Caterpillar took 250 orders for a trailer it hadn't built yet. Nobody who ordered one knew Caterpillar was behind it.
Nancy Yaklich ran new product concepts at Caterpillar, the construction equipment maker that turned 100 in 2025. For most of its history, Caterpillar figured out what to build by asking the contractors who already bought its machines, a small group of large customers. Nancy knew the limit of that habit: big customers tell you what they want — but what they want is often not what the wider market will actually buy. She wanted to hear from the contractors who never call Caterpillar. And she wanted to test ideas without the company's name attached, so a concept could fail quietly, instead of failing as Caterpillar.
This is the usual trap of the messy middle: a team commits to building something before it knows whether the market wants it. However, Nancy's whole setup was built to answer one question first, and cheaply: will anyone actually buy this?
Nancy set up a second company called Eco Forge Solutions, with its own website. Nothing about Eco Forge Solutions pointed back to Caterpillar. That let Nancy test a concept with contractors across the whole market, not just Caterpillar's own customers, and keep the Caterpillar name out of it — until she wanted it in.
Getting legal and finance on board was a challenge. A Caterpillar team operating under a hidden brand, free to act on its own, was exactly the kind of setup that made them nervous. Nancy got them on board by pairing that freedom with clear rules for what the brand could and couldn't do.
She still had to show the tests had discipline. Caterpillar's leaders wanted numbers, so she borrowed a phase-and-milestone system she had used at Cargill, the food and agriculture company where she worked before, and set the evidence each concept had to hit before it earned more money. Then, she built a rough formula for how many contractors to talk to, taking the top 400 US contractors from the trade list Engineering News Record, multiplying by a response rate, and dividing by two.
The formula gave leadership a single target number they could sign off on. It also set the bar Nancy had to clear: five contractors had to name the same problem without being prompted, and twenty-seven had to ask to hear more.
Nancy set those two numbers before running a single test. That is de-risking in practice: decide up front what evidence a concept has to produce, so spending more money depends on what the market actually does.
The first concept was the Net Zero trailer, a replacement for the site office every contractor knows. That office is usually a metal box with weak heat, weak air conditioning, and traps in the corners for mice, and the superintendent still spends most of the day in it. Nancy's team, just two people covering North America, talked to about twenty contractors, most of them outside Caterpillar's customer base. They kept hearing the same thing - site trailers are far too expensive. For example, one company, Pepper Construction, made good green site trailers, but at $300,000 each, well beyond what tight construction margins allow. Contractors wanted a trailer like that at a price they could actually afford.
Eco Forge ran the test. Nancy hired an outside firm and put the trailer through A/B ads that kept the pictures the same and changed the message, to see which benefit landed. She ran it in two waves. The first wave checked her guesses about who the customer was, which benefit mattered, and which channel worked. Google ads brought nothing, but LinkedIn did the job: a short form there collected each contractor's details and a first read on what they'd pay. The second wave built on that, dropping Google, tightening the LinkedIn targeting, and asking what size trailer to build.
Within two months she had validated the customer, confirmed the problem, and pulled in 179 qualified leads. "That's the beauty of doing things digitally," she says: the whole read on the market came back in about a month.
Then, she also tried industry publications, such as Construction Superintendent and Green Builder, on the theory that contractors don't spend their time on LinkedIn. Both drew few leads. That told Nancy the trade press wasn't worth the spend, a useful thing to know before running the next product through the same system. A final round tested price, to see what contractors would pay.
By the end, more than 250 contractors had raised their hand for the trailer, and one large construction company wanted to order more than 100. That order, Nancy says, "gave us a sense that we were on to something real”. The tests also gave her the whole go-to-market package, the value proposition, the channel, the target customer, and the price, all confirmed before anyone welded a trailer.
The messy middle is where innovation teams get stuck defending guesses they were forced to make. Nancy skipped the guessing. She measured demand for a product that did not exist and learned what the market would pay before Caterpillar spent money building it, all under a name nobody could trace back to the company. By the time the trailers go out, the orders are already in: eight built, one sold, a second buyer waiting, and two rented to two of the largest construction firms in the country. Another Caterpillar team is now running the same tests on a different product.
That is what the messy middle looks like when it produces real demand instead of guesses: prove the market wants it, then build.
