Agentic sales assistants for automotive lead conversion

Insights

  • Online sales of cars are predicted to reach 30% of the global sales in 2025, up from just 5% in 2022, and are forecast to grow at a CAGR of 12.6% from 2024 to 2032.
  • However, there are challenges to be addressed such as multiple navigation clicks, fragmented information, and unsatisfactory answers to queries, with AI playing a key role to address them.
  • Businesses use generative AI to create new content while agentic AI builds on capabilities to take autonomous decisions and actions.
  • One study reported an AI-enabled dealership achieving a 26% lead-to-sale conversion while the average rate is less than 6%.
  • Most car manufacturers have implemented an online sales channel but need agentic AI to be integrated into them to make the process actionable.

People looking to buy cars are increasingly shopping online. This shift started with the pandemic due to social distancing requirements, but more recently, it is driven by advances in digital tools and artificial intelligence (AI), convenience, pricing transparency, and omnichannel experiences, shuffling between online and physical in-store visits. Online purchase of cars saves time spent at a dealer’s shop for customers, reduces cost by letting them compare financing options, and provides customized experience in choosing a model and configurations. The online sales platform becomes a natural extension of the dealer showroom.

New car buyers reported a record-high 75% satisfaction rate with their shopping experience in the US in 2024. Customers prefer online channels during the early stages of a car purchase, with a shift to in-person communication during the final stages, such as a test drive and vehicle delivery. Online sales of cars are predicted to reach 30% of the global sales in 2025, up from just 5% in 2022, and are forecast to grow at a compound annual growth rate (CAGR) of 12.6% from 2024 to 2032. However, there are challenges to be addressed for this growth — both cultural and technical, with generative and agentic AI playing a key role.

Challenges in online car shopping

Changes in customer preference and online buying trends are a major challenge for dealers, over and beyond macroeconomic and industry level issues such as fluctuations in vehicle supply. If not addressed, these can translate into frustrated customers and missed opportunities for sales conversions at any stage in the life cycle of a car purchase (Figure 1).

Figure 1. Stages in the online car purchase life cycle and potential break points at each stage

Figure 1. Stages in the online car purchase life cycle and potential break points at each stage

Source: Infosys

The problems with online shopping include the need for multiple navigation clicks through a website, leading to repeated visits without satisfaction. Buyers face confusing and inconsistent navigation across websites or applications, fragmented information, unsatisfactory answers to customer queries, and a lack of personalization and context-sensitive continuity of information with virtual assistants and their responses to queries across channels. They spend time unproductively as they sift through conflicting information due to the lack of a centralized, reliable product repository. The experience is further marred by generic virtual assistants that fail to resolve queries comprehensively. Notably, 50% of users report increased frustration with traditional chatbots that provide irrelevant and generic information not tailored to their needs.

To compound these challenges, customers are not always shown relevant videos or images that demonstrate critical vehicle features and accessories, creating further frustration for potential customers. Moreover, leaving crucial purchasing information spread over various third-party sites such as marketplaces and demand aggregators rather than a single platform complicates the process. That makes sales conversions less likely and increases the potential for customers to abandon their purchases, with almost 30% of customers leaving the brand after a negative chatbot experience.

However, emerging technologies like generative AI and agentic AI — advanced AI systems where individual agents can act autonomously and together to retrieve information and make decisions to achieve their objectives — offer opportunities to address these issues. ChatGPT for example can now recall and refer to past conversations with users.

The ability of AI to integrate information from across inputs with omnichannel sales strategies promises enhancements to lead conversion ratios and get more people to buy online instead of dropping off the website without making a purchase.

Challenges in building agents

Building agentic AI has its own challenges. The following are the four categories of challenges in building and implementing agentic systems, and how they can be addressed:

  • Fragmented and inconsistent data: Fragmented, poor quality, and inconsistent data hamper the accuracy of decision making. Lack of access to real-time inputs in a dynamic business that changes by the minute hampers the effectiveness of agentic AI. Continuous access to real-time data is important for agentic AI to be accurate and effective.
  • Lack of workflow alignment: The agentic persona should supplement the existing process flow and not resist or manipulate the process. It should be flexible to adapt to changing scenarios.
  • Unclear prompts: Vague or poorly worded prompts can lead to irrelevant or inaccurate information. Crafting context-aware and effective prompts for each stage of the sales process is crucial.
  • Ambiguous decision autonomy: Drawing a boundary between what the agent does autonomously and where a human should be involved is a challenge. Clear roles and boundary conditions need to be defined as to where the agent should be autonomous and where it should rely on human intervention, especially based on the complexity of the use case. Human oversight on decisions taken by the agent should be in line with brand values and ethics.

AI agents to build better customer journeys

Broadly speaking, businesses use generative AI at least partly to create new content. Agentic AI builds on the capabilities of AI technologies to take autonomous decisions and actions. In automotive sales, generative AI can be used to create personalized content such as vehicle configuration details matching a customer's interests dynamically and at scale. AI agents can go further: Collate vehicle details and present options to a customer, suggest configurations based on the user’s preference, and could even orchestrate sales steps, negotiate a price, and arrange delivery to create an efficient, personalized end-to-end customer experience.

However, at present, most of this AI-enhanced process is limited to the presales stage, meaning the smooth end-to-end journey isn’t created for the customer. To overcome this, the AI solution should not limit its interaction with the customer to just the initial stage of sales, but act as a constant bidirectional bridge between the customer and the sales consultant. For the vendor, this AI-powered automation using AI agents can keep track of customer interactions and trace their feedback at each stage. This would leverage agentic AI’s long-term, persistent memory of customer interactions across channels — web, voice, social media, email — with a feedback loop for reinforcement learning to keep track of customer interactions and analyze them. Agentic AI can constantly evaluate customer expectations and tailor vehicle positioning to offer appropriate choices, while keeping the continuity of the lead intact for a successful conversion.

Adopting this approach can deliver tangible benefits: One study reported an AI-enabled dealership achieving a 27% higher rate in setting showroom appointments compared to the average share of customers scheduling a showroom visit after initial inquiries. The study also reported a 26% lead-to-sale conversion rate for customers who eventually buy a car, while the average conversion rate for automotive dealerships is less than 6%. Their ability to analyze lead behavior and recommend actionable next steps can boost conversion rates, reduce customer dropout, and increase customer loyalty.

Mercedes-Benz for example is using AI to improve its online customer experience by deploying an AI sales assistant that uses natural language conversations to improve customer interaction for services such as queries, to schedule a test drive, or to initiate a purchase process. Research found that 75% of customers appreciated the company’s transformed dealership experience.

Financial services company Capital One has implemented a proprietary multiagentic AI tool to help its customers with the car shopping experience. A multiagentic AI tool is a system where multiple agents collaborate with each other to solve a complex task, typically one agent dedicated to one task. Capital One designed the multiagent system keeping in mind how humans interact with customers – some to provide information, some to take action, and some to evaluate other agents. Deployment of the system resulted in up to a 55% increase in customer engagement metrics in some cases.

Best practice guidelines

To ensure the ethical and effective implementation of agentic AI, adopting a process framework is critical. This framework should be designed and implemented to identify critical break points in the sales process and automate intelligent solutions to maintain momentum and enhance lead conversion. Key components of this process framework include constant engagement with leads to uncover requirements, continuous monitoring of customer progress, analyzing drop-off reasons, and providing actionable insights to close deals effectively.

AI usage must also adhere to responsible AI guidelines to align with regulatory and data security standards, ensuring customer information is handled responsibly. By focusing on regulatory alignments and data security, businesses can instill confidence in their customers that their data will be protected, which is paramount in building lasting relationships. This responsible approach not only facilitates better sales outcomes but also enhances trust and transparency between buyers and sellers in the digital marketplace.

The future of agentic automotive sales

For companies ready to evolve, the key to success is to integrate these AI-driven strategies into their sales frameworks and leverage them to offer novel customer experiences. Potential car buyers in the US see agentic AI as a gamechanger in the car buying experience, with 61% of respondents to a study wanting an agent to find and recommend the best car matching their needs. In the near future, a hybrid approach blending AI assistance with human expertise will be a solution of choice.

Most car manufacturers have implemented an online channel for new car sales, along with AI. But these channels provide only content and answer queries. Agentic AI should be integrated into them to make the process actionable, with agents executing tasks such as fixing the price and finalizing a car deal. Agentic AI can understand each customer’s unique perspective, not just provide personalized content but take real-time decisions and steer the online sale system toward the objective, which is to convert each lead into a car purchase.

“The true promise of agentic AI will be unlocked when we have interenterprise agents talking to each other across organizational boundaries, eventually transforming into an agentic ecosystem,” says Prof. Mohanbir Sawhney.

Connect with the Infosys Knowledge Institute

All the fields marked with * are required

Opt in for insights from Infosys Knowledge Institute Privacy Statement

Please fill all required fields