industry

Fifty ads in ten minutes

Twenty years in the trades, two years inside a construction tech company, and a growing suspicion that nobody is helping contractors figure out what's real.

By John Dye · AUG 04, 2026 · 7 min

Disclosure

John Dye was the founding go-to-market executive at XBuild (August 2024 to July 2026). DIALED covers XBuild using only public sources, and never for pay. Read the standard.

I was sitting at the airport a few weeks ago with time to kill, scrolling Facebook.

An ad came up for a construction tech product. Then another one. Then another. So I started screenshotting them.

In under ten minutes I had fifty. Fifty companies, fifty products, all aimed at contractors. Every one of them promised to change everything. Revolutionary. Game-changing. The future of the trades.

Fifty. In ten minutes. On one guy's feed.

That bothered me more than I expected, and I have been thinking about why ever since.

What contractors actually told me this year

The most common thing I heard from contractors in the last twelve months was not "I need better software."

It was: my tech stack is out of hand.

A CRM. An estimator. A photo app. A measurement tool. An answering service. A review platform. A scheduling tool. A payments processor that does not talk to the accounting software. Three of them overlap and nobody remembers who signed up for which. The subscriptions stack up and the line item grows every year, and every one of them was supposed to simplify things.

Meanwhile there are fifty more companies in the feed promising the same thing the last five promised.

Nobody is drowning because there is not enough technology. They are drowning because there is too much of it and no fast way to tell what is real.

The uncomfortable part of my own story

I was a contractor. Twenty years in insurance restoration, supplementing, and roofing field operations. I built an audience teaching this industry how to get paid properly on supplements, and along the way I introduced a lot of you to a lot of products. Some of them changed how people ran their businesses. That was real and I am proud of it.

Then I took a role as the founding go-to-market executive at XBuild, an AI construction tech company out of Silicon Valley. Two years building something meant to make your work easier.

And while I was on the inside, the noise on your side got completely out of hand.

I was not watching the flood. I was standing in it. I know how these companies get built now, how fast they can ship, how they price, how they raise money, and what happens to a product eighteen months after somebody buys the company.

I also learned how easy it has become to build something that looks like a real product. That part is new, and it is the whole reason this needs to exist.

The cost of picking wrong went up

Anyone can build software now. A person can describe what they want to an AI and have something shippable in a weekend. The barrier that used to keep unserious products out of the market is gone.

That is not a complaint. In a lot of ways it is good, and some of the best tools in this industry are being built faster than they could have been five years ago.

Here is what did not change: your business is still on the line.

Your software holds customer addresses, photos of the inside of people's homes, signed contracts, and payment details. When it breaks in October during storm season, somebody has to answer the phone. When the company gets acquired and the price goes up forty percent and the integration you depend on gets deprecated, you are the one explaining it to your crew.

The cost of building went down. The cost of picking wrong went up.

Nobody is standing in that gap.

Nobody is answering the actual question

There is trade press for enterprise construction. There are vendor blogs, which are ads with a byline. There are review sites where vendors bid for placement and say so in the fine print. There is a Facebook group where somebody asks which CRM to buy and gets thirty answers, half with a referral link attached.

None of it answers the question a contractor is actually asking, which is not "is this good."

It is: does this work for a business like mine, and is anybody like me still using it a year later?

A tool that transforms a four million dollar painting company is a nine hundred dollar a month paperweight for a thirty million dollar GC. Context is everything, and the difference never shows up in the marketing.

What DIALED is

Here is the part I want to be exact about, because it determines whether any of this is worth your time.

I am not going to tell you what software is good.

I cannot. Nobody can, from the outside. What I can do is tell you what a product is, what it costs, who owns it, and what its privacy policy actually permits. And then ask the people using it whether they still are.

That is the whole idea. We record what is true. Contractors supply the verdict.

The Index. Every tool being sold to you, with the things nobody publishes. Real pricing. Who owns the company and who owned it before. What the privacy policy permits them to do with your data and your customers' data. Which integrations are verified versus merely claimed. Every fact carries a source and a date, and you can see which ones we checked ourselves and which ones the vendor told us.

The panel. This is the one that matters most, and it needs you. Contractors register the software they run and when they started with each tool. Every quarter we ask one question per tool: still using it? That is where "still using at twelve months" comes from. Not a survey, not a star rating, not somebody's mood on the day they wrote a review. An actual count of what people kept, broken out by trade and company size.

Your name is never attached to it unless you choose to attach it. The number counts either way.

The Field Test. You nominate a company. We go to a jobsite, a contractor runs the same real task on their current stack and on the challenger's, same inputs, same clock, one take, on camera. The company covers production, and we fund several a year ourselves. That fee buys the test and it has never bought a result. Publication is unconditional and nobody sees the episode before you do.

The show and the Summit. A podcast where a vendor brings a contractor who actually uses the product, and the contractor says what really happened. Then everyone in one room in Dallas on January 12 and 13.

Three rules

Everything above is worthless without these, so they are on the standards page permanently.

Nobody pays for coverage, placement, or a rating. No sponsored posts, no affiliate links, no vendor paying to change a listing or remove a review.

Conflicts are disclosed. Mine included. I spent two years at XBuild through July of this year. I do not work there anymore and I am not in those conversations. I will cover them the same way I cover everyone else, from public sources, and that disclosure appears every time they come up.

We report. Contractors judge.

A note to the software companies

This is not a hit piece operation, and if you are building something real you should want it to exist.

Right now you compete against companies that will say anything, and there is no referee. A directory that asks everyone the same questions and publishes the answers is good for the people telling the truth. Fill your page in. Take the Field Test. Bring a customer on the show.

The ones who do will be obvious, and so will the ones who don't.

The point

Your business is not a use case. The way you run crews and price jobs and handle customers is specific, and most of it is specific for good reasons that took years to learn.

You should not have to become a software analyst to buy a CRM. Evaluating one properly is fifteen or twenty hours of demos and reference calls and reading terms of service that nobody reads. You do not have fifteen hours. You have jobs to run.

That is what this is. Those hours, already done, kept current, and free.

Register your stack. It takes two minutes, it is what makes the numbers real, and it is how you find out what is happening to the software you already pay for.

Errors are corrected on the page with a dated note. See our editorial standard.