How Mahindra Runs Its Test Booking Journey Across Meta CTWA, Website and Voice AI

Our Impact

• Fully automated buying journey, from first click to confirmed dealer slot
• Chat + Voice AI deployed across Meta CTWA, mahindra.com and inbound calls
• Faster go-live on every new vehicle launch, with the intelligence layer reused rather than rebuilt

Company Overview

Mahindra runs some of India's most closely watched vehicle launches, driving strong demand through Meta CTWA campaigns around every new model. The challenge wasn't getting attention. It was what happened after the click.

Challenge: Big Launch Spend, Weak Funnel After the Click

Mahindra's Meta CTWA campaigns brought interest in cheaply and reliably. But leads arrived as raw form-fills and missed calls, with nothing in between the ad and the dealer to tell a serious buyer apart from someone comparing on-road prices in another city. Dealer callbacks routinely lagged behind the buyer's own research, and every new model launch meant rebuilding the qualification journey from the ground up.

What Mahindra needed wasn't a chatbot. It was a buying journey: discovery, comparison, pricing, qualification and booking, running end to end and reusable across every future launch.

Solution: The Buying Journey, Deployed Across Three Channels

LimeChat built the same buying journey (discovery, comparison, pricing, qualification, booking) across Meta CTWA, mahindra.com and inbound Voice AI:

• Demand capture: Every CTWA click opens directly into a live WhatsApp conversation, with the agent already contextualised on the model in the ad. No landing page, no form.
• Discovery, native to WhatsApp: Buyers compare models and variants through interactive carousels, media galleries and quick replies, with on-road pricing for their city. No app download, no redirect.
• Qualification and booking: WhatsApp Flows capture buying criteria through structured in-chat forms. The agent scores intent and completes the test drive booking (date, time, dealer location) in the same conversation.
• Website: The same agent runs on mahindra.com, answering buying questions for high-intent visitors and converting them into booked test drives.
• Voice (inbound): AI voice agents take inbound calls on the questions that decide a purchase, including on-road pricing, variant comparisons, EMI and financing, and fuel-type options, and book the test drive inside the call.

How It's Built

The journey runs on a hybrid architecture: deterministic logic handles anything with a single correct, commercially binding answer, such as pricing lookups, variant specs, dealer routing, slot availability and booking confirmation. The LLM is reserved for judgment calls: intent classification, open-ended comparisons, objection handling, multilingual input. Pricing and specification responses always come from retrieved data, never generated, so a hallucinated on-road price isn't an available failure mode.

Guardrails and Handoff

The agent stays inside the buying journey by design. Out-of-scope questions, competitor comparisons, abusive language, and safety- or religion-adjacent topics are blocked and redirected or escalated. Service and after-sales queries are routed elsewhere, a deliberate scope boundary. Handoff to a dealer happens the moment a customer asks for a human, or when the query needs dealer-side or specialist judgment: negotiated pricing, exchange valuation, financing approvals, fleet enquiries.

Deployment Speed: Built Once, Reused Every Launch

Each new vehicle model used to mean rebuilding the qualification journey from scratch. With LimeChat's Agentic Studio, the intelligence layer sits above the channel, and every new launch reuses it instead of rebuilding it, cutting go-live time dramatically with each successive model (XUV700 to XUV 3XO to Thar Roxx). Mahindra's own team can now add and edit test cases, adjust prompts and journey copy, and modify flow branches without LimeChat engineering involvement.

What's Next

The current build handles inbound voice; the nearest expansion is outbound: dormant-lead reactivation, no-show recovery, and follow-up on stalled bookings. Beyond that, the same intelligence layer is extending to Mahindra Electric, where EV-specific discovery (range, charging infrastructure, battery warranty) is the delta, and to Mahindra Farm Equipment, where voice-led journeys in regional languages are a natural fit.

Conclusion

Mahindra's rollout shows what a conversational buying journey looks like when discovery, qualification and booking are built once and deployed consistently across every channel a buyer shows up on. Every new launch now starts from a proven playbook instead of a blank page.

‍