Culture change in a digital transformation, and where it stalls

Most digital transformations do not stall on the technology. They stall on the working habits the technology exposes: who is allowed to decide, what happens to someone who raises a problem late, and how work moves between teams that have never had to share it before.

That is what people mean when they call it a culture problem. The phrase is unhelpful, because culture sounds like something you address with a values poster. The parts that actually block a transformation are more specific, and most of them are visible in how the organisation already behaves.

What to look at when someone says culture

Four things determine whether new tools and processes get used as intended.

Decision rights

Who can approve a change, and how many people can stop one. Transformations usually add new decisions without saying who owns them, so they default to whoever owned the nearest old thing.

What gets rewarded

Not the stated values, but what appears in objectives, promotions and pay. If delivery volume is measured and rework is not, people will ship things that create rework.

What is safe to say

Whether a team can report a slipping date, a broken assumption or an unusable tool without it being treated as a performance issue. This determines how late you find out about the others.

How work is handed over

The points where something moves between teams are where a transformation either works or quietly reverts to the old process running alongside the new one.

Risk-taking, collaboration and continuous learning are the usual headings for this, and they are too abstract to act on. The four above can be checked against records that already exist.

Assessing the culture you have

You do not need a large survey programme to get a usable picture, and a survey on its own tends to produce the answers people think are wanted.

More useful sources:

• Decision records and approval chains from the last two or three significant changes, which show where decisions actually stopped rather than where the process says they should
• The points in the last year where a date slipped, and how long the slip was known about before it was reported
• Handover points between teams, and whether a shadow process runs alongside the official one
• Exit interviews and internal transfer requests, which tend to be more direct than an engagement survey
• Objectives and performance criteria for the roles the transformation depends on

Structured instruments have their place. The Competing Values Framework and the Denison Organisational Culture Survey both give you a vocabulary and a comparison point, which is worth something when you need to show a board that the assessment was not improvised. Treat their output as a description of how people answer questions about their organisation, which is not the same as how the organisation behaves under pressure.

Interviews are worth doing if you interview across levels rather than at the top. The people who know where a process breaks are usually the ones operating it.

Where transformations stall

A few failure patterns recur often enough to look for deliberately.

The new process becomes optional. The old system stays available "for now", and it keeps working, so it keeps being used. This is a decision about switching things off, and it is normally avoided because someone senior is worried about the exception cases.

Middle managers are given the change to deliver but not the authority to change anything. They are measured on the old numbers throughout, so they protect the old numbers.

The transformation is communicated as a completed decision. People who can see a problem with it conclude there is no point raising it, and the problem surfaces later at higher cost. If you want the objections, you have to ask for them before the decision is announced, and be visibly willing to act on some.

Training is delivered too early, before the tool exists in usable form, and is not repeated when it does. People form their working habits in the gap.

What leadership alignment means in practice

Leaders do not need to agree that the transformation is a good idea. They need to give the same answer when their team asks what happens to the old way of working, what happens to the numbers they are currently measured on, and who decides the exceptions.

Where leaders give different answers to those questions, teams follow the answer they prefer, and the programme develops two versions of itself. This is the most common form of what gets described as resistance, and it is not resistance. It is people acting reasonably on inconsistent instructions.

The observable part of leadership behaviour matters more than the messaging. If a leader announces that raising problems early is expected, and then reacts badly the first time someone does it in public, the announcement is dead. Nothing recovers it except a visible instance of the opposite.

Being clear about who holds which role in the change avoids much of this. Stakeholder engagement roles, and who holds each one sets out the distinctions, and what a stakeholder engagement framework is covers how to structure the wider engagement.

Interventions that hold, and ones that do not

Training programmes, workshops, coaching and incentives all appear on the standard list. They work when they change something structural and fail when they are the only thing on offer.

Interventions worth the money tend to have a structural component:

• Changing the objectives and performance criteria for the roles the transformation depends on, before asking those roles to work differently
• Setting a date to switch off the old system, with a named owner for the exception cases
• Giving one team the authority to decide the questions that keep escalating, and publishing what they decide
• Putting people who will operate the new process into its design, early enough that their objections can still change it

Incentives deserve particular care. Paying people to demonstrate a behaviour reliably produces the appearance of that behaviour. If you reward collaboration without measuring an outcome that collaboration would improve, you will get meetings.

Signals that it is working

Culture measures are proxies, and it is better to say so than to present an index as a fact. The ones that carry information are mostly counts of behaviour rather than statements of attitude.

Worth tracking:

• How long a known problem takes to reach someone who can act on it, compared with a year ago
• Whether the old system is still being used, and by whom
• How many decisions escalate above the level that was supposed to own them
• Whether the exception process is being used at the volume expected, or well above it
• Turnover in the roles the transformation depends on most

Attitude surveys are still useful for direction over time, provided you accept they measure willingness to answer. Run one twice with the same questions before drawing any conclusion from a movement.

Setting these up properly is the same problem as any other measurement exercise, and how to set KPIs and marketing objectives covers the mechanics of picking measures that will not mislead you.

What communication contributes, and what it does not

Communication cannot make an unpopular decision popular, and it cannot substitute for authority that was never granted. What it can do is remove the ambiguity that lets two versions of the programme coexist, and give people a route to report problems that does not require courage.

That is a narrower claim than the usual one, and it is worth being honest about the boundary. Internal communications and employee engagement sets out what comms can and cannot carry. If you need the structure for the transformation programme specifically, a transformation communication plan you can fill in is the practical version.

If you are starting an assessment, the decision records are the place to begin. They are already written, nobody has to be asked how they feel, and they show you where decisions actually stop.