Before deciding where AI should go, we drew the whole machine
Several service lines, 34 interviews and one shared map of the commercial process, to find out where artificial intelligence would actually change something.
The challenge
The client distributes technical equipment to industrial companies and, around that equipment, provides installation, commissioning and maintenance services. The business is not won on the catalogue; it is won on the process that runs from a customer raising a need to the order being delivered and working.
That process is a long B2B sale with many hands in it. A request does not arrive as a closed order; it arrives as an enquiry that has to be translated into a technical configuration, quoted, reviewed, adjusted when the customer changes the scope, approved, ordered from the factory, coordinated with logistics and finished off with an installation. In between sit commercial, technical, logistics, service and administrative areas, each with its own tool and its own criteria.
The company wanted to bring artificial intelligence into that journey. The real problem came first: nobody held the whole process in their head. Each area knew its own stretch in detail and had little idea what happened before or after, and a good part of the actual work happened outside the corporate tools, in spreadsheets, emails and informal arrangements.
Without that map, any decision about where to apply AI would have been a bet: the visible would have been automated instead of the expensive.
Approach
We started from the journey the company already had and turned it into a service blueprint, which is the same journey seen from behind: what has to happen backstage for the customer to get what they get.
We worked on one common process, the same for the whole company, and on top of it broke down how each service line differs. The processes look alike but are not the same, and those differences are exactly where the errors gather.
- 34 in-depth interviews with the people who do the work, not only with those who run it
- One common reference process, with the variants of each service line laid over it
- A clear split between what the customer sees, what happens backstage, and the support processes underneath
- Particular attention to the handovers between areas, where information gets lost in a B2B sale
- A shared priority scale for every finding, separating quick wins from structural change
- The draft refined with the same people who were interviewed, to validate the map before drawing conclusions
- A roadmap keeping immediate improvements apart from the medium and long-term plan
Key outcomes
1
common reference process, with each service line's variants laid over it: for the first time the whole journey fitted in one place.
34
interviews turned into a living map, then refined with the interviewees themselves.
120+
support processes documented step by step that had never been written down anywhere.
150+
improvement initiatives identified and prioritised, separating what could be fixed in weeks from what needed structural change.
And something nobody had planned for: the map put numbers on blind spots the organisation was not measuring. Recurring errors showed up in order entry, in invoicing and in coordination between areas, and none of them appeared in any report before the project.
Why this counts as a result
This project was a diagnostic, so what it delivered was not savings already banked but the basis for making them: a shared map, a prioritised portfolio of initiatives, and a few uncomfortable numbers where there had only been competing opinions. With that on the table, the conversation about where to apply artificial intelligence stopped being a wish list and became a decision with criteria.
Got a process that hurts?
Tell us about it in half an hour. No forty-page deck. If we are not the right people, we will say so and point you towards someone who is.