The Market Moves
Faster Than
Businesses Evolve

Chapter 9

1988

In 1988 a group of us at high school produced a newspaper entirely on a computer. As far as anybody around us knew, it was the first one.

From here that sounds like nothing. It was not a small thing.

Producing a newspaper was a trade. Typesetting was a skilled job that people trained for and did for a living, and paste up was a craft with its own tools, its own vocabulary and its own quiet professionals who were very good at it. If you wanted a page laid out, you went to somebody who knew how, and you paid them, and you waited.

We were doing it in a room at school because we thought it was interesting.

By 1989 I had turned it into a business, on a Macintosh SE and a laser printer.

It did not start with an ambition. I had a job at a church. I was helping people with their materials and their branding for nothing, because I liked doing it and they needed it and nobody else was going to. It became a side business by accident, which has happened to me more than once. Almost everything I have built started by being given away.

Here is what I want you to sit with, because it is the whole argument of this book arriving early and in miniature.

The skill I had in 1988 was gone by the middle of the nineties. Not devalued. Gone as a distinct skill. It became a menu command. Anybody with the software could do the thing I had been paid for, and the specialists I had watched with something close to awe as a teenager mostly had to become something else.

I was seventeen and I did not understand what I was looking at. I thought I had found a clever way in. What I had actually found was the first of five times I would watch the same thing happen, and the only reason I came out ahead that time is that I was young enough to have nothing invested in the old way.

The people who lost were not worse at their jobs than I was. They were considerably better. They had simply spent twenty years becoming excellent at the exact thing that was about to stop being scarce.

Chapter 10

Four Times

The skill I had in 1988 was gone by the middle of the nineties. Not devalued. Gone as a distinct skill. It became a menu command, and the specialists I had watched with something close to awe mostly had to become something else.

Then websites did it again. Building a website in 1998 was a specialist act, and by 2010 it was a template and a weekend.

Then search did it. There was a period when knowing how search actually worked was a genuine edge, and then the knowledge spread, the tools got cheap, and the edge went where edges go.

Then content did it. Then AI did it, and it is still doing it, faster than the others and to more categories at once.

Four times. Five, depending on how you count. Every time, the same shape. A skill that is hard becomes a skill that is easy, and the money stops living there and moves somewhere else. I have been in the room for all of it, mostly on the winning side, sometimes only because I happened to be young.

None of which stopped me from getting the next one wrong.

Chapter 11

The One I Got Badly Wrong

I was working at a high-tech firm in Santa Barbara. We sold website builds, hosting, and T1 lines, which tells you roughly when this was. I was inside the industry. Not adjacent to it, not reading about it, inside it, with a decent view of who was actually buying what and how the money moved.

Four months before the crash, I took money that had been set aside for something important and personal and I put all of it into technology stocks.

And I did not buy the big safe names, which is the detail that matters. I bought the risky ones. The companies I thought had genuinely great ideas, because I was in the industry and I could see which ideas were great.

Every single one of them went to penny stocks.

The ego of it is awful, and it took me a long time to be able to tell it without softening the edges.

Because notice what actually happened there. I was not an outsider gambling on a hot sector I did not understand. That would be an ordinary story and a forgivable one. I was an insider, and my expertise is precisely what made me certain. I had more information than the people around me and I used all of it to be more confident, not more careful.

That is the failure I keep coming back to, and it is not a failure about stocks.

Expertise does not automatically make you a better reader of change. Past a certain point it can make you a worse one, because it hardens into certainty, and certainty stops asking questions. I have met a lot of very experienced people who were extremely sure about what AI was going to do to their industry, in both directions, and I now discount confidence more than I used to. Including my own, especially my own, and it is still the hardest one to discount.

Chapter 12

I Screamed Shark

Santa Monica Beach, 1987. I was in eleventh grade, Donnie Flaig was in ninth, and Don Kick was out there with us. Neither of them had surfed before that day, as far as I know, which is a hedge I have been using for nearly forty years and neither of them has ever bothered to correct.

I brought the boards. I brought everything. I was the one who knew what he was doing.

The three of us were sitting out past the break when I saw a fin.

I screamed at them to paddle in. Then I paddled in, which I want to be clear about, because in the version I would prefer to tell I stayed out there shepherding my friends to safety. I did not. I went, and I yelled at them from the beach.

They stayed where they were. Sitting on their boards. Laughing at me.

It was a pod of dolphins.

They still bring it up. I have been hearing about it since the Reagan administration and I expect to hear about it at my funeral, and they have earned every minute of it.

Here is why it is in this book, because it is not just a story about being wrong at the beach.

That day I was the one trying to protect Donnie. I had the experience, I had the equipment, I was the one who could read the water, and I was completely wrong about what I was looking at.

Since then he has been the one protecting me. He has protected my family through their trust and their planning. He has protected my business through a situation I did not understand while I was standing in the middle of it, which is a sentence I can write and cannot elaborate on, because that part belongs to me and stays with me.

What I can tell you is how he worked, because it is the thing I actually learned.

He went and got the facts. Not the opinions, not the version anybody was telling, not the shape of the thing as it appeared to the people inside it. The facts, assembled, in order. And because the facts were assembled, the decision that followed took about four minutes and there was nothing to argue about.

I have thought about that a lot, because it is the same argument I make for a living and I did not recognize it until somebody did it to me. Measurement beats opinion. Not because measurement is morally superior, but because a decision made on assembled facts is fast and a decision made on competing opinions is slow and never really finishes.

The man I was trying to save from a dolphin turned out to be the one who could see what was actually coming.

And Don, the third one out there laughing at me, is the man who years later built the wooden stand that held a laptop over my face when I could not sit up. There is a chapter about that later. I would not have predicted either of those things from a beach in 1987.

I am not sure there is a cleaner illustration of expertise being the wrong instrument, and it is a great deal easier to write about now that everyone involved has had four decades to enjoy it.

Chapter 13

What a Platform Says

One habit has served me better than any framework, and it is not sophisticated. When a platform announces something, I treat the announcement as a claim rather than as information, and I go and measure what actually happened to the accounts I can see.

I did not arrive at that from principle. I arrived at it from being wrong. Early on I took product announcements at face value, adjusted budgets around them, and told clients things I had not verified, on the reasonable-sounding grounds that the company that built the thing presumably knows what it does.

They know what they built. That is a different question from what it does to your money in your category this quarter, and they are not measuring that on your behalf.

The awkward part is that checking is slow, unglamorous, and produces conclusions that make you unpopular in a room full of people who have already made a plan. Nobody thanks you for it in the moment. It is also the single most valuable thing I have ever done for anybody, and it is a large part of why I think this work stays valuable when the tools get good.

The largest example I have of this is the next chapter, and it is also the one place in this book where I am going to ask you to take my word for something.

Chapter 14

Sixty Thousand Dollars Back

Google announced a new release of broad match at our franchise conference, which was held at Google's own offices.

Broad match is the setting that lets Google decide which searches your ads should appear for, rather than you deciding. The pitch for the new release was the usual one: give the system more freedom and it will find you customers you would not have thought to bid on.

I went and checked what it was actually doing to our accounts.

What I found was wasted clicks. Bogus traffic. Money going out against searches that had no plausible relationship to a commercial cleaning contract, at a volume that only becomes visible when you add it up across roughly a hundred locations rather than looking at one account and shrugging.

I want to be careful with the language here, because there are two different accusations available and only one of them is mine. I am not saying Google took the money dishonestly. I am saying the traffic we were paying for was not real traffic in any sense that mattered to the business, and that the new setting was what let it through.

I put the measurement together and I took it to them.

I asked for a hundred and fifty thousand dollars.

I got sixty.

Sixty on its own is the better headline, and a hundred and fifty makes it look like I lost. I am printing both because that is what happened, and the gap is the most honest thing in the story. I did not walk in and get handed a settlement. I made a case, the case was strong enough that they did not dismiss it, and then it got negotiated the way these things get negotiated, and I did not get everything I asked for.

Over sixty thousand dollars came back to the company. I have worked with Google since the beginning and I have never seen a rebate like that, before or since.

Now here is what I cannot give you.

I have no documentation. No screenshots, no ticket number, no letter, nothing I can put in front of you. It was years ago, it happened inside somebody else's accounts, and I did not leave with a file. This is my first-hand account of something I did, and that is all it is.

I am also not making a general claim about broad match, then or now. I measured our accounts, in our categories, at that time. What was true for us is not automatically true for anybody else, and Google never characterized the payment or agreed with any explanation of it.

This book spends a whole section separating what can be documented from what cannot, and this story sits on the wrong side of that line.

I am including it for two reasons.

The first is that the chapter before this one argues that a platform's announcement is a claim rather than information, and that somebody has to go and check. This is the largest example I have of doing exactly that, and leaving it out to keep the sourcing tidy would have made the book safer and less true.

The second is that I took the payment as evidence the case had been taken seriously. That is my reading of it and not a fact about Google. You do not have what I had in front of me at the time, so weigh the story accordingly.

Take it as what it is. One person's account of one thing he measured, with the outcome attached, the ask and the settlement both stated, and the evidence missing.

Chapter 15

Down 8-2

When I was born, the experts told my mom I would never be able to walk.

I was walking at nine months.

By 1987 I was playing junior varsity volleyball at Chatsworth High, and I was not where I needed to be. I could get better, or I could quietly stop being part of it. I had to elevate or vanish, literally. I had to practice more.

So I built a volleyball court in the back yard.

Not a net strung between two trees. A full grass court. My best friend Greg Weaver, my dad and I set metal poles into cement foundations six feet deep, with a wrench crank to keep the net tight, and hung a beach style net between them. There was one light. If you looked straight at it, you were blind.

And everybody came. Greg. Francis Chan. My sister played. Too many people to name. We all played.

I made up a song called "I Always Win." I am the only one who can sing it, and I sang it on that court many times.

I say this all the time: I am not a sore loser. Not at all. I have a lot of practice, in fact (haha). But when I win, I like to rub it in a little. Sometimes way too much. Hence the song.

On my own I did jumping drills in the back yard, and I lifted with John Fairchild, my weightlifting coach and my mentor. Years later we worked together: he was the high school pastor at Rocky Peak when I was the middle school pastor.

The next spring I moved up from JV into the starting rotation. My coach, Steve Berk, told the Los Angeles Times I was an extremely hard worker and very smart.

Senior year we did not lose a match.

We got to the City final against Palisades, at UCLA, and went up two games to none. Then we lost the next two. In the fifth game we were down 8-2.

We won it 15-11. John Fairchild was in the stands.

Here is the part that matters.

We did not come back from 8-2 at UCLA. We came back from it in my back yard, two years earlier, on nights that did not count, and the final was just the first time most people saw it.

A middle blocker has one job, and it is to go up. A blocker who stays on the ground is not a smaller version of a blocker. He is a hole in the net, and everyone on the other side can see it.

Most business owners I talk to about the ground moving under them already know they are not where they need to be. Nobody has to tell them. What they usually do not have is the court: a place where the work happens every day, with other people, whether or not anyone is keeping score, long before the game that matters.

A plan does not close that gap. Repetition does, in a place built so the repetition is hard to skip.

The score will be 8-2 at some point. For most businesses, it eventually is.

The only question worth asking before then is where your back yard is, and whether anything is being played on it.

Chapter 16

The Truck

I was built slightly wrong and nobody knew what it would cost.

My right leg is shorter than my left. The foot on that side is a size eight against a size twelve on the other. There is no hamstring on the short side. None of it stopped me playing volleyball at a level that got me a partial scholarship, and none of it was treated as a problem at the time, because I was young and it did not hurt yet.

The summer after my senior year a commercial truck rear-ended me.

I never played competitively again.

I have told that in one sentence because that is how it happened. There was no arc, no decision, no moment where I weighed my options. There was a truck, and then there was a different life, and the second one arrived without any warning that it was starting.

For years afterward I thought of the truck as the cause of everything that came next.

It was not. I had been playing on a body that was going to charge me for it eventually, and that was true long before the truck existed. The truck did not create the problem. It collected on it early, and it collected all at once instead of over the twenty years it would otherwise have taken.

I did not know I had a bill outstanding. Nothing hurt. Every measure I had said I was fine, and every measure I had was measuring the wrong thing.

I have since watched businesses run on exactly that arrangement, and I have listened to a lot of owners explain what happened to them.

AI killed us. Google changed the rules. A competitor came in and destroyed the market. Advertising stopped working. The economy turned.

Sometimes that is entirely true. Markets really do move and some of them move violently.

But often what I am being told about is a truck.

The event is real, it is sudden, and it is the thing everybody can point at. And underneath it, for years, there was a structural condition nobody had a measure for, because everything looked functional and the numbers that were being watched were the wrong numbers.

The truck gets blamed for the whole thing. The truck is rarely the whole thing.

The uncomfortable follow-up question, and I ask it of myself before I ask it of anybody else, is this: what is currently true about my business that is not hurting yet.

Chapter 17

The System Was Moving

Stand completely still and you are still moving.

The planet is turning under you. It is carrying you around the Sun, and the Sun is carrying the whole arrangement through the galaxy. You will never feel any of it, because everything you can see is moving with you at the same rate. People only know it is happening because somebody built instruments to check, and then checked.

That is the entire use I have for the astronomy, so I am going to leave it there. What it buys is one sentence, and the sentence is the reason this chapter exists.

A company can repeat the same marketing motions and never return to the same market.

I did not understand that at Stratus until well after I should have. I have already told you the part about the name, and about the four years it took to change it. This is the part underneath that story, and it is the part I actually think about now.

The playbook still looked like the playbook

For most of that period we were running a recognizable operation. Google Ads brought in leads and we knew what they cost. Search mattered, so we did the search work. Reviews mattered, so we chased reviews. There was a CRM with a pipeline in it, and campaigns that started and finished, and a reporting rhythm that made the whole thing feel measured.

None of that was wrong. The loud version of this argument is that search stopped working, or that the funnel died, and I do not believe either one and never have. Search still sends people. The funnel still describes something real, which is our own internal sequence for handling a lead once it exists.

The problem is narrower and worse. Those systems were no longer a complete description of where a customer went to decide about us.

They had been, more or less, for a long stretch. There was a period when a facilities manager looking for a cleaning vendor typed something into Google, looked at a page of results, clicked two or three, and the entire encounter with our category happened inside a surface we could work on directly. Get the listing right, get the page right, get the ad right, and you were in the room where the decision happened.

That is the assumption that quietly expired.

Where the deciding actually moved

By the end of the period I am describing, the deciding was scattered.

Somebody would meet us first in a map listing with a star rating attached. Or in a review, written by a customer of one franchisee in one market, which they had no reason to read as anything other than a statement about the whole company. Or in a comparison page written by a third party who had never spoken to us and had ranked us anyway. Or in a social post from an operator. Or, increasingly, in a generated answer that summarized the category and named some companies in it, assembled from sources we did not control and did not always know about.

Then, sometimes, after all of that, they arrived at something we had actually written.

Every one of those surfaces existed before. What changed was the order and the weight. The material we controlled moved from being the first thing a customer met to being the thing they checked our reputation against, and by the time they got to it the useful part of the decision had frequently already happened.

There is one outside measurement in this book and it belongs here as well as where I first used it. The Pew Research Center found that Google users clicked a result on 8 percent of visits where an AI summary appeared, against 15 percent where one did not. I am not going to build a theory on one study. I am going to say what it is: independent evidence that the answer surface and the click surface have come apart, published while I was still arguing internally about a name.

The point is not that any single one of these channels became dominant. It is that the number of places where a stranger forms a first opinion went up, our share of those places went down, and almost none of the new ones offer a correction mechanism.

The correction stopped being available

I said earlier that a person will let you correct them. Somebody reads Stratus Building Solutions, decides we put up buildings, and you get four seconds and a small laugh to fix it. I watched that happen thousands of times and treated it as free.

A ranking system does not offer the four seconds. It reads the name, the category, the description, the listings, the reviews and the content, notices whether those agree with each other, places you somewhere, and moves on. If it places you adjacent to your category rather than in it, nothing errors. You are simply not in the answer, and no report you run afterward has a line on it for the answer you were not in.

So the confusion tax got worse without anything about the confusion changing. The same ambiguous name, priced by a different system, at a moment when more of the deciding was being done by that system. That is the mechanism I could not articulate at the time and can now.

The sentence I wish I had said in 2022

Here is the thing I was arguing for, and here is the thing I should have been arguing for.

I was arguing that a clearer name would perform better. Which is true, and it is a claim about our position inside an arrangement I was treating as fixed. A better headline, a better category signal, a better result on a page I still pictured as a page.

Stratus did not need a better position in the old system. It needed to recognize that the system itself was moving.

Those two arguments call for different work and they call for it on different timelines. The first one is a project with an end date. The second one says that whatever you do now has to keep agreeing with itself while the ground it sits on keeps moving, which is not a project at all.

I had the smaller argument and I made it for four years.

What the response actually had to be

Once you take the second version seriously, the shape of the work changes and it stops being a marketing job.

The name had to answer the category question, because that signal is read by everything. But the reviews had to be real, recent and consistent across roughly a hundred locations, because that is the most portable trust signal a local business has and it is read by people and machines at the same time. The listings had to say the same thing as the site. The description had to say the same thing as the listings. The CRM had to know what actually happened to a lead, because a company that cannot see its own follow-up cannot tell the difference between a market problem and an operator problem. The operators had to deliver the experience the reviews were promising, and that meant recruitment and training, which are not marketing functions at all.

None of those items is exotic. What is hard is that they have to agree, and they have to keep agreeing, because the thing being scored is the agreement.

That is why the four years of work that happened while the name argument dragged on turned out to matter as much as the identity did, and possibly more. Reviews and the trust work underneath them, franchise recruitment, training, the HubSpot data and the accountability that came with it, standardized reporting, the social strategy: those moved for four years while the name did not. I spent a good deal of that period frustrated that the visible piece was stuck. Looking back, the system was being rebuilt underneath the sign the whole time, which is roughly the opposite of what it felt like.

I did not have a name for that way of working then. I call it Adaptive Brand Management now, and the definition lives with the company that sells it rather than here.

What StratusClean did and did not do

When the identity went live in April 2026, the thing it removed was interpretive work.

A stranger reading the new name does not have to place us. A search system reading it does not have to reconcile a construction signal with a cleaning service. A franchisee explaining the company at a chamber breakfast does not open with a correction. The four seconds go away, at every one of those surfaces at once, which is the only reason a name is worth four years of anybody's political capital.

What it did not do is cause the growth that preceded it, and I want that stated plainly in the chapter rather than only in the back of the book. The name did not exist for most of the four-year period. It cannot be credited with what happened during it. The tidy version of this story, in which a company changes its name and the numbers go up, is not true, and it is the version I would have to tell if I wanted this to read as a case study.

What this is actually a story about

The reason I keep this one and retell it is not that it worked. Plenty of things work.

It is that every part of it was visible years before anybody acted, including to me, and it still took four years, and the delay was not caused by stupidity or by anybody being asleep. It was caused by the fact that the old system was still producing revenue, the drag was distributed across thousands of interactions nobody was counting, and the environment underneath the whole operation was moving at a speed that no monthly report is built to display.

A company in that position does not get a moment where it is required to act. That is the whole difficulty. There is no quarter where the graph breaks, no customer who writes in to say they could not tell what you did so they hired somebody else, and no notification from anybody that you have stopped appearing in an answer you used to appear in.

You just become slightly harder to find, slightly harder to understand, slightly harder to verify, and slightly easier to leave out. Repeat that across enough surfaces and enough years and it does not read as a failure. It reads as a company that used to come up more often than it does now.

That is what vanishing looks like from the inside, and it is why the argument in this book is about moving before anything forces you to.

Chapter 18

Why Waiting Is So Reasonable

Most businesses do not miss a shift because they are lazy or stupid. They miss it because the old system still works. It just works slightly less well than it used to.

Revenue is still coming in. The team is busy. Acquisition cost has crept up but not alarmingly. Organic traffic is down a bit, and there are always reasons for that. None of it presents as an emergency, and the pressure to change is diffuse instead of acute.

Then one quarter the numbers do not bounce back. Or a competitor who started three years ago is suddenly the obvious answer in your category and you cannot work out when that happened. At that point catching up costs a multiple of what staying current would have cost, and you are paying it from a weaker position.

The moment to move is while the old thing is still working, which is exactly the moment when the case for moving is hardest to make and the person making it sounds like an alarmist.

I have been that person in that room. I understand why nobody wanted to hear it. I also think, having now watched it play out, that the room was wrong and so was I, in different ways at the same time.

Expectations jump, they do not drift

The other thing that catches people is that customer expectations do not rise smoothly. They sit still for a while and then they move in a step, usually because somebody in an adjacent category raised the standard and now everybody has been trained.

People now expect to find accurate information about a business before they engage with it. They expect reviews, plenty of them, recent ones. They expect a site that works properly on a phone. They expect the answer to arrive in a few seconds, from whatever they happen to be asking, without them having to go and look.

None of those expectations were built by your competitors. They were built by the largest companies in the world spending enormous amounts of money, and then applied to you.

If you built those systems years ago, the jump is a moat. You did not do anything on the day the standard moved, and you were suddenly further ahead. If you did not, then every customer who was considering you is doing the same research and finding somebody who did.

That asymmetry is why I keep pushing the timing argument. The work is the same work either way. It is only the cost that changes, and it changes a lot.

Adaptation is a posture, not a project

The other mistake, and this one I see constantly, is treating adaptation as something you complete.

The website gets rebuilt. The branding gets refreshed. A review campaign runs. And then everybody exhales, because the adapting has been done, and the market is expected to hold still for a few years out of politeness.

It will not. It never has. The businesses that stay in front of this are not the ones that had one excellent year of investment. They are the ones that built noticing into how they operate, so the next shift arrives as an adjustment instead of an emergency.

That is a harder thing to sell than a project, because it never finishes and there is no ribbon to cut. It is also the version I have seen work.