of selected clients profiled from external data
One of the world’s largest commercial vehicle manufacturers asked a simple question — why do customers who show interest end up buying elsewhere? In four weeks, Cleevio turned external data into working prototypes that tell a sales team who to call, and when.
The challenge
Daimler Truck noticed a recurring pattern: companies expressed interest in their trucks and ultimately chose a competitor. The brief was to replace assumptions with evidence — understand what was driving customers away, and what would bring the right ones in. Daimler partnered with McKinsey and Cleevio for the strategy; Cleevio was also the technology partner, responsible for building the tools to uncover and act on the insights.
Two questions framed the work. Retention: how well do we understand the customers we already have, and when will they be ready to buy again? Acquisition: where are the companies that should be buying from us, and how do we find them before the competition does? With thousands of companies in every territory and market developments happening daily, answering either meant more than consulting — it meant collecting, connecting and interpreting data from dozens of sources, while keeping the human relationship at the centre of the sale.
- Thousands of companies per territory, no way to rank them
- Purchase timing left to periodic check-ins and chance
- Service history and sales outreach living in separate worlds
Our solution
Cleevio worked alongside Daimler Truck’s sales and strategy teams as both consulting and technology partner — first exploring what modern AI and data analytics could do, then building it.
Data landscape
Dozens of external sources evaluated; company data, geospatial information, news monitoring, public tenders and company websites selected.
Customer profiles
External data matched against the existing customer base — detailed profiles for about 50% of clients.
Similarity model
Around 50 company characteristics — segment, structure, geography, operations, technology — to find prospects that look like the best customers.
Growth-signal detection
Revenue growth, acquisitions, new branches and facilities correlated with historical truck purchases.
Retention signals
A four-year replacement cycle and service-visit patterns turned into outreach timing and customer-care triggers.
Territory-based profiling
Lead prioritisation frameworks that stay inside data-protection rules.
4 weeksDiscovery and prototyping
~50%Clients matched with external data
50Company traits in the similarity model
Retention
Knowing when a customer is ready — and when to leave them alone.
The first step was the existing customer base. The data showed that many customers replace their trucks every four years — enough to predict when a company is likely to be in the market again, and to plan the conversation before the competitor does.
Servicing data turned out to be just as useful, in the opposite direction. A truck that has been in the workshop repeatedly over the past two months is a customer with a problem, not a prospect. Reaching out with a sales pitch at that moment is tone-deaf; the prototype routes those accounts to customer care first, so the experience improves before a new deal is on the table.
Acquisition
Finding tomorrow’s customers in today’s public data.
New-customer acquisition started as a discovery phase: generate hypotheses, test them against real data. Growth signals from company registries — businesses whose revenue points to fleet expansion. LinkedIn announcements of new branches, which tend to align with new vehicle needs. Geospatial analysis of property expansions and new facilities. Public tenders. Company websites.
On top of that sits a similarity model: roughly 50 characteristics of what makes a successful Daimler Truck customer — industry segment, organisational structure and size, geographic presence, customer demographics, operational patterns and technology usage. Companies that closely match the profile of the best existing customers are the ones most likely to have the same needs, and the same purchasing behaviour.
Early warning
From raw data to a brief the sales team can act on.
The most promising prototype is an early-warning system for sales opportunities. Analysing the history of one long-standing customer revealed sequences of growth signals — revenue increases, acquisitions, recovery periods — that preceded truck purchases. The sample is small and the findings are preliminary, but the mechanism is clear: combine enough sources and the timing of outreach stops being a guess.
Just as important is what the sales representative sees. Not raw data — a daily brief of important developments at the companies in their territory, a prioritised lead list, and alerts when a company shows signs of being ready to buy. The system augments the salesperson’s judgement; it does not replace it.
Results
The research phase delivered working prototypes and first validations — early-stage by design, honest about sample sizes, and explicit about what to build next. It also settled the question Daimler Truck cared about most: AI can make the sales process sharper without removing the people from it.
- About 50% of selected clients matched with external data for customer similarity analysis
- Key business events identified that correlate with truck purchases
- Actionable insights delivered to Daimler Truck’s sales teams to improve targeting and efficiency
- A foundation and roadmap for production-ready AI sales tools
company traits in the similarity model
the replacement cycle the data made visible
focus areas — retention and acquisition — with prototypes for both
Through the tailored support and profound insights from Cleevio AI in the critical second phase of our GenAI project, we were able to make strategically important decisions and successfully chart our course. We are grateful for the outstanding expertise and reliable partnership of Cleevio AI.
Stefan HauserAI & Sales Lead, Daimler Truck
Read next
Awards & recognition
Second opinionAsk someone
Ask someone
who isn't us.
We've pre-loaded the context.
Ask what we're good at — and what we're not.
Opens a new chat, context already written.
Talk to us
Our AI sales agent knows the work cold — what we build, how a POC runs, what it'd cost for your case. Ask her now; she'll follow up by email to get you started.