The Last Hundred Meters
Every outcome is the expression of a system (System Expression)
Alex de Verteuil ·
Make the rehearsal harder than you expect launch day to be.
When I train for a race, I don’t run every workout on flat ground. I want to know what happens to my pace when the grade changes and what happens to my form when I’m tired. I’ll go all out in training to find the point where I start to lose form. That helps me judge what pace I can hold from start to finish on race day.
By the last hundred meters, I’m drawing on work I did long before race day. The extra rep when I was tired taught me what good form felt like under fatigue. The harder workouts gave me room to make mistakes, recover, and try again while there was still time to learn.
I want the same thing from a workflow before a launch. Make the rehearsal harder than you expect launch day to be. Bring in the interruptions, handoffs, and competing demands the team will face. Then give people time to fix what the rehearsal reveals.
Here’s what that looks like before a launch. At a collection site, a missed handoff during a rush can leave a specimen unusable, and the patient comes back for another blood draw. In an assay evaluation, a missing signal can lead a team to repeat the full evaluation when the imaging setup was what needed checking. Both use up time the team had planned for other work.
Axiom II. Every outcome is the expression of a system.
Find the stall while it’s still small
In my experience, when a problem shows up at scale, someone has often seen a smaller version of it already. A person found a workaround, got the run finished, and moved on. The issue was manageable at that volume, and the team had a deadline. The workaround became part of how the work got done.
That’s why I pay attention to what it took to finish a pilot. Where did someone have to step in? If the same thing happens during a routine shift or at several sites at once, who will see it, and what can they do about it?
Consider an outpatient collection site where a specimen needs to wait before it can move to the next step. On a quiet morning, the person who collected it is still at the station when it’s ready. During a rush, that person gets called to check-in. The colleague who takes over needs to know what has already happened and what’s still pending. If that information lives with the person who stepped away, the workflow depends on getting them back at the right moment.
The lab records a specimen rejection. The collection team has to reach the patient, arrange another visit, and collect another specimen. The patient has to make a second trip, which might mean arranging a ride or taking more time off work. A single outcome creates different work for each part of the system. If the team responds with a refresher when the actual problem was a handoff, the next person covering the station still has to guess.
The cost of fixing the wrong thing
Now imagine an assay evaluation in which part of the expected fluorescence signal is missing. The assay may have failed. The imaging setup may also have been unable to detect the signal under the conditions used for the evaluation.
Those explanations send the team in different directions. A product investigation can pull specialists out of the field, pause scheduled evaluations, and use up the time set aside between validation and go-live. If the team repeats the run on the same setup, it may get the same result and still have no idea which explanation is right. A check with a known signal could help the team determine whether the imaging setup can detect what the assay is expected to produce.
The first miss creates work. Acting on the wrong explanation creates more. Even after the cause is clear, the customer visits and evaluations you moved still have to happen. Sometimes there’s only one launch window, and the time spent investigating the wrong thing comes out of the time you had to address a real finding.
Before I ask a team to repeat a run, I want to know what we expect to learn from it. If the answer is that we’ll see whether the same thing happens again, I’d ask what we can check first to narrow down why it happened.
Know what keeps you on time
I can show up late because I underestimated how long the previous thing would take. I can also show up half an hour early because I set aside more time than I needed. What I want is to arrive on time, ready to do what I came to do, with enough room for the delays I can reasonably anticipate.
High-performing teams know what keeps them on time. They know which handoffs tend to stall, which exceptions need a decision, and how much room to leave before a commitment. During validation, that might mean testing the assay and imaging setup together early enough to investigate a finding before go-live. At a collection site, it might mean rehearsing a handoff during a rush and checking whether the next person can see the specimen’s status without tracking down a colleague.
A buffer has a job. Give the team room to handle the variation it expects, then revisit that allowance as the workflow improves. Time the team can reliably reclaim can go back into serving customers, improving the process, or testing the next idea. When routine recovery keeps absorbing those hours, it gets harder to make time for any of that.
In training, I know I’ve learned something when I can hold my form on the next hard rep without having to think through every correction. I look for the same thing in a workflow. If a rehearsal exposes a fragile handoff, I want the next person covering that station to be able to pick up the work on an ordinary shift. The change has to live in what they can see, what they know to do, and what they can decide.
That’s what I mean by embodying a principle. Sustainment is part of the improvement itself. When a team carries what it learned into everyday work, it raises the baseline for what it can reliably do. The next round of improvement starts there. That’s how a system evolves.
What the repeat costs
The Hidden Work Simulator makes that trade-off visible. The comparison starts with the same incoming workload and staffing. Then it changes how the team responds to a problem. What happens if you repeat the work before you know the cause? What happens if you check the likely cause first? What does it cost to rehearse the workflow early enough to catch the problem before it affects a patient visit or a launch?
Each response uses capacity. A targeted check may prevent another failed run, but a patient whose specimen is already unusable still needs a redraw. A rehearsal takes time up front. Its value depends on what it helps the team catch and how much recovery work it prevents. The comparison makes it possible to see which appointments or evaluations get displaced, what carries over, and how much time is left to improve the process.
Before the next launch, I’d ask the team to name one stall they already know how to work around. Rehearse it under the conditions you expect on a busy day. See who notices it, who can act, and what it takes to get the work moving again. Then come back on an ordinary shift and see whether the team can do it again. That’s the work I want behind me when I reach the last hundred meters.
Compare a repeat, a targeted check, and an early rehearsal in the Hidden Work Simulator →
Explore the premise and five axioms of the A. de Verteuil Framework →
If you are preparing a launch and want to talk through what your rehearsal is showing, get in touch.