How Make's AI helps Stellantis &You UK auto-close 18,000 customer messages

• Industry: Manufacturing – Automotive

• Use case: Customer support automation

• Country: UK

• Company size: 1,001-5,000

• Apps used: Knack, OpenAI, ClickSend

Stellantis&You UK, a retail division of Stellantis, is one of the largest UK manufacturer-owned car dealership groups, overseeing 30+ locations in the United Kingdom. With an overflow of customer messages and phone calls, they quickly encountered a critical issue. Their Net Promoter Score (NPS) had declined rapidly. Using the power of Make’s automation, the Stellantis &You UK team built a two-way messaging process using ClickSend (SMS service) and Knack (no-code/low-code app builder) Make modules, enhancing it with OpenAI-powered sentiment analysis. This workflow not only automates essential customer interactions, such as booking confirmations and reminders, but also analyses clients’ intent and sentiment for more accurate support. Thanks to AI, Stellantis &You UK significantly reduced non-value-adding work, allowing the Aftersales department to respond more quickly to incoming messages and focus on issues that affect overall customer satisfaction. Conor Power (Digital Process Improvement Manager) and Daniel Cheng (Digital Systems & Database Engineer) explain how Make helped them analyze over 47,000 customer messages, auto-close over 18,000 of them, and massively improve the decreasing NPS.

As someone who was selling cars just four years ago, I never imagined I’d be crafting AI-supported communication solutions! That’s the power of Make. Making complex technology accessible to everyone, helping them grow their business.

Daniel Cheng, Digital System and Database Engineer

The challenges

As Stellantis &You UK grew, their team soon realized their customer support resources weren’t robust enough.

Communication challenges: Before the two-way messaging solutions, clients had two means of communication with the Aftersales Departments. By a phone call or in person. The overwhelming volume of calls left the team struggling to balance support with their other responsibilities, leading to significant delays and frustrated customers. As a result, Stellantis &You UK’s Net Promoter Score (NPS) dropped noticeably, with a lack of communication as the top complaint.

Resistance to automation: Stellantis &You UK employees were unfamiliar with automation, leading to initial resistance and skepticism. Many trusted traditional IT systems more or feared that the automation software would be too complex. The Stellantis &You UK team needed a simple, user-friendly solution that was accessible even to those without technical expertise.

Neither Conor nor I have a technical background,  yet we’ve implemented powerful AI solutions thanks to the available tools. For example, I was able to quickly test, engineer, and version control different AI prompts for our messaging analysis. Currently, we are on version 13 of the prompt, which has remained unchanged since its deployment a year ago and continues to accurately instruct AI on how to analyse messages correctly.

Daniel Cheng, Digital Systems and Database Engineer

Compliance in a consumer-facing industry: Beyond the operational challenges, deploying AI-powered messaging in the UK automotive sector meant navigating the UK GDPR. Routing customer messages through a third-party AI model raised questions that all procurement and legal teams would ask before any solution goes live: where is the data processed, who can access it, and could it be used to train an AI model? Stellantis &You UK needed answers before they could build.

The solutions

Make helps analyze incoming messages and determine whether the Aftersales Department should be notified. This is done by using AI to recognize customers’ sentiment.

AI-powered context analysis

Using Knack (a “no-code” app-building tool), ClickSend, and OpenAI, Stellantis &You UK analyzes the context of inbound messages from customers and improves the communication process for service and maintenance bookings. 

Thanks to advanced AI intent classification within their two-way SMS solution, the automation determines whether the Aftersales Department requires notification of the customer message. If not, the AI concludes the conversation by marking it as “No response required”.

AI’s ability to grasp subtle meaning differences is why we use it in our workflows. With Make, our AI distinguishes similar messages based on context. For example, if an automated status update says, 'Your booking has been confirmed' and the user replies 'Okay', no response is needed. But if a message about vehicle work goes out and the customer replies 'Okay', the AI correctly interprets it as approval. This nuance ensures customer decisions are clearly conveyed to Aftersales Department.

Conor Power, Digital Process Improvement Manager

Context-aware sentiment recognition

The AI within Stellantis &You UK’s messaging solution recognizes dissatisfaction, even if it’s not expressed explicitly. By recognizing subtle language nuances and underlying sentiment, the AI knows when the customer is upset and notifies the Aftersales Team. This ensures that the team can proactively address concerns before they escalate.

Most automations can catch clearly negative messages like 'your service is terrible.' But Make’s AI goes further. It doesn’t just look for negative words. It reads between the lines. Even if a message doesn’t sound bad on the surface, like 'thanks, took you long enough,' the AI can still pick up on the tone and context and flag it to our team.

Conor Power, Digital Process Improvement Manager

A compliance-first approach to AI

The team started with low-cost pilots using no real customer data, testing AI models before any personal information entered the picture. OpenAI qualified and proved the right fit in terms of accuracy and cost.

Once that was decided, Stellantis &You UK engaged their GDPR consultants, Merrion Data Consultants, to run a full Data Privacy Impact Assessment. The assessment covered data processing, potential risks, mitigation strategies, and a review of OpenAI's GDPR and SOC 2 compliance documentation.

Data residency was another formal part of their vendor evaluation process. Stellantis &You UK prioritized UK or Ireland data storage, and a supplier's inability to offer either would have been a deciding factor against them. OpenAI's support for UK and EU data residency meant it cleared that bar.

On model training, an enterprise-level agreement with OpenAI guarantees that their data is never used to train AI models.

"From the off, we knew our data and messages would not be used to train the AI models. OpenAI and other enterprise AI providers offer similar assurances to businesses."

David Male, Head of Transformation at Stellantis &You UK

Make's built-in data anonymization capabilities add a further layer of control for scenarios where minimizing personal data before processing is required.

"We know there are modules in Make that support data anonymization too."

David Male, Head of Transformation at Stellantis &You UK

The results

Thanks to Make’s AI capabilities, the Stellantis &You UK team managed to: 

• Make “Communication” one of the main satisfaction reasons for the company’s top Net Promoters (score 9 or 10)

• Analyze over 47,000 messages and automatically close over 18,000 of them using AI over 12 months. 

• Based on a 30-second review time, save 151 hours of non-value-added work by auto-marking messages as “No action required”.

• Stellantis &You UK also runs an impressive 700+ Make automation scenarios for their key processes

Why are Make and Stellantis &You UK a perfect match?

Stellantis &You UK runs over 700 active Make scenarios, automating many of its key processes. Despite initial reservations, Make’s adoption rate and usage among employees are high. As Daniel and Conor said, the platform gives even non-technical team members a unique opportunity to build powerful, robust workflows quickly.

According to the Stellantis &You UK team, one of the biggest Make’s advantages is the number of API integrations available, providing a bigger palette of tools to choose from. And they all agree on one more thing. 

It makes daily operations easier.

Make simplifies the complex through its visual power. We are really excited about what comes next. The platform is transforming our journey day by day, and we still aren’t done building our future with it!

Conor Power, Digital Process Improvement Manager