The Course They Were Told To Stay

What the Industry Was Doing

While I was restructuring CI Web Group, the rest of the trades agency industry was operating as though nothing had changed.

I want you to understand what that meant in practice, because the practice is the part of the story that matters. This is not an abstract argument about industry trends. This is a description of what most contractors’ actual marketing partners were doing, every day, between 2023 and the moment you are reading this book.

The agency that most contractors were paying in 2024 was building WordPress websites. The same WordPress, with the same plugin architecture, with the same code-to-content ratios, with the same security exposures and the same load-time issues that I described as a strategic liability in the previous chapter. The agency was not building anything else, because the agency had not yet absorbed that anything else was needed. WordPress had worked for the previous decade. WordPress was therefore going to keep working. That was the operating assumption.

The content that agency was producing for the contractor was a blog post per month. Sometimes two. The blog post was usually written by a junior copywriter or, increasingly through 2024 and 2025, by an AI tool that produced generic, formulaic, unsourced text the agency was passing off as original work. The blog post had a target keyword. The keyword had been chosen by a tool that scraped Google’s suggestions for the contractor’s service area. The post got published. The post sometimes ranked. The post mostly did not. The agency reported the rankings the post did achieve and did not report the ones it did not. The contractor read the report, did not understand most of it, and signed off on another month.

The traditional SEO work that agency was doing was the same work it had been doing in 2018. Title tags. Meta descriptions. Header tags. Internal linking. The schema markup that was state of the art in 2017. The Google Business Profile management that was standard practice in 2019. The agency had not added anything substantial to its SEO methodology in five to seven years, because the methodology had been working in a pre-AI search era and there had not been a reason, the agency thought, to update it.

The reporting that agency was sending to the contractor every month was a dashboard pulled from Google Analytics, Google Search Console, and the contractor’s call-tracking platform. The dashboard had charts. The charts went up and down. The narrative around the charts said the work was producing results. The contractor was not in a position to evaluate whether the results being claimed were the actual results, or whether different work would have produced different results, or whether the trend lines were reflecting the agency’s contribution at all rather than reflecting the contractor’s organic referrals from neighbors and existing customers.

That is the agency landscape that most trades contractors were buying from in 2024 and 2025.

It was, by every standard of the previous decade, professional, competent, and adequate. It was not, by the standards of the era we are entering, any of those things. The agencies operating that playbook are going to spend the next three to five years either rebuilding around AI or going out of business, and the contractors paying them are going to spend the same three to five years either changing partners or losing ground.

That is what was happening underneath the part of the trades industry that was not paying attention. While I was sitting in a Dallas office watching my CTO and my AI engineers build agentic workflows on n8n and LangChain, most of the agencies my industry was competing with were still arguing about whether to upgrade their WordPress plugins.

The Pressures They Were Not Watching

Meanwhile, the contractors those agencies were serving were getting their ass handed to them.

I want to use that phrase deliberately because the polite version of it understates what was happening. Trades contractors in 2024 and 2025 were facing a level of structural competitive pressure that none of them had faced before, and most of their advisors were not telling them about it.

Private equity was rolling up the trades. Five years ago this was a regional phenomenon. Today it is national. PE-backed roll-ups are buying HVAC, plumbing, electrical, and roofing companies in markets across the country at a pace that is reshaping the competitive landscape of every trade in the United States. The contractors being absorbed are getting paid for their businesses, often well, and walking away. The contractors not being absorbed are now competing against well-capitalized, professionally-managed, technology-enabled regional brands that did not exist five years ago. The independent contractor whose grandfather started the business in 1962 is now competing against a PE roll-up that has the marketing budget of a Fortune 500 division. That is a different competitive environment than the one most contractors trained for.

Major retailers were entering the trades. Home Depot, Lowe’s, and the other national retail chains have been quietly building out their HVAC and plumbing service operations for years. The customer who needs a furnace replaced in 2026 can now buy one from Home Depot and have Home Depot’s installation network do the install. The customer does not need to find a local contractor anymore. The retailer has become the contractor. That is a different customer-acquisition environment than the one most contractors trained for.

Energy companies were entering the trades. Several major North American utilities have launched HVAC service offerings in the last three years, leveraging their existing customer billing relationships to cross-sell installation, maintenance, and repair services. The utility customer who has been paying their gas bill for fifteen years now gets a marketing email from the utility offering them a furnace replacement. The trust the utility has built through fifteen years of bill-paying is now competing with the trust the local contractor has built through fifteen years of in-home service calls. That is a different competitive trust environment than the one most contractors trained for.

BlackRock and other institutional investors were buying single-family rental properties at a scale that is changing the customer base of the trades industry. The homeowner who used to be the trades industry’s primary customer is being replaced, in market after market, by a corporate property management company that has procurement contracts with national-scale service vendors. The contractor whose business model was built around homeowner relationships is now competing for the property-management contract that the corporate buyer signs once a year and renews based on price. That is a different sales environment than the one most contractors trained for.

And the technology that was supposed to be helping contractors keep up with all of this was lagging.

I want to flag one specific data point because it tells you everything you need to know about how far behind the industry’s technology layer actually is. As I write this book in 2026, the major field service management platforms in the trades industry — the systems that run dispatching, scheduling, customer records, technician routing, and work-order management for hundreds of thousands of contractors — do not have an MCP server available. Not one of the majors. The Model Context Protocol has been the emerging industry standard for AI-system integration since late 2024. Every other major business-software category has shipped MCP integrations. The trades field service management platforms have not. That single fact tells you that the technology layer underneath the contractor industry is operating eighteen to twenty-four months behind the rest of the business-software economy. The contractor whose dispatching system cannot talk to an AI agent is the contractor who cannot deploy the operating capabilities the rest of the economy is deploying. That contractor is being held back by their own software vendor.

So the contractor in 2026 is competing against private equity, major retailers, energy companies, and institutional landlords — while operating on technology that cannot speak to the AI systems that are about to define how all of those competitors operate.

That is the competitive environment. Most of their advisors were not telling them about it.

What the Advisors Told the Contractors

What the advisors were telling them, instead, was to stay the course.

I want to be precise about who I mean by advisors, because the category is broader than the trades agency landscape I described in section one. The advisors include the agency partners. They also include the industry consultants, the conference speakers, the business coaches, the franchise systems, the manufacturer dealer programs, the trade-association leadership, the trade-publication editorial voices, the buying-group executives, and anyone else whose job involves telling contractors what to think about the operating environment they are inside of. That is a substantial economy of advisors. Most of those advisors built their careers on the playbook of the previous decade.

The playbook of the previous decade said: optimize your Google rankings, run paid search ads, manage your reviews, train your call-takers, run a referral program, send out direct mail, attend the right conferences, and keep your truck wraps clean. That playbook worked. It worked because the customer-acquisition pipeline of the trades industry was, for two decades, dominated by a small number of channels — Google search, Google Maps, Yelp, the Yellow Pages becoming Angie’s List becoming Angi, word of mouth, the local newspaper. Those channels had operating physics that the playbook had been built around.

The advisors who built their careers on that playbook are now telling contractors that the playbook still works.

They are telling contractors that AI is overhyped. They are telling contractors that ChatGPT is a fad. They are telling contractors that Google’s rankings still matter most. They are telling contractors that the same SEO methodology that worked in 2018 still works in 2026. They are telling contractors that the agency they are paying is doing the right work. They are telling contractors not to chase shiny objects. They are telling contractors that the people who are warning about AI are the same people who warned about every other technology and turned out to be wrong.

They are wrong.

And I do not have to make that case alone anymore. While I was finishing this book, one of our clients — John Dean, who owns Superior Air Duct Cleaning in Pennsylvania — sent me his account of watching this exact dynamic play out around him, including a call he was invited to whose entire purpose was to convince him that leaving the old playbook was a mistake. His full story is in Voices from the Clients, at the back of this book. This is the sentence of his I cannot improve on:

Change is expensive. Staying the same is comfortable. It's much easier to criticize the future than invest in building it.

I want to be careful with how I say this because the people I am describing include some of the most influential voices in the trades industry, and many of them are people I respect personally. They are wrong anyway. The fact that they are people of integrity and experience does not make them right about the present. They are operating from inside a model of how the trades customer-acquisition environment works that was true ten years ago and is no longer true now. They have not yet absorbed that the model has changed. The advice they are giving is therefore advice that worked under a model that no longer applies. Their integrity is not the issue. Their model is.

And the contractors taking their advice are paying the price for the model lag.

Every quarter that an advisor tells a contractor to stay the course is a quarter that the contractor does not invest in the AI infrastructure that the next decade is going to require. Every quarter that the contractor stays the course is a quarter that a competitor who is not staying the course gains ground. Every quarter that ground is gained is a quarter the contractor cannot recover, because the variable being lost is not money or attention or capability. The variable being lost is time.

The Cost of Waiting for Proof

The argument the advisors keep making, when pressed, is that contractors should wait until there is proof.

Proof that AI search is real. Proof that consumer behavior is changing. Proof that the agentic-browsing era is going to arrive in the way the early movers say it is. Proof that the cost of restructuring an agency or a contractor business now will pay off in three to five years.

That argument sounds reasonable. Most contractors who hear it nod. Most advisors who make it sleep well at night believing they are protecting their clients from getting ahead of themselves on a technology trend that may or may not pan out.

I want to tell you why the argument is wrong, and I want to tell you using a parallel from the history of this industry that the contractor reader will recognize immediately.

In 2008, when I keynoted the Mitsubishi Diamond Dealer conference and asked who in the room was running search-engine optimization on their websites, three contractors raised their hands. The rest of the room did not. The advisors of that era were telling contractors the same thing the advisors of this era are telling them now: wait until there is proof. Wait until you can see Google rankings translating into measurable lead volume. Wait until your peers have done it first. Wait until the technology is mature.

The contractors who waited never caught up.

That is not a metaphor. That is the literal observable history of the trades industry over the last eighteen years. The three contractors who raised their hands at Mitsubishi in 2008 became, over the next decade and a half, dominant operators in their markets. The contractors who stayed the course in 2008 became, over the same period, the operators trying to figure out why their lead volume kept declining and what they could do about it. The early movers built positions in organic search that the late movers could not catch up to, because the late movers were arriving four, six, eight years after the early movers had already established the authority that Google’s ranking algorithm was rewarding. The late movers could buy paid search to make up the difference. The cost of paid search kept rising. The early movers were not paying the same costs because the early movers had organic. The gap kept widening. The gap is still widening.

The contractors who waited never caught up.

The companies that focused on Google search early on, on web content, on the foundational SEO work — if they did not make any major errors — are now the juggernauts of today’s organic search environment. They will never slow down so somebody else can catch up. There is no version of competitive dynamics where the established juggernauts of an organic-search position voluntarily lose ground to allow late-arriving competitors to catch them. The late arrivals have to compete in different categories, with different strategies, in different addressable markets, because the position the early movers built is not available anymore. It belongs to the early movers. Permanently.

That is the dynamic that is playing out right now with AI search.

The contractors who are building AI-era infrastructure right now — the website architecture, the content strategy, the agentic readiness, the AAO framework, the technical foundation that lets AI agents read and rank and recommend their businesses — are building positions that will be uncatchable to the contractors who wait for proof. The proof, when it arrives, will arrive in the form of the late movers realizing that the early movers have already taken the AI-search positions in their markets and that there is nothing the late movers can do to displace them. The proof will arrive too late to act on. That is what *too late* means in competitive dynamics. Not that the technology stopped working. That the positions had already been taken by the operators who acted before the proof was available.

The writing is on the wall. The advisors keep telling the contractors to wait until they have proof. The contractors keep waiting. By the time they have proof, it is too late.

That has happened before. It is happening again. It will keep happening, in different forms, for as long as there is an industry that has advisors and contractors and a competitive environment that rewards early movers. The only protection a contractor has against this dynamic is to recognize it before it has finished and to act on the recognition while there is still time.

The Only Thing You Cannot Change Is Time

There is one variable in business that cannot be recovered.

Money can be replaced. Money is not the constraint most contractors think it is. Most contractors who tell me they cannot afford to invest in AI-era infrastructure are mistaking the cost of the investment for the cost of not making it. The cost of not making the investment is, over a five-year horizon, larger than the cost of making it. They will pay either way. They will pay more if they wait.

Technology can be acquired. Most of the technology that any individual contractor needs to operate inside the AI era can be bought, licensed, partnered for, or built by an agency partner who has already done the work. The contractor does not have to build their own agentic stack. The contractor does not have to learn n8n. The contractor does not have to understand the technical depth of MCP or Synapse-style data lakes or HydraOS-style orchestration. The contractor needs a partner who has built that depth. The partner exists. The partner has been operating for several years now and is taking on clients.

People can be hired. The shortage of AI-fluent operators in the trades industry is real but solvable. The agencies that have built AI-era teams have built them in the last three years. There is no reason a contractor who decides to invest in AI-fluent talent cannot have the talent in place inside twelve to eighteen months. The barrier is the decision, not the supply.

Time, however, is different.

The only thing you cannot change is time.

That sentence is the central operating insight of this entire book. I want you to register it carefully. Money you can earn back. Technology you can acquire. People you can hire. Skill you can develop. Capacity you can build. Reputation you can rebuild. Customers you can re-acquire. Markets you can re-enter. All of those variables are recoverable.

Time is the asymmetric variable. The quarter you spent waiting cannot be retrieved. The year you spent waiting cannot be retrieved. The five years you spent waiting cannot be retrieved. While you were waiting, your competitors who were not waiting were building positions that compound. The compounding does not pause to let you catch up. Every quarter you delayed, the position you would have had to build to compete fell further out of reach, because the position your competitors had built had been compounding the entire time.

This is the asymmetry that the *wait until there is proof* advisors do not understand. They are evaluating the cost of acting now against the cost of acting later. They are not evaluating the cost of acting now against the cost of acting later *adjusted for the compounding gain that competitors will accumulate during the waiting period.* The adjusted cost of waiting is dramatically higher than the unadjusted cost. The unadjusted cost is what the advisors are calculating. The adjusted cost is what is actually happening.

The early movers in AI search are compounding right now. They started compounding eighteen months ago. They will continue compounding for as long as the AI search era continues to develop, which is going to be at minimum another decade. Every quarter that a contractor stays the course is a quarter that the early-mover compounding curve pulls further away from the contractor’s starting position. The math gets worse, not better, with time. There is no version of waiting that gets cheaper as the waiting continues. The waiting gets more expensive every quarter.

And the medium where the compounding is happening is itself changing.

The web that contractor businesses have been operating inside for two decades is a web of pages, links, search results, and human browsing behavior. Customers typed queries into Google. Google returned ten blue links. Customers clicked the links. Customers visited websites. Customers made phone calls. That has been the architecture of trades customer acquisition since roughly 2005.

That architecture is now being replaced.

The web of 2027, 2028, and 2030 is a web of agents. Customers will not browse the internet. Customers will not sift through blue links. Customers will not visit websites. Customers will tell their AI assistants what they need, and their AI assistants will do all the heavy lifting — evaluating options, comparing providers, reading reviews, scheduling appointments, even authorizing payment in some cases. The contractor whose digital footprint is optimized for human browsers in a world where customers no longer browse is a contractor whose digital footprint is invisible to the customers themselves.

The customer is no longer the human. The customer is the agent. The agency that wins the AI decade is the agency that optimizes for the agent. The contractor who wins the AI decade is the contractor whose agency optimizes for the agent on their behalf. That work is happening right now. The contractors building it are building positions that the late arrivals will not be able to catch.

The only thing you cannot change is time.

What This Means For The Reader

I want to close this chapter by being honest about what your situation probably is, as a contractor reading this book in 2026.

Your agency is probably still building WordPress sites. Your agency is probably still publishing one blog post a month. Your agency is probably still selling you on a 2018 SEO methodology dressed up in 2026 language. Your agency is probably telling you that AI is overhyped, that the technology will mature in time, that you should focus on what you can measure today, and that staying the course is the responsible thing to do.

Your industry advisors are probably reinforcing all of it. The conference speakers who told you to stay the course in 2018 are still on the same stages telling you the same thing in 2026, often with the same slides updated to reference AI as a passing trend rather than an architectural reorganization of how customer acquisition works. The trade publications you read are running articles by editors who do not yet understand the operating implications of agentic search. The franchise system or the buying group or the manufacturer dealer program you are part of is, with a few exceptions, telling you the same things they told you a decade ago. The model is the same. The advisors trained inside the model. The advisors cannot easily train themselves out of it.

You are probably looking at private equity rolling up your competitors, the major retailers expanding into your service area, the utilities starting to compete for your installation work, and the institutional landlords changing the customer base — and you are probably wondering, on Sunday afternoons at the kitchen table, whether the marketing partner you are paying is going to help you respond to any of it.

The honest answer is probably no. The honest answer is that the marketing partner you are paying is operating inside the same playbook that the rest of the industry is operating inside, and the playbook has not been updated for the operating environment you are now competing in.

I am not telling you this to scare you. I am telling you this because you cannot make a clear decision about what to do next without seeing the situation clearly. The agencies operating the old playbook are not going to tell you the playbook has changed, because telling you would require them to admit that the work they have been billing you for has not been the right work. They are not going to do that. The advisors operating the old model are not going to tell you the model has changed, because telling you would require them to admit that the advice they have been giving has not been the right advice. They are not going to do that either. The information has to come from somewhere outside the system that is invested in keeping the information from reaching you.

That is what this book is for.

The next decade of your business will be shaped by a single decision, which is the decision about whether to stay the course your advisors have been telling you to stay or to act on the recognition that the course is wrong. I am not going to make the decision for you. I am going to tell you, as plainly as I can, what I see, what I have built, what the trade-offs look like, and what it has cost me — and I am going to leave the decision in your hands, where it belongs.

Whatever you decide, decide with the time variable in mind. The asymmetry is real. The compounding is happening. The advisors who are telling you to wait until there is proof are telling you to wait until it is too late.

The only thing you cannot change is time.

Before I close this book, I owe you five more chapters.

The first is the chapter where I am honest with you about why this transition is genuinely hard for most operators in the trades industry. The camp dynamic Chapter Eight critiqued is not the whole story. There are real cognitive, structural, and economic constraints that are keeping operators stuck in the previous decade’s playbook, and the contractor reader who is trying to understand why their current advisor has not made the AI investment will recognize their advisor in the four problem states this chapter names. Empathy without endorsement. The constraints are real. The path through them is also real.

The second is the chapter where I tell you what good industry leadership actually looks like in this transition, because the contractor reader deserves to know who is getting it right. There is a coalition of operators in the trades industry who have been leaning in to the AI transition with the seriousness and the sophistication the moment requires. I am going to name them. I am going to tell you what they are doing. I am going to make the case for why the contractor reader should be looking at these operators as the cohort to watch.

The third is the map. AI is reorganizing the entire economic and human system in 2026, and the reorganization is happening from six different perspectives at once — the labs creating the technology, the platforms adopting it, the consumers absorbing it, the businesses adapting their marketing for it, the businesses leveraging it operationally, and the individuals deciding whether to evolve alongside it or be replaced by it. You cannot make a clear decision about your own business without understanding all six.

The fourth is the forecast. Where AI is going next, what we are racing to build, and what contractor businesses need to start preparing for now even though the future they are preparing for does not yet exist. Trying to build for something that does not exist is hard. I have been doing it for three years. I am going to tell you what I see.

The fifth is the chapter I have been quietly writing toward since the Foreword — the chapter that the Prologue’s final sentence pointed at, and the chapter that the line that runs through this book makes inevitable. It is about the hardest decision I have ever made, in a hospital in Seattle, with my mother on one side of a hospital bed and a trial team of doctors on the other. It is also about what AI could have done for him then and what AI is going to do for the next person like him now. It is the chapter where I tell you that AI is scary, that the fears are legitimate, and that the good AI could do is also legitimate — if we lean in, and if we act responsibly. That chapter closes the book.

First, the diagnostic.