Fragmented adoption
AI interest existed in pockets across the regional teams, with no shared roadmap or governance connecting them.
Manufacturing and bottling
Aligning a multi-market group before releasing AI investment
Executive enablement and a governed roadmap, top down
A group operating across several regional business units had AI interest in pockets and no shared view of what it was for. Before any investment could be committed, leadership needed one definition, one set of priorities and one route from awareness to implementation.
- Sector
- Manufacturing and bottling
- Footprint
- Several regional business units
- Engagement
- Executive enablement and AI roadmap
- Pillars
- Sovereign · Productive
Client identity withheld. Sector and footprint are generalised to preserve anonymity. A reference call with the engagement sponsor can be arranged.
None of it was technical
Interest in AI had appeared independently across the regional business units, which is normal and not in itself a problem. What was missing was anything connecting those pockets: no shared roadmap, no common definition of what AI was for in this group, and no governance model that would apply when something moved from experiment to production.
That made investment impossible to commit with confidence. Senior management were being asked to fund something that meant a different thing in each region — and when a proposal cannot be compared against another proposal, the honest executive answer is to wait. The delay was not caution about AI. It was the absence of a common basis for the decision.
The second gap was sequencing. There was no structured route from initial awareness to practical implementation, so effort dispersed instead of compounding: several regions doing early work, none of it building on the others, and nothing that could be presented as group progress.
None of the three constraints were technical. No model needed choosing, no infrastructure decision was blocking anything. The constraints were alignment, sequence and governance — which is the pattern in the large majority of groups that tell us their AI programme has stalled.
The challenges
Leadership alignment gap
Senior management needed a common understanding of AI's role before greenlighting group-wide investment.
No phased path
There was no structured route from initial awareness to practical, sequenced implementation across the group.
What we did
- 01
Executive enablement
A working session with group leadership that produced an agreed statement of AI purpose, the priorities that follow from it, and the guardrails the group would operate inside. The output was one page, agreed in the room, rather than a strategy document circulated afterwards for comment.
- 02
Regional workshops
AI literacy and use-case sessions with the management teams in each regional business unit. Two purposes: raise the floor so the conversation was informed, and collect candidate use cases from the people who know where the work actually piles up.
- 03
Roadmap design
Candidate use cases scored on return, margin impact, feasibility and time to value, then sequenced with named owners and stage gates. The scoring mattered less than the fact that everyone could see why one thing was ahead of another.
- 04
Phased execution
The roadmap moved into active follow-through, with the governance model applied from the first phase rather than introduced once something was already running.
What changed
Shared understanding
Leadership and regional management hold one common view of AI priorities and of what happens next. Proposals can now be compared against each other.
A structured path
A phased roadmap, already moving into its next stage, gives the group an ordered route from awareness to implementation.
Governance by design
Scaling proceeds deliberately, with governance built in from the outset rather than retrofitted once something is in production.
If you are in the same position
The tell is that AI is being discussed in several places in your organisation and none of the discussions can be compared with each other. That is an alignment problem, and no amount of technology evaluation fixes it.
Start at the top and keep it short. Half a day with the leadership team, ending in a written statement of purpose, priorities and guardrails, is usually enough to unblock a decision that has been drifting for months. The value is not in the document; it is in the disagreements that surface while it is being written.
Then go wide before you go deep. Regional or functional workshops do two things at once — they raise literacy and they collect candidate use cases from people who can see where the cost actually is. Use cases proposed by the centre are systematically worse than use cases proposed by the line.
And build the governance model at the start, when nothing is running yet and it is a design conversation. Retrofitting governance onto three live systems is the same work done under pressure, with an audience.
Leave with your top three risks documented
Thirty minutes with a senior practitioner. No slideware, no sales engineer.