article

Would you trust an AI bot to filter your sales leads? SAIC did.

Martyna Kozlowska
Marketing Manager

...and thanks to AI the company generated ¥800M ($112 million).

Obtaining new leads is every automotive company’s goal. But having many leads means taking care of and turning them into customers.

That is why SAIC Motor has deployed AI in a powerfully practical way to solve this very problem… and the results speak for themselves.

Through its SAIC-GM-Wuling venture, the company introduced AI bots with a straightforward yet highly effective mission: intelligent lead screening.

The engagement and qualification bot

Let's take a step back for a moment and look at how this works. AI intervention begins with a sophisticated chatbot deployed across SAIC Motor's websites.

  • Engagement: This tool provides immediate, 24/7 support by answering product-specific questions (e.g., "What is the battery range of ...?"), drawing information from a constantly updated knowledge base.
  • Qualification: It actively works to differentiate casual browsers from serious prospects by asking targeted qualifying questions. These could include, "Are you interested in scheduling a test drive this week?", "Would you like me to connect you with a specialist to discuss trade-in options?", or "Can I help you get pre-approved for financing?"
saic chatbot AI

Lead screening bot

The next step in the process is simple but rigorous.

Every incoming lead is contacted by an AI bot three times within a two-day window. If a lead is unresponsive or deemed invalid, it's immediately filtered out and no longer passed to the sales team.

This simple step has dramatically improved both the quality of leads and the efficiency of their sales consultants.

saic chatbot AI

"The initial chatbot is a classic application of a Retrieval-Augmented Generation (RAG) model. It queries a vectorized knowledge base (product specs, FAQs) to provide instant, accurate answers, effectively handling the top of the funnel.


However, the key is the chatbot’s ability to pivot from information provider to a qualification agent using targeted questions to gauge intent. We know that this crucial first filter is hard to achieve, but doable, as we worked on it ourselves at Reffine." says Tomasz Badowiec , CTO at Reffine"

Tomasz Badowiec CTO, Reffine

But the AI's job doesn't end there. For leads that consultants mark as "invalid," a human-led call center steps in for a final reactivation attempt, ensuring no potential customer is left behind.


Results

The impact? This AI-driven system was a key factor in SAIC-GM-Wuling achieving an incredible ¥800 million ($110 million USD) in e-commerce revenue in 2024.


Bigger picture

This tactical success is built on a deeper, group-wide strategy.

At SAIC-Volkswagen, the company has been building its foundational data capabilities since 2016. By creating a robust Customer Data Platform (CDP) that unifies customer, vehicle, and business data, they are enabling true "thousand-person, thousand-face" (千人千面) marketing.

This means that the company adheres to a user-centric approach and leverages massive data to deeply understand user preferences and deliver personalized experiences.

Currently, SAIC’s supply chain ecosystem includes 2,300+ global suppliers and 1,100+ dealers, with over 20,000 daily platform users and tens of millions of data exchanges.

Rather than one-size-fits-all marketing, users see uniquely relevant marketing messages, offers, and product suggestions in real-time.


Trend

SAIC’s implementation of AI bots aligns with wider trends: many enterprises report 25–50% improvements in lead conversion rates and marketing efficiency, with AI chatbots and automated systems active 24/7 to generate faster lead qualification and higher order values.

"Data enrichment and scoring is a must have in the modern automotive industry. AI will drive substantial disruption in that area and we are going to be part of it in the MENA region. You can already see that with giants such as Clay (3.5B valuation), Toma, a new automotive startup in the US.


We are building data validation, enriching with AI voice agents that are able to handle calls with clients for test drives but also follow ups or technical interviews with much higher quality, personalisation, and scale than was imaginable even 2 years ago.


To give you an example, AI can call a client to book a test drive and have a completely different vibe focused on their family when calling an Indian customer from Dubai or a classy, calm, super polite Arab individual from Saudi Arabia,”

Adam Wardach CEO, Reffine
Stay ahead

Subscribe to our newsletter and stay up to date with AI in Automotive

Every two weeks our analysts sift through both mainstream English white papers and the harder-to-find Chinese-language articles and reports, so you get a full 360° view of how AI is reshaping the automotive world. Want it delivered straight to your inbox? Subscribe now on LinkedIn and stay a step ahead: