Nov 28, 2019
Nov 10, 2023

AI-Powered Pricing: PartsTrader's Improved Tender Process

PartsTrader operates online automotive marketplaces in New Zealand and the USA, playing a pivotal role for collision repairers comparing and purchasing parts.

Essentially the "stock exchange" for automotive parts, their monthly operations involve over 140,000 parts request tenders and millions of parts and quotes from suppliers, resulting in annual sales surpassing $1.5 billion.

The Challenge

For PartsTrader, ensuring customers receive prompt and accurate quotes is paramount.

Given the bustling automotive parts market, predicting the best quote price can be a daunting task. Repairers need unwavering confidence in the provided quotes. The overarching goal was clear: Refine the procurement process, arming repairers with insights to make optimal decisions for their clientele.

Our Approach

  1. Personalised Collaboration: We worked closely with PartsTrader, diving into various statistical and AI methodologies to find the best fit.
  2. Embracing Simplicity: After extensive testing, we identified a straightforward approach that ensured easy implementation for PartsTrader, rivalling even the more complex machine learning solutions. Despite its simplicity, this method delivered robust results, standing toe-to-toe with more intricate machine learning solutions.
  3. Creating a Dynamic Model: Our design ensured the AI model could be updated daily with new marketplace transactions, keeping the data fresh and relevant.
  4. Providing a Future Vision: We charted a clear roadmap highlighting how integrating new data sources could further refine prediction accuracy. This not only gave PartsTrader an immediate solution but also a path forward.
  5. Addressing Historical Challenges: Recognising past hurdles, we specifically designed the model to filter out extreme pricing data, which had been a recurring challenge.

The Outcome

The synergy of AI and statistical techniques has now enabled PartsTrader to provide faster, pinpoint-accurate quotes.

This pivotal transformation not only streamlined their tender process but also fortified the confidence of repairers during the procurement phase.

Key Client Benefits

Efficient Quote Predictions

Leveraging statistical and AI techniques, we significantly enhanced the accuracy of quote predictions. This saves both time and money for repairers, ensuring optimal part selections for their customers.

Streamlined Pricing with Outlier Detection

Our model offered quick and efficient pricing. By identifying and excluding outlier data such as extreme prices, we reduced potential inaccuracies, enhancing trust and satisfaction for platform users.

Adaptable Model for Real-Time Updates

Our design allowed the model to adapt daily, reflecting fresh transactional data. This ensured responsiveness to changing market conditions, keeping PartsTrader ahead of the curve.

Clear Future Roadmap

Beyond the immediate solution, we equipped PartsTrader with actionable recommendations on enhancing the model's predictive power, promising even greater value in the long run