Notes on operational excellence.

Who Checks at 2 am on a Sunday? How a Compliance Scorecard Ensures Adherence to Standard Operating Procedures

Published 13 October 2026

In my last two blogs, I wrote about the importance of standard operating procedures, and of making sure they are trained, with refresher training and competency assessments carried out on an ongoing basis. But refresher training is all well and good, and when competency checks are done, usually during a day shift session, operators tend to learn to adhere to the procedures. That is when people know they are being watched, and it leaves a gap I want to look at this week.

The 2 am problem

So what happens when procedures are not adhered to, at 2 am on a Sunday morning, when there is no management present, no Gemba walks, no CCTV cameras, and no one to enforce adherence?

I experienced this first hand. I took a step back and did some thinking about the best way to enforce adherence without damaging morale.

The compliance scorecard

The answer was a compliance scorecard. After discussion with leadership, I came up with a daily check of compliance to the standard operating procedures. It also covered:

• Data and information capture. This mattered to the business, because a lack of data capture meant management did not get timeous information on which KPIs were not being achieved.
• The status of problem solving for shift team KPIs.
• Quality at source testing by operators, which I will talk about in more detail in a future blog.

The checks were done by operational excellence staff every day, and each shift operations team was scored on its compliance for the preceding day. The scores were then circulated daily across all shift teams.

From enforcement to pride

The scoring quickly escalated into healthy competition between the operational teams, as no team wanted to have the lowest compliance score at the end of the month, quarter or year. It became a sense of pride. Management took it further by building the compliance score into team goals and KPIs, so that shift team performance was evaluated with compliance as a leading indicator.

From a low base of 29%, the overall compliance score improved to the point where each shift team consistently scored over 90%. While the direct impact of the compliance score on KPIs was not tracked, some KPIs showed a long term improvement which could be partly attributable to improved compliance. For example, downtime decreased from a daily average of 52 minutes to 11 minutes!

Once teams were consistently scoring over 90%, leadership reduced the frequency of the checks to three times a week, and then twice a week, on random days, so that teams could not let up.

Why it worked

Looking back, three design choices stand out:

• The score was visible and comparable. Teams could see where they stood against each other, which turned compliance from something done to them into something they owned. It is the same mechanism that created healthy competition when shift metrics were made visible in the short interval control introduction.
• The checks were independent. They were done by operational excellence staff, which kept the scoring objective.
• The checks became unpredictable. Once they moved to random days, a team could never know which day would be checked. Behavioural psychology has long found that unpredictable checking sustains behaviour better than a predictable schedule, and a team that cannot predict the check has to comply every day.

There is a risk worth naming. Once a measure becomes a target, it tends to stop being a good measure, as people learn to improve the score rather than the underlying behaviour. This is often called Goodhart's law. Independent scoring, random check days, and a scorecard that goes beyond a simple check list to include data capture and problem solving are what help keep it honest.

A leading indicator the workforce owns

Last week I described breakdowns and lost sales as lagging indicators, and adherence and competency checks as leading indicators. The compliance score is a good example of a workforce leading indicator of KPI achievement. It tells management, every day, which shift teams are following the procedures that deliver the KPIs, long before the KPI results themselves show a problem.

This follows the thread running through this series. The strike showed that standard procedures only work when they are followed. Last week's story showed that they only keep being followed if someone keeps checking. This week's shows how to keep checking when nobody can be there. Supervision only works while the supervisor is present. A score the team cares about is present at 2 am on a Sunday.

Sources

Strathern, M. (1997), "'Improving ratings': audit in the British University system," European Review, 5(3), 305–321.

Don't Let Up: What Happens When Refresher Training and Competency Checks on Standard Procedures Stop

Published 6 October 2026

In my last blog, I narrated the story of a strike that improved efficiency, and of the standard procedures, retraining and competency assessments that went on to deliver a 16% efficiency gain over five years. Getting procedures trained and assessed is a real achievement. But it is not a once off; it must happen continuously. This week I want to look at what happens when the training and the competency checks quietly stop, because in my experience that is where hard won gains start to disappear.

Why procedures drift

Without regular checks, operators tend to find shortcuts. Each one usually makes sense at the time, for the operator, and causes no immediate problem. The trouble is that the consequences arrive much later, as quality defects, unplanned downtime and breakdowns, long after anyone can connect them to the shortcut that caused them.

Sociologist Diane Vaughan gave this pattern a name, normalisation of deviance, in her study of the Challenger space shuttle disaster: small departures from the documented procedure that cause no harm become the new normal, until one day they do. Refresher training and regular competency checks are how an operation catches that drift while it is still small.

What happens when the training stops

Here is another real life example. In this case, refresher training and competency checks on procedures were allowed to lapse. The consequences took time to show, but when they did, they were costly. Machinery breakdowns during a peak trading period left product missing from retailer shelves, and the company lost market share at exactly the moment it could least afford to. When the root cause was traced, the breakdowns led back to non adherence to procedures.

Looking back, three connected causes stood out:

• Refresher training had stopped. Without it, adherence to procedures slowly declined, and with the competency checks gone, nobody was measuring it.
• There was no structured onboarding. New operators were taken on without formal training in the procedures for the areas they were responsible for. They learnt by watching the operators around them, who were not following the procedures anyway, so they learnt the wrong way of doing things. This is tribal knowledge at work: a shortcut that was never written down, passed from one operator to the next.
• The procedures were outdated. They had not been updated when processes changed and new equipment came on board, so the documents no longer matched the equipment and processes they were meant to describe.

The second cause is critical. A new operator will copy what they see. Without a structured onboarding plan, whatever they see first becomes their standard, and the drift spreads to everyone who joins.

The fix

The recovery was methodical, a very practical application of the Plan-Do-Check-Act (PDCA) cycle:

• Every procedure was checked to confirm it was still relevant.
• Procedures were updated where processes or equipment had changed.
• All operators and first line supervisors were retrained on the updated procedures.
• A defined onboarding process, with defined training plans, was put in place for all new operators.
• Procedures were then checked at defined frequencies (or when a process or equipment change was done) to determine if they required updating.
• If updates were needed, retraining and competency assessments were redone after updating.

Including the first line supervisors in the retraining mattered. They are the people who reinforce the procedures every shift, so if they are not trained and aligned, the operators' training fades again.

Questions worth asking in your own business

• When was each procedure last reviewed, and who owns keeping it current?
• When did each operator last pass a competency check on the procedures in their area?
• What does a new operator's first month look like, and who are they learning from?
• When equipment or a process changes, what triggers an update to the procedure?
• Are first line supervisors trained to the same standard as the operators they supervise?

The thread through all of this

This ties together several threads from earlier blogs. Kotter warned about declaring victory too soon, and about neglecting to anchor changes firmly in the culture. Letting refresher training lapse is a quiet way of doing both. A Gemba walk is how you see drift early, by standing at the machine and watching whether the procedure is actually being followed. And short interval control shortens the time to detect and the time to react.

It also comes down to what you measure. Breakdowns and lost sales are lagging indicators: they tell you what went wrong months ago. Competency check results and adherence are leading indicators: they tell you what is about to go wrong. Once the competency checks stopped, the only signals left were the lagging ones.

The lesson from the strike was that standard procedures only work when they are followed. The lesson from this story is that they only keep being followed if someone keeps checking.

Sources

Vaughan, D. (1996), The Challenger Launch Decision: Risky Technology, Culture, and Deviance at NASA, University of Chicago Press.

Kotter, J.P. (1995), "Leading Change: Why Transformation Efforts Fail," Harvard Business Review.

The Strike That Improved Efficiency: What It Taught Me About Standard Work

Published 29 September 2026

In this blog I tell a counterintuitive tale of a strike that actually resulted in improved production efficiencies, which goes a long way to explaining why standardised work is a crucial baseline for success in any business. Here is a true story from early in my career that makes the point better than any theory could.

The strike that improved efficiency

At a multinational I once worked for, in the early 1990s, the labour force went on strike for around six weeks. Head office sent leadership and management into every manufacturing facility to keep production running. While some of them had operated plant equipment, it was well in their past. Others had never operated the plant equipment before.

When the strike ended, someone at head office ran the numbers to assess the damage to the business. What they found was that production efficiency had actually improved while leadership and management, not the trained operators, were running the machines.

What the investigation found

A detailed root cause investigation followed, and the results were not what anyone expected. Leadership and management, with no prior experience on the equipment, had reached for the documented standard operating procedures as their first and only move, and followed them religiously. The regular operators (who were on strike), by contrast, mostly ignored the documented procedures and ran the machines the way they personally preferred.

This was a 24/7, shift based operation, and different operators running the same equipment across different shifts did things differently. Machine settings, sequencing, and pace all varied shift to shift, sometimes hour to hour. That variability showed up as inconsistent product quality, unplanned downtime, and efficiency that depended more on who was on shift than on the equipment itself.

Why this is a compliance problem, not a documentation problem

The procedures existed the whole time. The plant was not short of documentation before the strike, it was short of shared compliance. A document tells you what should happen. Compliance is what automatically happens, repeated the same way, by whoever is on shift, without anyone having to deviate from procedures.

This is close to a principle Toyota built its production system (TPS) on: you cannot improve a process you cannot repeat. If every operator runs a machine differently, there is no stable baseline to compare against, so there is no way to know whether a change makes things better or worse. Standard work, on this view, is not about control for its own sake. It is the precondition for being able to improve anything at all. Ironically, in this instance, the operators who prized their own experience and personal technique were the ones producing the least improvable, least consistent result. Management, who had no experience at all, were producing the most consistent one, purely because they had nothing to fall back on except the procedure as written.

The fix, and why it held

The response was not a directive from leadership instructing operators to comply. It was a series of focus group sessions where operators from different shifts worked out, together, the best way to run each machine. Only once they had agreed on best practices in equipment operation, did the business document these best practices as standard work procedures, aligned to ISO 9001, retrain everyone on them, and begin random competency assessments to check if the standard work procedures were actually being followed, not just filed.

That sequence matters. The standard was not handed down as a directive, it was built by the people who knew their equipment and the best way of operating it. This approach and sequencing built a sense of ownership with operators, they owned and lived the standard work procedures. And the competency checks were not a compliance exercise, they were what turned a written document into a standard, the same reinforcement loop that separates a Gemba walk that changes something from one that doesn't.

What it delivered

Over the following five years, production efficiency increased by 16%. That single number carried a wider set of business results: a better return on investment on existing machinery, shorter payback periods, higher and more consistent product quality, and enough freed up capacity that the business could grow production without new capital expenditure. Employee morale improved too, which should not be a surprise as they built the standard in the first place.

The thread through all of this

The lesson isn't that documentation doesn't matter. It's that a document only becomes value adding once it's built with the people who use it, and checked often enough that it stays real. The same thread runs through every post in this series: the tool was never the hard part. What people actually do with it, and whether anyone keeps checking that they're still doing it, is what determines whether an operation performs the same way regardless of who's on shift, or who's just walked out the door.

From Gemba Walks to Short Interval Control: What an Excel Sheet Taught Me About Empowerment

Published 22 September 2026

Last week I wrote about Gemba walks, and the research showing that presence alone changes nothing, it's the follow through that matters. This week I want to demonstrate that idea with a story from early in my operational excellence career, in the middle 2000s. This story is one of my major career highlights, as the idea did not only move KPIs in the right direction - it empowered the workforce to use a tool developed from the Gemba walk, in their personal lives outside of the work environment.

The problem the Gemba walks kept surfacing

Regular Gemba walks with the workforce kept turning up the same disconnect. Operators knew, in the moment, whether a shift (this was a 24/7 operation) was on track or falling behind. That knowledge lived in their heads, scribbled on a piece of paper or notebook, at or near the shift control room. Leadership only saw shift performance the next day, once someone had collected the paper, typed it up, and circulated a report. By the time the report indicated there was a problem to leadership and management, the shift that experienced the problem was long gone (and probably at home asleep), and so was any chance to identify the root cause to prevent reoccurrence. And most of the time, the problem did reoccur, if not the same day, then within the following month...

Building something simple: an Excel based short interval control system

The solution wasn't sophisticated or complex. We built a short interval control system in Excel, with no special software or tools, nothing exotic. The design decision that mattered most wasn't in the spreadsheet at all: it was who captured the data. Instead of a team leader, supervisor or shift manager transcribing key performance metrics after the fact, operators recorded their own shift performance themselves, at designated frequencies, and were held personally accountable for it, goal documents were set up so that operators were measured on these. Using simple Excel tools such as conditional formatting immediately flagged if a performance metric was out of specification: if a metric was flagged in red on capturing, then there was an issue somewhere, which had to be investigated and corrective action taken immediately on. This prevented reoccurrences, and the term "quicker time to detect and time to react" became a mantra for the workforce.

Why this shouldn't have worked

On paper, using a simple Excel based system was a strange choice. However, the culture and capabilities of the environment are critical to planning any change intervention, a one size fits all approach will inevitably result in failure ("Culture eats strategy for breakfast"). On the culture and capabilities background, most of the operators did not have tertiary education. Many came from the local underprivileged community with only a secondary education dating back to the 1980s, and had no prior computer and technological skills at all (this was back in the 2000s, when smartphones had yet to make their presence felt). Handing that workforce an Excel based system, and asking them to own it, was a real risk. It could easily have become another initiative that leadership and management were proud of and the workforce quietly resented.

What it actually delivered

It didn't fail. It did something better than hit its original target of "quicker time to detect and time to react", it changed people, in two ways no one fully anticipated when we started.

• Empowerment. Learning to use the system was, in effect, on the job computer training. Operators who had never switched on a computer learned to navigate a spreadsheet, enter data accurately, and check their own numbers. Those are not workforce skills, they're life skills, and several operators thanked leadership and management for providing them with a (free) opportunity to learn new skills so that they could keep up with changing technology. Operators proudly told stories of how they bought personal computers for their home use (having a company computer assistance scheme also helped) which were now being used to do budgeting, and to help their children format and prepare school homework (no AI then!). The system built capability that walked out the gate at the end of every shift and kept paying dividends long after the shift ended.
• Ownership. Because operators were personally inputting and being accountable for their own shift's performance metrics, the data stopped being something management collected about them and became something they owned. They could see, directly and immediately, how their shift's performance rolled into the site's overall result. That line of sight, from my effort to this number to the plant's performance, is a different kind of motivation to being handed a report with your name on it after the fact. It also resulted in healthy competition among shift teams, as all shift teams' metrics were visible, no team wanted to be seen to be doing the worst of all. This elevated overall site performance to a significant extent.

The site at which this was done is also the site behind the MBA research I've referenced through this series of blogs. Across that change programme, six of the eight primary success indicators we tracked showed statistically significant improvement. I can't claim the short interval control system alone drove that result, change is rarely that compartmentalised, but it was one of the contributory factors. It's the system I remember best (and am very proud of) because of what it did for the people running it, not just the impact it had on site performance.

On the subject of change management, a structured change management system (using the Kotter 8 step principle) was used for the short interval control system, and it proved its worth. Initial take up was slow, with lots of coaching from leadership and management in the first two months during the Gemba walks. By the third month, critical mass had been reached, and the system was fully accepted and operational by the workforce by the fourth month. Paid time off also dropped as operators now did their bill payments online, instead of physically going to a bank or store for mundane tasks.

This system also proved to be sustainable: even though I moved on from this site in the middle 2010s, I am aware that this Excel based system was in use until at least 2021/22. Speaking to a former colleague at a conference three months ago, I was told that with the advent of Big Data, AI and online monitoring, the system was finally retired in 2022/23.

The thread through all of my blogs

This is the same thread running through every blog of mine to date: the tool, whether it's a change model, a Gemba walk, or a spreadsheet, was never the hard part. What actually moves a business, and sometimes the people in it, is what leadership does with what the tool shows them, and how much of the process it's willing to hand to the people closest to the work.

Why Some Gemba Walks Don't Change Anything: What the Data Says About the Ones That Do

Published 15 September 2026

Last week I wrote that leadership (and management) who live a change, rather than just sponsor it, have a five point list of surprising habits. The first habit was showing up at workstations regularly enough that a visit isn't an event, and asking probing questions rather than only inspecting. This habit has a name in operational excellence: the "Gemba walk". It's worth a closer look, as the evidence on it's impact, is interesting, and more useful, than the age old advice to "get out of your chair/meeting room and walk the floor."

What a Gemba walk actually is

Gemba is Japanese for "the real place." In a manufacturing environment, that's the shop floor: the actual workstation, the actual machine, the actual point where a problem is happening, not the meeting room where it's being discussed. A Gemba walk, a routine that grew out of the Toyota Production System, is a structured visit to that "real place" so leadership and management can observe the work being done, first hand. They can (and should) ask the people doing it what's actually making their work difficult, instead of relying on a second or third hand status report or a dashboard that's already had the inconvenient detail smoothed out of it.

This sounds almost too straightforward to be a leadership and management discipline. That's exactly why it's not always executed successfully.

The uncomfortable finding: presence alone doesn't help

The most rigorous study on Gemba walks I've seen isn't a Lean case study at all. It's a randomised field study of Management By Walking Around, a productivity tool which is very similar to the Gemba walk. The study was run across U.S. hospitals by researchers from Harvard Business School and the Harvard School of Public Health. Twenty hospitals were randomly assigned to run an 18 month programme in which senior managers walked the frontline, asked staff to identify problems, and worked with them to resolve issues.

That finding surprised the researchers as much as it probably surprises you (and me, at first!!). It would have been easy to move past it and only quote the studies that make Gemba walks look good. But the more useful finding is buried in the detail of why it failed on average, because it identifies what separates a walk that works from one that doesn't.

What actually made the difference

The treatment hospitals identified an average of 17.3 problems per work area and resolved 9.1 of them, just over 50%. When the researchers investigated which work areas improved and which didn't, two things came to light:

• Work areas that solved a higher proportion of easy, quick to fix problems experienced better results. In the study's terms, each one point increase in their problem resolution measure was associated with a 26% improvement in staff perceptions of performance. Chasing the single biggest, hardest problem first, and leaving the small stuff, did not work nearly as well.
• Work areas where a senior manager took personal responsibility for making sure a problem got resolved saw a 21% increase in the performance measure. Where the walk ended with a note taken and no named owner, nothing changed.

The researchers' conclusion is worth repeating almost word for word: it's the actions taken on identified problems on the walk, not the presence of leadership or management, that improves how workforce experience their work. Walking the floor, noting problems, not doing anything and then going back to your desk is not a Gemba walk. It's a photo opportunity.

What this looks like in practice

This lines up exactly with what I see in businesses, and it talks directly into the two change management steps I highlighted last week as the ones leadership most often gets wrong: removing obstacles and recognising wins. A structured, action driven, Gemba walk is the mechanism that embeds and sustains these change management steps. A well executed Gemba walk will always focus on the following:

• It goes with genuine questions, not a checklist to tick off. "What's slowing you down right now?" beats a compliance walk every time.
• Small, easy fixes get actioned fast, even ahead of bigger ones. Visible momentum matters more early on than chasing the single highest value problem.
• There must be an action focus - every problem raised gets a responsible person and a target date, ideally someone senior enough to actually remove the obstacle, not just note it.
• The loop closes. The person who raised the problem gets told what happened, even if the answer is that it couldn't be fixed and why.
• It never turns into an audit. The moment a Gemba walk starts to feel like an inspection, people stop telling you the truth, and you've lost the only thing the walk was for.

None of this requires new software, capital expenditure or a bigger budget. It requires treating the walk as the start of a small, trackable commitment, not the end of a leadership box-ticking exercise. That's the same thread running through everything I've written so far: the tool or the technique was never the hard part. Following through on what it shows you is.

Sources

Tucker, A.L. and Singer, S.J. (2015), "The Effectiveness of Management-By-Walking-Around: A Randomized Field Study," Production and Operations Management, Vol. 24, No. 3.

Leadership by Example - If Leadership Doesn't Walk the Talk and Live the Change, How Can They Expect the Workforce to do so?

Published 8 September 2026

A large proportion of articles and blogs on lean and operational excellence emphasizes tools and techniques, such as value stream maps, DMAIC, statistical process control and short interval control. But in more than twenty years of planning and implementing operational excellence programmes in the workforce and in the boardroom, I've seen more programmes fail from something much simpler than a lack of tools and techniques: executive leadership said one thing and did another, and the workforce noticed.

I touched on this in my first blog, where the research behind my MBA dissertation showed a clear pattern in an operational excellence change programme, measured against Kotter's eight step change management model. The stages that worked well were the visible stages: building urgency, setting the vision, communicating it. The two stages that consistently let teams down were due to uncommitted leadership: not removing the obstacles standing in the way of new ways of working, and not visibly recognising the wins along the way. Workforce and management surveys said the same thing: the case for change was clear, but executive leadership did not provide support for the clearing of the obstacles for the programme to succeed; and did not ensure any recognition and celebration of small wins.

A workforce will attend all presentations on operational excellence programmes when the usual initial startup hype requires them to do so. But how will they believe in the programmes, and actually embrace the changes required for the programmes to succeed?.

Leadership do not give their workforces credit; the workforces are much more astute and observant than what leadership thinks. Workforce behaviour mirrors leadership behaviour – if leadership does not embrace and live the changes required by the programme (colloquially – “walk the talk”), this is soon observed by the workforce, who then start to question the motivation of leadership for the programme. Ultimately, this leads to lack of confidence and trust in leadership from the workforce.

This “lip service” by leadership in removing obstacles is not a single, isolated event. It's a manager who signs off on a new standard and then quietly reverts to the old shortcut under deadline pressure by leadership to meet productivity targets quickly. It's a shift supervisor who was never trained on the new visual management board (due to leadership declining training resource requirements and/or demanding improbable timelines), and simply stops using it. It's a plant manager who talks about continuous improvement at the town hall and is never once seen on the floor asking a team what's slowing them down, and what needs to be done. Every one of those moments is small. The workforce adds them up faster than leadership ever notices..

This isn't just a workforce impression. It shows up in research on why improvement initiatives fail.

• McKinsey's cited research pegs the failure rate of organisational change programmes at around 70%, attributing the high failure rate mainly to employee resistance and a lack of visible management support, not to the technical content of the change.
• On the positive side, McKinsey also found that organisations where leadership clearly defines roles and consistently communicates progress are as much as eight times more likely to see a change programme succeed.

Every successful operational excellence programme I’ve worked with has had executive leadership who recognized that their buy in and living the change process required, was vital to the success of the programme (and to the success of the business) . Conversely, I’ve worked with executives who expect that the success of an operational excellence program is solely dependent on their inhouse operational excellence specialist (or external consultant), and does not require their intervention, or more crucially, changes in their ways of working. These were the operational excellence programmes which fell flat and did not achieve any sustainable improvement for the business.

Very few leaders intentionally undermine an operational excellence programme they've just funded. It's about whether commitment is visible in behaviour, not just stated in a kickoff presentation. In practice, that comes down to a short list of unglamorous habits:

• Showing up in the workforce regularly enough that a visit isn't an event, and asking questions rather than only inspecting.
• Removing obstacles the workforce have flagged, quickly enough that the workforce notices the connection between raising it and it being fixed.
• Holding managers accountable for reverting to the old ways of work, exactly as consistently as holding workforce teams accountable.
• Recognising wins publicly, specifically and timeously, not folding it into a generic end of quarter results presentation.
• Sitting through the boring parts, the data review, the standard work audit, not just the launch event and the results slide.

None of this requires a bigger budget or a more sophisticated methodology. It requires leadership treating the change as their own discipline to model, not an initiative to sponsor and delegate. The realization by leadership of their crucial role, more than any tool in the operational excellence toolbox, is what separates the programmes that make a visible difference to the long term sustainability of the business, from the ones that quietly disappear from the next town hall agenda.

Sources

McKinsey & Company, "Changing change management," and "How to double the odds that your change program will succeed."

Why Change Programmes Are Not Sustainable: What the Data Says About the Ones That Do

Published 1 September 2026

Most businesses can point to a change or improvement programme that delivered results on launch day and quietly faded within eighteen months. The uncomfortable truth, backed by structured research into one such programme, is that even well resourced initiatives often only partially succeed when measured against a full set of performance indicators, not just the ones chosen as indicators of success.

The research, conducted by Collin for an MBA dissertation, tracked two linked change interventions against Kotter's eight step change model, then checked the results against real workforce KPIs rather than only using programme survey results alone. The pattern that emerged was consistent: of the workforce KPIs chosen up front as measures of success, roughly three in four showed statistically significant improvement. Of the wider set of KPIs that were tracked but never singled out as "indicators of improvement", fewer than four in ten did. Attention, as it turns out, is focused on "indicators of success"; everything else is not regarded as success indicators.

Mapped stage by stage, the change process told the same story. Building urgency, assembling a guiding coalition, setting the vision and communicating it landed well, and so did the later discipline of not letting up and anchoring new behaviour into how the business operates. The two stages that consistently let teams down sat in the middle of the model: removing the obstacles standing in the way of new ways of working, and visibly recognising the short term wins along the way. Workforce teams and their managers said the same thing in their individual surveys: the plan was clear and the case for change was clear, but not enough effort was focused in identifying and removing barriers to implementation and execution; and the wins weren't celebrated.

The practical implication for any change management process: measure broadly, not just indicators of success, and treat "removing barriers" and "recognising wins" as disciplines built into the change plan from day one, not steps assumed to look after themselves once the technical work is done.