you automate your purchasing decisions and improve accuracy.