Transformation governance that people actually use
Most large transformations have plenty of governance. Steering groups, working groups, weekly status, a RAID log, a monthly report pack. What they lack is decisions.
That gap — lots of governance, few decisions — is the most common pattern I see when I am brought into a programme that is not going well. The instinct is to add more oversight. It almost always makes things worse.
The test for whether governance is working
Ask one question: how long does an open decision sit before it is made?
Track the average age of open decisions over time. If it is stable or falling, governance is working. If it is growing, it is not, and no amount of reporting quality will change that. Decisions ageing out is what "the project is drifting" actually consists of.
Almost nobody measures this. It takes a column in a spreadsheet and it tells you more than the entire status pack.
Three properties of governance that works
1. The steering group can actually decide
A steering group that has to refer upward is a reporting meeting with an expensive attendee list.
Populate it with the smallest set of people who can commit: the business owner accountable for the outcome, the budget holder, the senior technology owner, the programme lead. Give it a written mandate stating what it can approve without escalation — scope changes up to a threshold, budget reallocation within a limit, go/no-go on milestones.
If a decision consistently escalates past the steering group, either the mandate is too narrow or the wrong people are in the room. Both are fixable in an afternoon and both are usually left for months.
2. Reporting short enough to be read in full
A forty-slide pack does not get read. It gets skimmed for the RAG status, and the RAG status is the least reliable thing in it.
A useful report is two pages:
- Decisions required, each with an owner and a date.
- Risks that could change the outcome — not the full register, the few that matter now.
- Value delivered against the business case, in the business case's own measures.
- Confidence in the next milestone, stated plainly in a sentence.
Activity completed can go in an appendix. It is the least decision-relevant content in most reports and typically occupies the most space.
3. Escalation with a name and a clock
"Escalate to the steering group" is not an escalation path if the steering group meets in three weeks.
A working path names a person and a time limit: unresolved after five working days, it goes to a named individual who can decide within two. Without the clock, escalation is a queue.
Reporting on value, not activity
The most common reporting failure is measuring the wrong thing. Programmes report milestones hit, tickets closed, workstreams green — all activity, all necessary, none of it evidence that the business case is being delivered.
Value reporting requires the business case to be decomposed into measures with an owner, a baseline and a target, before delivery starts. If it was not, that is the first piece of work, and it is usually revealing: business cases frequently contain benefits nobody will own.
A benefit without a named owner is not a benefit. It is an assumption.
| Report line | Activity framing (weak) | Value framing (useful) |
|---|---|---|
| Migration | "80% of records migrated" | "Service desk operating on new platform for 3 of 4 regions" |
| Training | "450 users trained" | "Adoption at 71% of active users, target 85% by Q3" |
| Integration | "12 of 15 integrations built" | "Order-to-cash flow running end to end in production" |
The right-hand column is harder to write. That is the point: it forces someone to state what has actually changed for the business.
Cut the number of forums
Governance sprawl is real. Programmes accumulate forums — a steering group, a design authority, an architecture board, a change board, a delivery stand-up, a business readiness group — each with its own pack and its own attendee overlap.
The cost is not the meeting time. It is that decisions get routed between forums, and every hand-off is a delay and a chance for the decision to be reframed.
Map every forum, its mandate and its membership on one page. The duplication is usually obvious and the merge is usually uncontroversial once it is visible.
Where tooling helps, and where it does not
Governance fails for structural reasons — unclear mandate, no decision owner, activity reporting — and no tool fixes any of those. Buying a platform to solve a governance problem is a well-worn way to spend money without changing the outcome.
What tooling does well is remove the friction once the structure is right: keeping decisions, risks, actions and value measures in one place so status is assembled from live data rather than rewritten every month, and so the age of an open decision is visible without anyone compiling it.
That is what we built TransformRadar to do — one EU-hosted workspace for planning, governance, programme health and value realisation, designed around how transformations are actually run rather than as a generic tool with a governance label attached. It reflects the practice described above; it does not replace it.
Get the structure right first. Then make it cheap to sustain.
The short version
Governance is working when decisions get made quickly by people with the authority to make them, and when reporting tells you whether the business case is being delivered.
If you want one change: start tracking the age of open decisions this month. It will tell you more about your programme than the next three status packs.
Frequently asked questions
- How often should a transformation steering group meet?
- Monthly for most programmes, fortnightly during a critical phase such as go-live. More frequent than that and the group becomes an operational forum; less frequent and decisions queue up beyond the point where they can still influence the outcome.
- Who should sit on the steering group?
- The smallest group that can decide without referring upward: the business owner accountable for the outcome, the budget holder, the senior technology owner, and the programme lead. Everyone else attends by invitation for specific items. A steering group that cannot decide in the room is a reporting meeting.
- What should a status report contain?
- Decisions required with owners and dates; the small number of risks that could change the outcome; value delivered against the business case; and a plain statement of confidence in the next milestone. Activity completed is the least useful thing in most status reports and usually takes up the most space.
- How do you measure whether governance is working?
- Count how long decisions sit open. If the average age of an open decision is growing, governance is not working regardless of how good the reporting looks. It is the single most diagnostic measure available and almost nobody tracks it.