Shrink the Loop: What Change Actually Looks Like Inside Companies Right Now
| Chris Dyer, named the #1 Leadership Speaker to Follow in 2026 by MSN.com, spent this year inside sales organizations, HR leadership groups, manufacturers, and associations working the same problem: change arriving faster than the organization can absorb it. His framework, Shrink the Loop, treats the gap between a change appearing and the business responding as the thing to measure, across four stages: See the Wave, Make Sense, Make Work Work, and Protect the Humans.This guide covers five lessons from those rooms, including why the order People, Process, Tools, Technology decides whether a rollout works, why saved hours quietly disappear, and what the meeting math actually costs. |
Table of Contents
- Why Change Feels Different This Year
- Shrink the Loop: The Four Stages
- Five Lessons From This Year’s Rooms
- The Monday Test
- Where I Got This Wrong
- Frequently Asked Questions
A sales leader told me in March that his team had been through four major systems changes in two years and had fully adopted none of them.
He was not complaining about the tools. He liked the tools. He was describing something else, which is that his people had learned that a rollout is a thing you wait out.
I heard some version of that in almost every room this year.
Why Change Feels Different This Year
The complaint is not new. The compression is.
Water power and textiles ran roughly sixty years. Steam and rail, about fifty-five. Electricity and the combustion engine, around fifty. Petrochemicals and aviation, closer to forty. Digital networks and software, about thirty. Whatever we want to call the current wave, the working estimate is twenty-five years and shrinking.
A manager who came up in a fifty-year cycle got to learn one operating model and run it for a career. A manager today gets a fraction of that, and the change arrives while the previous change is still half-installed.
The Anthropic Economic Index, which tracks how Claude is actually used for work tasks rather than how AI is discussed, gives a sense of where that pressure is landing. In its May 2026 data, tasks commonly done in Computer and Mathematical roles account for about 24 percent of global usage. Sales and Related accounts for roughly 9 percent, Management about 6 percent, and Business and Financial Operations about 6 percent. The split between augmentation and automation across all usage sits near even, at roughly 51 percent to 49 percent.
Those are observations about tasks, not predictions about jobs, and the index says so plainly. What they tell a leader is narrower and more useful: this is not confined to engineering, and the balance between AI doing work and AI helping someone do work is genuinely unsettled right now.
Which is why the useful question stopped being what is coming. Every leadership team I met this year could already name what was coming. The question is how long it takes them to do anything about it.
Shrink the Loop: The Four Stages
Shrink the Loop is the framework Chris Dyer uses to measure that lag. The premise is that organizations do not fail at change because they lack information. They fail at a specific stage, usually the same one every time, and the stage is identifiable.
| Stage | The job | How you know it is broken |
| See the Wave | Awareness | Your competitors, customers, or regulators tell you about the shift before your own team does |
| Make Sense | Translation | Every department has a different story about what the change means, and the frontline has a fourth one |
| Make Work Work | Adaptation | New tools sit on top of old workflows and nobody logs in after week three |
| Protect the Humans | Sustainability | You learn how tired people were during the exit interview |
Most organizations are strong at one or two of these and quietly terrible at another, and the weak stage tends to stay weak for years because nobody owns it. A company with excellent awareness and no translation capability generates a lot of accurate memos that change nothing. A company that adapts well but never protects the humans posts good numbers for six quarters and then loses the three people who were carrying it.
The diagnostic value is in the pattern. Have your leadership team score all four from one to five, then have three frontline managers score the same four anonymously. The gap between those two sets of numbers is the real baseline, and it is usually more informative than either score on its own.
Five Lessons From This Year’s Rooms
1. The sequence is almost always backwards
People, then process, then tools, then technology. That order does not change, and companies invert it constantly. They buy the platform, then try to bolt the humans onto it, then wonder why adoption stalls at week three.
The autopsy question I now ask first: what workflow changed before the tool arrived? If nobody can answer in one sentence, the purchase was premature. That pause costs a quarter. The shelfware costs two years and a chunk of leadership credibility, because somebody has to keep defending the decision in budget meetings long after everyone stopped using the thing.
2. Everyone sees wave one, nobody owns wave three
Ask a leadership team to name the change heading for them and you get a fast, confident answer. Ask for the one nobody in the building is watching and the room goes quiet, which is the actual finding.
The fix is unglamorous. Assign one named person to scan the category monthly, because scanning that belongs to everyone belongs to nobody. Add a ten-minute standing item to the leadership meeting that is a question rather than a report: what did you see this month that we did not expect? Then define the tripwire in advance, the specific signal that would force a move, so the debate happens once instead of every quarter.
3. Saved hours refill with admin unless you spend them on purpose
Teams that automate something almost never track where the recovered time goes, and it reliably drains back into low-value work within a month.
The sort that works: automate the routine and low-judgment tasks, augment the complex ones where AI drafts and a human verifies, double down on the relationship and judgment work that only gets more valuable, and monitor the unproven experiments without scaling them yet. Then do the part everyone skips. Write down the hours you actually saved on the first item and move those specific hours into a double down task. Time saved that quietly refills was never saved.
4. Meetings are the only recurring expense approved once and billed forever
Nobody re-signs the contract on a standing meeting. It gets created in a moment of need and then bills the company every week until someone leaves.
| The formula | One example | Per year |
| People x minutes x times per month, divided by 60 | 8 people x 60 minutes x weekly = 32 hours a month | 384 hours on one calendar line item |
Thirty-two hours a month on one meeting. Nearly four hundred a year. Run that math on every recurring item on your calendar and give each one a verdict: keep it, shrink it, merge it, make it asynchronous, or kill it.
The version of this that changes the most is the standing one-on-one. A weekly thirty-minute check-in with nine direct reports costs a manager eighteen hours a month, and most of that time goes to status that several people could have heard at once. Performance concerns, compensation, conflict, and anything personal stay behind a closed door. Status and pipeline updates do not. One caution learned the hard way: announce the trade out loud, because a cancelled one-on-one with no explanation gets heard as a manager who stopped caring.
5. Pace is the constraint, and it is rarely a vacation problem
Faster cycles only work if the people running them can keep cycling, and the depletion I see is almost never fixed by time off.
Sleep is one kind of rest. There are others, and they deplete separately. Mental rest, meaning space to think without eleven things open at once. Emotional rest, meaning somewhere you do not have to perform being fine, which customer-facing work burns through fastest. Sensory rest from screens and trade show noise. Creative, social, and spiritual rest, each with its own drain. A week in Florida refills the physical column and leaves the others exactly where they were, which is why people come back from vacation still flat and conclude something is wrong with them.
Run the scoring with a team after a big push. Their lowest collective score is information about how the work is being run.
The Monday Test
The research on learning transfer is unkind to conferences. Apply something within days or lose it. Which makes the only meaningful measure of any offsite, keynote, or planning session the same one: what happened on the first Monday back.
Ninety minutes on the calendar, blocked before anyone leaves the room. One page of work, not eight, because depth beats coverage. One person told what you committed to, since saying it out loud roughly doubles follow-through. One small thing live within seven days.
The reason this works has nothing to do with motivation. It works because a commitment with a date, a page number, and a witness attached is a different object than a good intention. Most conference notebooks fail at exactly one point, which is the gap between the room and the calendar.
Then a thirty, sixty, ninety. One experiment launched by day thirty. One system changed by day sixty, meaning a meeting killed, a charter signed, or a step deleted from how a customer buys from you. One measurable result by day ninety that you can show another human being.
Where I Got This Wrong
I ran PeopleG2 for fifteen years and made this mistake in the most expensive way available.
We bought a platform to fix a problem I had correctly diagnosed. Good software, real problem. I ran the rollout by scheduling training, which felt like doing the people part. Scheduling training is a logistics task wearing a people-work costume. Nobody had a reason to care, the underlying workflow stayed exactly as it had been, and within a quarter people were using the new system to do the old process slightly slower.
I defended that purchase in budget conversations for a long time, which is its own kind of tell. The tool never had a chance. I had skipped straight to it.
Bringing This Into an Event
Chris Dyer delivers Shrink the Loop as the keynote Thriving Through Relentless Change, and every audience gets a post-keynote workbook built specifically for that organization, with the frameworks above rebuilt as worksheets. Inc. Magazine named Chris Dyer the #1 Leadership Speaker on Culture, and he ranks #15 on the Global Gurus Top 30 Organizational Culture Professionals for 2026. A preview of the keynote is at youtu.be/5Y5weUZ4SrE.
Frequently Asked Questions
Why do change initiatives fail?
Most fail on sequence rather than substance. The order that works is people, process, tools, technology, and organizations routinely start at the end by purchasing a platform before redesigning the workflow it is supposed to support. Chris Dyer’s test is whether a leader can name the process change in one sentence before the money is spent. If not, the initiative is not ready.
What is the Shrink the Loop framework?
Shrink the Loop is Chris Dyer’s change framework, built around four stages: See the Wave, Make Sense, Make Work Work, and Protect the Humans. Rather than trying to predict what is coming, it measures the lag between a change appearing and the organization responding to it, and identifies which of the four stages the lag is concentrated in.
How do you calculate the cost of a recurring meeting?
Multiply the number of people by the length in minutes by how many times it happens per month, then divide by 60 to get hours per month. Eight people in a weekly hour-long meeting costs 32 hours a month and 384 hours a year. Run the calculation on every recurring item on the calendar, then keep, shrink, merge, convert to writing, or cancel each one.
What are the seven types of rest?
Physical, mental, emotional, sensory, creative, social, and spiritual. They deplete independently, which is why time off can restore the physical column and leave someone still exhausted. Chris Dyer has teams score all seven after a demanding stretch, on the basis that a low collective score is information about how the work is being run rather than a personal failing.
How long should a company take to respond to a market change?
There is no universal number, and chasing one misses the point. The useful practice is to measure your own current lag on a change you already lived through, name the stage where the delay accumulated, and work that single stage for a quarter while leaving the others alone. Organizations that improve fastest fix one thing at a time.
Where to Go Next
The culture system underneath all of this, including the seven pillars, is in The Power of Company Culture. The newest book is Moments That Matter, with a free companion workbook and no email required at chrisdyer.com/moments.
To bring Thriving Through Relentless Change to your team or conference, see chrisdyer.com/speaking, or contact Shannyn Downey at 6 Degrees Speaker Management: shannyn@6degreespeakers.com, 888-584-4177.
