The Process You Killed Is Still in Someone's Inbox

Launch day is easy to love. There is a town hall, a new logo on a slide, a Teams message that says 'we're live', maybe a cake in the Manchester office. The new workflow gets a name, a training session and a champion. Everyone is told what to do next.

Then the old workflow keeps happening.

Not because anyone is rebellious. Because the old form is still linked from the intranet. Because the shared drive folder is still called 'current'. Because the onboarding deck still says step four is to email the spreadsheet to Sandra. Because Sandra still has the spreadsheet open.

UK teams are good at launching change and poor at ending it. We treat retirement as an afterthought, something that will happen when the new thing beds in. It rarely does. The old process becomes a ghost workflow: still moving through inboxes, still collecting data, still teaching new starters the wrong version of the job.

Launch culture has a blind spot

A launch plan usually covers comms, training, access, support and success measures. A retirement plan covers the bit nobody wants to own: what stops, when, and who turns it off. Without that, change becomes accumulation. The new process does not replace the old one; it sits on top, and people learn to hedge.

Hedging looks sensible at team level. If you are not sure which form is official, you fill in both. If the new approval route is slow, you email the old approver 'just in case'. If the old tracker is the only place with last year's history, you keep updating it. Multiply that by a few dozen people and you have a second shadow operation running alongside the one you just funded.

Ghost workflows are expensive

The cost is not always dramatic. It is the new starter who is taught two versions of the same process and spends their first month confused. It is the manager who checks two inboxes. It is the analyst who reconciles a spreadsheet that should have been retired in March. It is the customer who gets a request for information they already supplied through a different route.

Sometimes the cost is serious. Old forms can collect personal data that no one is looking after. Legacy shared drives can hold documents that should have been deleted under a retention schedule. A retired approval route can leave a gap in the audit trail. In the UK, where data protection and public sector accountability are not abstract concerns, 'we forgot to switch it off' is not a great look.

A retirement plan is four simple things

You do not need a transformation programme to kill a process. You need a short plan that someone owns.

First, archive rules. Decide what happens to the old records. Do not just delete them. Make the old location read-only, name it clearly, and set a retention period that fits your legal and operational needs. If the old tool holds personal data, check that access is restricted and that the retention clock is real.

Second, redirects. Every old link, form, bookmark and auto-reply should point to the replacement. The intranet page should say 'this process retired on 31 October 2026; use the new booking form'. The old mailbox should bounce with a helpful message. The shared drive should have a readme at the top, not a lonely folder called 'old stuff'.

Third, a named custodian. One person accountable for the retirement, not a committee. They chase the stragglers, check the redirects and decide when the old thing can be switched off. This is not a permanent job. It is a role for the life of the retirement.

Fourth, a clear end date. 'Soon' is not a date. Put it in the calendar, communicate it, and stick to it. If the new process is not ready, move the date openly rather than letting the old one drift on indefinitely.

You do not need a funeral, but you do need a full stop

Change involves loss, even when the old process was rubbish. People built the spreadsheet. They knew its quirks. A small acknowledgement helps: a note that thanks the people who kept the old thing running, a short explanation of why it is going, and a clear signpost to the new route. You do not need sandwiches and a eulogy. You do need to mark the ending so people know it is real.

That is especially true in UK organisations where politeness can keep a dead process on life support. Nobody wants to be the person who turns off the form. A named custodian and a public end date give permission to do it.

The 30-day drift check

Retirement does not end when the switch is flipped. Check for drift. Search for submissions to the old form. Look at traffic to the old mailbox. Ask the support desk what people are still asking. If the old process is still being used, find out why. Sometimes the answer is a missing redirect. Sometimes it is a gap in the new workflow that needs fixing. Both are useful.

Put the check in the plan: 30 days after the end date, confirm the old route is closed and the new one is working. Then close the retirement record. This is not bureaucracy; it is how you stop the ghost from coming back.

Make retirement part of every launch

The simplest fix is to change the launch template. No new workflow goes live without a retirement plan for what it replaces. Add two questions to the project brief: what does this replace, and when does that stop? If the answer is 'nothing' or 'not yet', say so explicitly. That is a decision, not an oversight.

Then keep a lightweight retirement register. A simple list is enough: old process, replacement, custodian, archive location, redirect status, end date. It can live on the existing project board. It does not need a dashboard, a steering group or a new tool. It needs to be looked at.

Change that only adds is not transformation. It is clutter with a communications plan. The kindest thing you can do for a new process is to retire the old one properly. Give it an end date, a custodian and a redirect. Then let the team get on with the work.