Let me tell you about the most expensive meeting I’ve ever witnessed. It lasted maybe fifteen minutes, and it happened well over a decade ago, in a BPO I used to work at – long before tools like KYP existed. My team back then was called Delivery Excellence, and we had just rolled out a simple work-logging tool. No sensors, no AI. Each employee had their processes listed in it, and they clicked whichever one they were working on at that moment. And when they weren’t working on anything – a break, or simply no work left in the queue – they clicked stop. The honor system, digitized.
And there was this one employee whose data looked… provocative. Every morning: process, process, process. Lunch: exactly by the book. And then, almost every single afternoon: stop. One click. Until the end of the day.
His manager saw the dashboard and did what felt natural. Called him in for a difficult conversation. Voices were raised. Expectations were… clarified.
And you know what? It worked. The very next week, the improvement was spectacular. His afternoons filled up with process clicks. Stop button? Never heard of it.
Except then we looked at his output.
Identical. Same volume as before. Because here’s the thing – he was one of the most efficient people on that team. He simply finished everything he was assigned by lunchtime. And then he clicked the truth.
That man used to be the most honest row in our database. After one fifteen-minute meeting, he started lying like everyone else – and the company didn’t gain a single minute of work. What it lost was the only thing it actually had: a true signal saying “this person has spare capacity – he’s efficient, he’s underloaded, give him something bigger.” Oh, and one more thing: every fake afternoon click now landed on some process, quietly inflating its measured time. The lie didn’t just hide his free capacity. It poisoned the process numbers too.
Now – if you did the homework from our episode on what process intelligence actually collects, you’re holding five beautiful sentences about your sensor’s scope. And I know the instinct for what comes next: “Let’s not tie our hands with promises. We’ll decide what to do with the data once we see it.” That is exactly what that manager did. He decided once he saw it.
Today: the three rules of use you write down before the first dashboard ever renders – so that your best people never learn to click fiction.
Same data, trust or fear: why the rules of use matter
Quick recap for anyone who missed our episode on what process intelligence actually collects – it’s linked below, and the one-line version is: fear produces round numbers. Scared people produce fake data, so “what exactly are you collecting?” has to be answered out loud, in five decisions, before anything gets installed. That was the collection side. But at the end of that episode, one question was left hanging in the air – the one your works council asks after you’ve nailed all five decisions: “Fine. But what are you going to DO with it?”
Today we pay that debt. And to do it, I want to bring back a comparison from our conversation with Adam Bujak – one I haven’t been able to shake since.
Think about your doctor. They put a cuff on your arm and measure your blood pressure, and you don’t ask a single question. You don’t request a data retention policy. You don’t bring a lawyer. You pay them to take that measurement. You even say thank you afterwards.
Now imagine your insurance company calls and says: “Great news! We’d like to measure your blood pressure too. Same cuff, same number. Weekly. You know… for your own good.” Suddenly you have many questions.
Same cuff. Same measurement. Same number. So what changed?
Not the collection. The use. With your doctor, the rules of use are ancient, obvious, and one-sidedly in your favor: the reading exists to help you, it stays confidential, and it will never be quoted back at you as an accusation. With the insurance company, you can’t be sure the same number won’t be used to price you, flag you, or exclude you. The trust was never about the cuff. It was about what happens after the reading.
Your process intelligence sensor is that cuff. And whether your organization is the doctor or the insurance company is not decided by what you collect – we’ve covered that, in five decisions. It’s decided by what you have promised to do, and never to do, with the reading.
Here’s the mechanism that makes this urgent, and it’s the same one from the cold open. In the absence of written rules, your employees do not assume neutral use. They assume the worst case – every single one of them, from day one. Nature abhors a vacuum, and fear is excellent at filling it. Undefined use equals worst-imaginable use, in every head on the floor. And we established in that episode, in some detail, what fear does to your data.
And I can hear the objection already: “Andrzej, relax – we’ll just train the managers.” We did. In that BPO, my team spent real effort preparing managers for working with this data – how to read it, how not to weaponize it. Some of them got it. Sometimes it went well. And it took exactly one who didn’t, in fifteen minutes flat, to burn the honesty of one of the best people on the floor.
That’s the lesson I took from it: training is a suggestion. A written rule is a commitment. A rule that lives in a training deck dies in the first tense meeting – a rule that’s written down, signed, and public survives it.
So let’s write them down. Three rules. Starting with the sentence the homework told you to save for later – the sixth one.
Rule 1: The Never List – what process data is never used for
Let’s start with the sentence you saved. That homework ended with five sentences about what you gather – and a sixth one itching to be written: “this data will never be used for…”
Most companies never write that sentence. Not out of malice – out of caution. Leaving it unwritten feels like leaving your options open. But we’ve just seen what an open option costs: in that BPO, one manager exercised an open option, once, and the most honest data in the building turned to fiction. So today we close the options. On purpose. In writing.
This is rule number one – the Never List: a short, public, management-signed list of things this data will never be used for.
What goes on it? Four clauses that, in our experience, do most of the heavy lifting.
Clause one: no decision about an individual is ever made on this data alone. Ever. The data can start a conversation. It cannot end one. This is the cold-open manager’s mistake, written down and banned – he treated a dashboard as a verdict, when the dashboard was, at best, a question mark.
Clause two: this data supplements performance management – it never replaces it, and it is never automatically wired into it. The moment someone’s rating is calculated by a script reading the sensor, you’re not measuring the process anymore. You’re paying people to perform for the sensor. And by now you can guess exactly what they’ll produce.
Clause three: no attendance policing. When Adam Bujak from KYP was on the show, he said something I’ve been quoting ever since: the times when people were monitored are over – checking whether someone logs in at 9:00 sharp was important for some control freaks, but the point is the how. If your main use case for process intelligence is catching late arrivals, save the money: buy a punch clock and a worse company culture. It’s cheaper.
Clause four: no public shaming. A finding about an individual is never shown to a room. Not “as an example.” Not “anonymized, but everyone knows who it is.” A finding about a person travels exactly one path: a private conversation that starts with a question. More on that question in a minute.
Now – here’s the part of this chapter that matters more than any clause. And I can hear it coming, because I’ve heard it in real meetings: “Andrzej, this is all very noble, but we can’t put ‘this will never influence workforce decisions’ on paper. What if the data shows a team is genuinely overstaffed?”
Good. Then. Don’t. Write. It. I’m serious – this is the most important thing I’ll say today: the Never List is not a wish list of everything your employees would love to hear. Every clause is a promise you will have to keep in your worst quarter, not your best one. If restructuring based on aggregate data is a real possibility in your company, then a “never affects headcount” clause is not a promise. It’s a lie with a signature under it.
Here’s a rule I keep repeating to my own people, and it applies to every leader listening: the day your employees catch you selling bullshit is the last day they buy anything from you. Anything. Not this policy. Not the next reorg announcement. Not the town-hall optimism. Nothing, ever again. And a generous Never List you can’t keep isn’t generosity – it’s inventory you will be caught holding.
So the test for a Never List is three words: short, honest, signed. Short – it fits on one screen, because a forty-page policy is where promises go to hide. Honest – every clause survives contact with your worst quarter. Signed – by the highest name you can get: a rule signed by “the project team” weighs exactly nothing; the same rule signed by the COO is a commitment with a career attached to it. Then publish it, right next to the whitelist.
Because an over-promised Never List is just timesheet fiction – leadership edition. Fear produces round numbers. So does PR.
Right. You’ve written down what you will never do. Which leaves the harder question – the one that manager failed: when the dashboard shows you something ugly, and it will, what is the first thing you’re actually allowed to do?
Rule 2: Ask “What’s blocking you?” before any consequence
So the Never List is signed. You know what you’ll never do. Here’s the problem: the dashboard doesn’t care. Sooner or later it will show you something ugly – a team at, say, forty percent passive time, one person’s queue twice everyone else’s, a process taking three times longer than the SOP swears it does. And every management instinct you’ve ever been rewarded for says: you see a problem, you act on it. Fast. Decisively. That’s what leaders do, right?
That instinct is exactly what we watched in the cold open. That manager acted fast and decisively. Fifteen minutes, problem “solved”, dataset dead.
So the second rule of use – after the Never List – is the first-response rule: the first response to any bad reading is a question, not a consequence. And you already know the question, because it’s the same one from the collection episode, decision four: “What’s blocking you?”
Run the cold open again with that rule in place. The manager sees the empty afternoons. Instead of a verdict, he asks the question. And the true answer is sitting right there in the data: nothing is blocking him – he’s efficient and underloaded. You’ve just discovered spare capacity in one of your best performers. Give the man something bigger, and thank the tool. Same data, one question, opposite outcome. The question costs nothing. The verdict cost the truth.
And I can hear the pushback – I’ve had it, nearly verbatim, in a steering committee: “Andrzej, if the data can’t trigger consequences, what exactly are we paying for?” You’re paying for the questions. Because the honest answer to “what’s blocking you?” is almost never “I’m lazy.” It’s a broken handoff, a frozen system, a pointless approval – and each of those, fixed, is worth more than any dressing-down in the history of dressing-downs. Consequences fix one person. Maybe. Questions fix processes.
The first-response rule also has a second half: altitude – who gets to ask. Every finding has a level, and it gets acted on at that level, by whoever owns the process – not whoever manages the person. A team-level pattern goes to the process owner. A four-day approval queue goes up, to the leadership that owns the approval – and yes, that direction works, because in decision five we put management on the measurement too. Meetingorrhea, decision bottlenecks, the director’s inbox where processes go to hibernate – those are findings, and “what’s blocking you?” gets asked upward as well. And I’ll tell you from experience: nothing buys credibility for the whole program faster than the first time a leadership process gets fixed because the sensor said so.
Does that mean managers of people see nothing? No – it means role-based access from decision three finally earns its keep: the altitude you act at is the resolution you see. The manager who could misuse individual data simply never receives it. And if a manager is the micromanaging kind – let’s be honest, they were already micromanaging, just with less data and more bias. Written rules don’t create that risk. They constrain it for the first time, and they turn misuse from “a Tuesday” into “a violation with a name.”
One more pattern, because it answers a fear your best people carry and never say out loud: the fast colleague who is fast because they skip steps. The data sees how work is done, not just how fast – skipped compliance checks show up as deviations; the industry even has a name for it, Maverick Processing. And here the first-response rule pays off twice. Because when someone skips a mandatory step, there are exactly two possibilities: they’re cutting a corner – or the step is pointless, and they’ve been quietly proving it for a year. A question finds out which one. A verdict never does. Either way, your diligent people finally stop losing to the fast-and-sloppy ones – for the first time in the history of office work, thoroughness is visible.
Which brings us to rule three. The one people actually believe. Because nobody believes paper – they believe what you do first.
Rule 3: Pick a first use that builds trust
So the rules are signed, the comms are sent, and here comes the instinct nobody questions: “now we go where the money is. The dashboard says the biggest opportunity is process X – that’s our first project, the CFO is already excited.”
Slow down. Because while the steering committee is reading business cases, your employees are reading something else: the first move. Nobody on the floor believes paper – we’ve established that. What they believe is precedent. The first public use of this data answers the only question they actually care about – “what is this thing, really?” – and whatever the answer is, it sticks for years. Which means the first use is not an ROI decision.
It’s a trust decision – and choosing it deliberately is rule number three. And if your biggest business case happens to smell like headcount, you’ve just confirmed every fear in the building on day one. You already know what fear does to your data.
So what’s the right first move?
Let’s play one more time, because in our deployments one first use outperforms everything else at building trust, and it’s rarely the one at the top of the business-case ranking. It’s not the biggest saving. It’s not the most impressive automation. I’ll give you a second.
The most credible first use of process intelligence is fixing the oldest complaint on the floor – the one the data just proved.
Remember the flip? The scope gets published, people see their own data, and somebody points at the screen: “THIS. This is the stupid, repetitive work I’ve been complaining about for months.” That – that exactly – is your first project. Not because it’s the biggest number, but because of the message, and the message is unbeatable: this tool works for the people who complained, not only for the people who paid for it. For years, that complaint was one voice against a manager’s gut feeling. Now it was a number, and the number won. Do this once, visibly, and something remarkable happens: employees stop hiding from the sensor and start feeding it. “Since you fixed that – could you look at this?” Congratulations. You’ve just recruited the entire floor as analysts.
And once that precedent is set, the next uses write themselves. Relieve a quiet hero – remember decision three: the person silently carrying fifty-plus-hour weeks whom nobody saw until now. The relief itself happens in private, no names on any slide – but word of it travels, and the message travels with it: the sensor protects. Or use the data to defend something your people already love. When the return-to-office debate arrives – and it always arrives – you’re the company that can show, with evidence, that hybrid teams deliver the same outcomes. Zuzanna asked this question in her material: do they really need to come back to the office four days a week? Without data, flexible work hangs on executive goodwill and dies in the first downturn. With data, it’s a proven operating model. Imagine being the employer whose measurement tool is the reason people get to keep working from home. That’s a first-year story worth more than any town hall.
Which leaves the last objection, and I’ll quote it the way I hear it: “Andrzej, this is all lovely, but our works council will never approve any of this.” Here’s what years of these conversations have taught me: works councils don’t reject rules. They reject the absence of rules. What gets torn apart in consultation is vagueness – “the data may be used to improve efficiency.” May? Used by whom? Decided how? So don’t arrive with a vendor brochure. Arrive with your draft Never List and your first-response rule, and ask them to pressure-test both. Two things will happen. First, they’ll punch holes in it – excellent; holes are cheaper now than in a crisis. Second, somewhere in that meeting they’ll realize this data serves their agenda too: fair workload distribution, overtime made visible, evidence for headcount arguments. I told you on this show to bring your works council, and I promised it would be fun. It is fun – when you show up with rules. It’s a massacre when you show up with a brochure.
Summary: the 3 rules in 3 sentences
So let me give you the whole episode in three sentences – the version for your steering committee.
One: the trust was never about the cuff – the exact same measurement is the doctor or the insurance company, depending entirely on the written rules of use. Two: there are three of those rules – a Never List that is short, honest, and signed; a first response that is always a question, asked at the right altitude; and a first use chosen for trust, not for ROI. Three: leaving the rules unwritten doesn’t keep your options open – it hands the pen to your employees’ imaginations, and their draft is always the worst case.
Homework: write down rules you can keep
Your homework – one page, one conversation. Tomorrow, write your three Never List clauses. Only the ones you can keep in your worst quarter – if a clause wouldn’t survive a bad year, it doesn’t survive the draft. Then take that page to the most skeptical person in your company – the works council rep, the privacy officer, the developer who reads terms and conditions for fun – before anyone sees a dashboard.
If they can’t punch a hole in it – you have governance. If they can – even better: they just saved you a crisis.
And if you haven’t done the five-sentence homework from the collection episode, start there. Collection before use. Cuff before rules.
