Thirty two thousand hours

The rollout was already live. The target was missing.

An 8 to 10 week sprint. One senior analyst and two juniors. A €19 billion European energy and refining major, across its project organisation.

The client had launched a Management Operating System across its project organisation. The tools were in place. Short interval controls, structured meeting cadences, escalation paths. What was not in place was a single number that anyone could be held to.

They were not short of a diagnosis. An internal survey had already told them where the time went: meetings, escalations, and email. Everybody had read it. Nobody disputed it.

What they did not have was a quantity. And without a quantity there is no target, so the rollout was running against an aspiration.

Three things had to be true at once for this to persist, and all three were.

No baseline. The operating system had no leading indicators. Project level KPIs only became visible one to three years after a project completed, which meant nothing anyone changed this quarter could be attributed to anything for a very long time.

No monetary value. The internal survey named the categories and stopped there. Meetings were a problem. How much of a problem, in euros, nobody had asked.

No reduction targets. The rollout was live across all three categories with no number to drive against in any of them.

That is the whole situation. It is not a sophisticated failure. A large, competent organisation knew exactly what its problem was and had never counted it.

The loss is distributed, so nobody owns it

It is worth being precise about why this particular measurement does not happen on its own, because the reasons are structural and they will be the same in your organisation.

No single person loses a day a week to badly run meetings. Several hundred people lose twenty minutes each. Twenty minutes is beneath the threshold at which anyone raises a complaint, and it is far beneath the threshold at which anyone starts a project. The aggregate is enormous and the individual experience is a mild irritation. Costs that behave this way do not surface, because there is nobody for whom surfacing them is worth the effort.

Each person’s own week looks fine. Only the sum looks broken, and nobody sees the sum.

The only witnesses are the people doing it. There is no system log for a meeting that ran forty minutes past its usefulness. There is no field in any ERP for the escalation that existed because two functions had not spoken. The evidence lives in people’s recollection, which means any measurement is self reported, which means the first response to any number you produce is that it is soft. That objection is correct, and we come back to it below, because how you answer it decides whether the work is used or filed.

The feedback loop is longer than anyone’s attention. One to three years to see a project KPI move is longer than most tenures in the role, and considerably longer than the interval at which an improvement programme is judged. So even a genuinely good intervention cannot prove itself in the window where proof would matter.

And there is a fourth reason that is more human. Counting the cost of meetings is an accusation. Somebody scheduled those meetings. Somebody escalated. To measure it you need an instrument that describes a system rather than blaming a person, and you need somebody outside the room to run it.

None of that is a knowledge gap. The client’s own people understood their operation far better than we did, and they still do. What they did not have was two weeks of somebody’s undivided attention and a method for converting a shared complaint into a defensible number. That is the trade, and it is worth naming plainly, because it is the only honest reason to bring anyone in from outside.

Three numbers, and one word doing a lot of work

Read the last line carefully, because the word in it is doing real work. It says potential. It is a modelled figure describing what recovering the mapped hours would be worth, not a saving anybody has banked. Nothing in this engagement measured euros arriving. The measurement was of time, and time was converted to value through a model whose assumptions are written down and can be argued with.

We keep saying this in an inconvenient way, because the alternative is how modelled numbers quietly become achieved ones, and it is not a small problem. It is the difference between a firm you can check and a firm you have to trust.

A chemical company, a different function, the same missing number

A €3 billion global chemical company wanted its plant supervisors spending more time on safety and frontline performance. Everyone agreed with the goal. Everyone also agreed the supervisors were buried in reporting, meetings and administration. And there the conversation had sat, for years, because “free up supervisor time” is a wish, not a target.

So it was counted. The supervisor’s week broke into seven task themes: safety, reporting, supervising and training, maintenance and production, meetings, project related work, and everything ad hoc.

The three priorities that came out of it were meeting practices, human resources, and unfinished projects. Meetings were reported as lacking a clear agenda, running off topic, and being attended by people who had not prepared. Training materials were not adapted to local plant language or machinery. And projects were routinely planned without the supervisor’s input on time and budget, then postponed, then redone the following year.

One detail is worth pulling out because it is the most actionable thing in the whole study and it is not a strategy. Work permits were named twice, independently, in two different parts of the analysis: once as a long and arduous task, once as a cumbersome and bureaucratic process. A permit workflow is a thing a company can change in a quarter without a transformation programme.

The reallocation analysis, where it was done in detail, offered exactly two destinations for a supervisor’s hour. Give it to a system, by automating monitoring and consolidating reports into one interface. Or give it to a different role, by upskilling the handling workforce so repairs do not need a supervisor to execute them. There is no third option. Time does not get freed. It gets moved, and somebody has to accept it.

A correction, and it matters more than the rest of this section. The headline figure attached to this case, in our own material and on our website, has read “80% supervisor time saving potential”. The underlying study does not say that. It says that 80% of supervisors saw some potential to save time in reporting and meeting related tasks. Those are different claims, and the second one had been restated as the first. 80% of people seeing some scope is not 80% of a week being recoverable. We are correcting it. If you have read that figure somewhere and made anything of it, that is on us.

Everyone was asked. Nobody was observed.

Both of these engagements measured the same way: they asked people. In the energy case, 200 people were asked about frequency and magnitude. In the chemical case, supervisors were asked to estimate what they would save if their specific task challenges were fixed, answering in bands rather than in exact hours.

Neither study observed anybody. There was no time and motion work, no calendar extract, no system telemetry. It is worth saying that in plain language rather than letting an hour count imply a stopwatch.

It is good enough to rank. If several hundred people independently rate meeting structure as a more severe problem than supplier communication, the ordering is real even if every individual estimate is wrong. Ranking survives measurement error in a way that magnitude does not.

It is good enough to set a target against. A target’s job is to be specific, shared, and revisited. A target built on a documented survey of 200 people is enormously better than no target, which is what the rollout had.

It is good enough to find the actionable detail. Nobody would have found the work permit workflow from a spreadsheet. It came out because 55 people were asked what made their week hard, and two independent parts of the analysis named the same process.

It is not good enough to promise a saving. Self reported estimates of hypothetical savings are optimistic, in a direction people cannot help. Anyone presenting €2.4 million from this method as money in the bank is misrepresenting the instrument, and if they do it once they will do it with your numbers too.

The count tells you where to work and gives you something to be measured against. Whether the recovered hours turn into output is a separate question, answered later, by re running the measurement.

Four things, named because leaving them out would be the same error as the 80%

The period the 32,000 hours covers is not stated in our own record of the engagement. It is the aggregate mapped by the study across the surveyed population. Read it as the size of the pool that was measured, not as an annual run rate, unless and until we can point you at the assumption.

Whether recovered hours become output. An hour taken out of a meeting is not automatically an hour of productive work. The monetary model assumes a conversion. That assumption is the single largest source of uncertainty in the €2.4 million and it is the first thing to argue about.

Neither study observed anyone. Everything above rests on what people reported about their own week. That is a legitimate instrument used carefully, and it is not the same as measurement.

Our own internal record is inconsistent on how many projects were analysed, giving two different figures on consecutive pages. We have not used a projects figure anywhere in this article for that reason, and we are reconciling it.

Where every figure above comes from

The energy case. A €19 billion European energy and refining major. A lost time survey across 200 employees capturing frequency and magnitude of meeting, email and escalation issues across projects and functions. Twelve interviews with the roles operating the Management Operating System. Three per category monetary value models. An Excel tracking tool delivered to the client to re run the measurement annually. Reported outcomes: 32,000 hours of avoidable lost time mapped, a 57% inefficiency rate benchmarked, €2.4 million of productive output potential unlocked. The client’s Director of Operations described the result as understanding exactly where, how much, and what fixing it was worth.

The chemical case. A €3 billion global chemical company. 55 supervisors and plant managers interviewed and surveyed across more than 60 sites in more than 10 countries, validated with HR, safety and operations leaders and consolidated through six stakeholder workshops. Seven task themes. Time saving potential collected as banded self estimates. The client’s Vice President of Operations described the result as clarity on where supervisor time was going, delivered in weeks rather than the months it would have taken internally.

Every figure above is either a mapped, benchmarked, identified or potential figure, and is described as such. Where our own materials have upgraded one of those into an achieved result, as with the 80%, we have said so.

Clients are described by revenue scale and sector, never named.