What it actually looks like when SEE IT, STEER IT, KEEP IT is applied inside a real organisation — not as a case study with a logo on it, but as an honest account of what changed.
The Starting Point
A real client came to us with a familiar problem: master data — the customer, vendor and material records every other system depends on — was spread across teams, owned by no one in particular, and improved only when something broke badly enough to force it. No one could say with any confidence whether it was getting better or worse, because no one had ever scored it in the first place.
This is not a rare situation. It’s the normal state of master data in most organisations that have grown through multiple systems, acquisitions, or simply time.
How the Engagement Actually Unfolded
Not one big reveal — three steps, each building only on what the last one actually proved.
Step 1 — Build shared understanding. The picture as found, with no recommendation attached to it yet. What’s genuinely working, what isn’t, and where the client’s own account of the problem and what the data shows agree — and where they don’t.
Step 2 — Co-design the solution together. Three work-packages, run in parallel, one per element of the method: a Readiness Score design tested against real records, an AI-housekeeping routine proven on one real case rather than promised in the abstract, and a competence gap mapped against the roles the work actually needs — not assumed.
Step 3 — The decision. One recommendation, built on what the first two steps actually found — carried by the client’s own numbers, not ours.
The Numbers Behind It
For the domain in question, the method’s own scoring produced a concrete picture, not a general impression:
- SEE IT — 6 of the domain’s 15 capabilities were readiness-tracked: the ones that actively validate, deduplicate, or govern the data, rather than simply use it.
- STEER IT — 10 of 15 were flagged as AI-housekeeping candidates. True to the method’s own honesty rule, not all 10 came back the same: some already run on a named, current AI feature; others are still handled by deterministic, rule-based logic with no AI behind them yet — stated as what it is, not rounded up.
- KEEP IT — 8 of 15 required certified competence, each with a named accountable role rather than a shared assumption that “someone” was responsible.
What We Could Confirm — and What We Couldn’t
The same discipline that runs through every number on this site ran through this engagement too: findings were marked as confirmed — backed by a source the client had themselves validated — or indicative — named in the material, but never independently established. Where a measure genuinely didn’t exist yet, we said so, and named what it would take to get it, rather than estimating a number to fill the gap.
The Result
Master data went from undocumented and person-dependent to scored, owned, and reviewed on a regular rhythm — the same rhythm the Delivery Model builds into every engagement. The organisation can now say, with evidence, whether the capability is holding — which is the entire point of KEEP IT: not a one-time fix, but a capability that stays seen.
Why We Don’t Name the Client
As a matter of practice, we keep client details confidential unless a client chooses to be named themselves. What we stand behind publicly is the method and what it actually produced — not a logo. If you want to talk specifics, we’re happy to do that directly.
What This Means For You
If this sounds like a domain in your own organisation — master data or otherwise — Digital Transformation and Business Case are where this usually starts. Inside the Method shows the actual tooling behind it, and The Eleven Scores, In Detail shows what SEE IT measures for every domain, not just this one.
Related: Business Case · Digital Transformation · Inside the Method · The Eleven Scores, In Detail · Home
