The HR transformation that was announced and then stopped

A manufacturing group announced an HR transformation at a town hall. There was a roadmap with eleven workstreams, a steering committee, and a named programme by the end of the week. People were told things would be meaningfully different within the year.

By month five, the job levelling exercise was stuck waiting for role descriptions, the role descriptions were waiting for the organisation review, the organisation review was waiting for a decision the board had deferred twice, and the HRMS selection had quietly become the only workstream anyone could point to. At the next town hall, nobody mentioned the transformation. Everyone noticed.

The announcement is the first mistake

Announcing a change programme before anything has been delivered converts goodwill into expectation, and expectation has a short half-life. From the moment of the town hall, the business is spending credibility it has not yet earned, at a rate set by how long people are willing to wait.

There is a further cost that is easy to miss. Once a transformation has been named, every ordinary improvement gets absorbed into it. A manager who fixes the leave approval route is no longer someone who fixed something; they are a workstream, now subject to alignment, sequencing and reporting. Useful small changes slow to the pace of the slowest thing in the programme.

The alternative is not secrecy. It is announcing changes as they land rather than as they are planned — which sounds less impressive in a town hall and produces a great deal more actual change.

Eleven workstreams is one workstream, eleven times over

HR functions in growing businesses are small. A programme with eleven workstreams, in a team of five who also run payroll, recruitment and everything else, is a scheduling fiction. What happens in practice is that each workstream advances to roughly the point where it needs a decision from someone outside HR, and then stops.

Everything then sits at forty per cent. Forty per cent of a job levelling exercise has no value at all — it is a spreadsheet of grades nobody uses. The same effort spent finishing two things would have produced two working things.

Sequence by dependency, not by importance

The usual sequencing question is which change matters most. It is the wrong question, because in HR the dependencies are unusually rigid and several important things simply cannot go first.

  ORGANISATION STRUCTURE
     v  (who exists, who reports to whom)
  ROLE DEFINITIONS
     v  (what each role is accountable for)
  JOB LEVELS / GRADES
     v  (how roles compare to each other)
  PAY STRUCTURE          PERFORMANCE PROCESS
     v                      v
  PROGRESSION AND CAREER PATHS

  SEPARATELY, AND EARLIER THAN PEOPLE THINK:
  CLEAN EMPLOYEE DATA -> everything above and any system

  ANYTHING ATTEMPTED OUT OF ORDER GETS REDONE.

Businesses routinely attempt pay banding before roles are defined, or a performance process before anyone agrees what each role is accountable for. Both can be completed. Neither survives, because the foundation moves underneath them and the work is done twice.

Employee data deserves its position. It looks like an administrative task and it blocks nearly everything — structure work needs accurate reporting lines, levelling needs accurate titles and tenure, and any system implementation needs all of it. Businesses that treat employee information management as a phase rather than a prerequisite lose months rediscovering the same discrepancies in every workstream.

Pick the first thing for visibility, not for importance

The first change to land sets whether anyone believes the rest. That argues for choosing something quick and noticeable over something significant and slow, even though the second is more defensible on paper.

1. 1

   Something people experience directly

   Leave applied for in a system rather than over WhatsApp. The roster published a week ahead. Payslips available without asking. These are felt by everyone, unlike a competency framework.

2. 2

   Something finishable in six weeks

   Long enough to be real, short enough that the people who heard about it are still paying attention when it arrives.

3. 3

   Something that does not need a board decision

   Anything requiring approval from a body that meets monthly has a timeline you do not control. Save those for when you have credibility to spend.

4. 4

   Something you can say was finished

   With a date. "Done" is a word HR programmes use very rarely and it is worth more than any amount of progress reporting.

After two or three of these, the conversation changes. Requests start arriving rather than resistance, because people have seen that things get finished. That shift is the actual asset a transformation needs, and it cannot be created by a roadmap.

The steering committee problem

Large programmes acquire governance, and governance in a small business is frequently the thing that stops the programme rather than the thing that steers it.

A monthly steering committee means every decision waits an average of two weeks, and a decision deferred once waits six. In a function with two workstreams and a clear owner, that machinery costs more than it protects. It is worth being honest about which decisions genuinely need a committee — usually only those involving money or organisation structure — and which need one named person who can be told they were wrong afterwards.

The test is simple: if a decision could be reversed in a week at little cost, it does not belong in a committee. Most HR process decisions meet that test, and treating them as though they do not is a large part of why HR transformation and change management programmes move so much slower than the equivalent work done informally.

Who has to be convinced

HR change fails at the line manager layer more often than anywhere else. Leadership approves it, HR designs it, employees are told about it, and managers are expected to operate it — usually having been consulted last, if at all.

• Managers are the operating layer for every HR process. A performance process they do not believe in produces completed forms and no conversations, which is worse than nothing because it looks like compliance.
• They have a legitimate objection worth hearing: time. Every HR change adds to a manager's workload. A change that does not say what it replaces is asking for more with nothing given back.
• Consult a handful before designing, not after. Three managers in a room for an hour will identify the two reasons your design will not work in their context, and they will then defend it.
• Tell them before their teams. A manager who learns of a change from a direct report has been undermined, and they will remember it during the next one.

What to do with the programme you already announced

If the town hall has happened and things have stalled, the recovery is to reduce scope publicly rather than to quietly let it fade.

Say that eleven workstreams was too many, name the two that will be completed and by when, and say what is being deferred and why. This is a difficult conversation and it is considerably better than the alternative, which is a programme everyone knows has stopped and nobody has been told about. Silence is read as failure regardless, and it costs the credibility of the next attempt as well.

Then finish the two. Publicly, with a date, and with a clear statement that they are done rather than progressing. A transformation delivered as a sequence of finished small things arrives at the same destination as the roadmap did, roughly as quickly, and without the period in the middle where nobody believes it — which is why most workable HR transformation and change management engagements look far less dramatic than the announcement would suggest. Where the sequence keeps stalling on the same unresolved question about who is accountable for what, the missing prerequisite is usually structural, and organisation design is the thing that has to be finished first.