Horizontal technology wins industrial pilots and then stalls at the purchase order, because of how plants buy.

Three seats supply the enthusiasm, and none of them can sign. The others rarely appear in a pilot, and each can stop the deal alone. Security does so most often, and late.

The usual loss is not a competitor. In more than 2.5 million recorded sales conversations, 40 to 60% of deals ended in no decision, and more than half of those came from indecision, not a preference for the status quo (Dixon and McKenna, The JOLT Effect).

Papers: The committee nobody mapped · Through the integrator

Published filings show the route is a choice, not a default. One automation vendor transacts about 65% of global sales through independent distributors; one engineering software vendor sells about 75% direct, and another about 94% direct (FY2024 and FY2025 10-Ks). Which one fits your product is what the partner tests decide. → The seven routes and how each is tested

Thing Company's founder has sold through four of the routes above himself: developer-led, with an open-source software platform that reached enterprise adoption from a free download; OEM-embedded, with a white-label subscription sold through carriers and device makers and billed on the partner's bill; direct, with enterprise sales from founding to exit at a subscription billing company its first channel partner acquired; and distributor-led, with partner programs built on deal registration, market development funds and rebates, run from an industrial venture capital firm. Today he sees the channel from inside a group that is at once an ISV, a reseller, an integrator and a managed services provider. → How the practice started

How do you handle conflicts with integrators, resellers or vendors we compete with?

They are checked before scoping. Current roles are disclosed in writing before scoping, and we take an engagement only where nothing the practice or its lead is involved in could bear on the verdict. No equity, referral fees or vendor compensation in anything we assess, and the fee is fixed before the research starts. Where there is a conflict, we say so and, where we can, point you to someone who fits.

Should we sell into plants as SaaS, on-premises or at the edge?

Test it with the people who can stop it. OT teams run controllers and historians under strict change control, and in Cisco's 2026 research 40% of industrial decision-makers named cybersecurity as the biggest obstacle to scaling AI. Whether a plant will run your product in the cloud, on its own servers or at the edge decides who reviews it and which budget pays. For autonomous systems, price and package for the step the buyer is on; the agentic AI and physical AI pages set out the steps.

Why does our industrial pipeline fill and not convert?

Usually one of three causes: an ideal customer profile that fills the pipeline with accounts that will never close, a value case that persuades the champion and misses the economic buyer, or a commercial model that does not fit how the plant approves spend. More sales capacity fixes none of them. Interviews with the accounts that stalled, not the ones that closed, show which one you have.

We are a system integrator. What would you test for us?

The practice before you build it: which plant functions will pay for it, at what services rate, which vendor lines to back, how many projects until the practice earns a return, and whether plants will accept you as accountable after go-live.

Will industrial plants buy software by subscription?

Often for business software. For software that runs the plant, the accounting matters: a perpetual license, or a cloud deal that includes a software license, can be capitalized as an intangible asset and funded from the capital budget, while pure SaaS is a service contract, expensed as it is used, though its implementation costs can be capitalized and expensed over the term of the arrangement. Some buyers need a subscription restructured before finance will approve it. Test which form the signing function will approve before pricing and hiring around a subscription.

Why does horizontal technology struggle in industrial markets?

Because of how plants buy: a committee rather than a champion, capital and operating budgets approved separately, IT, OT and security review that can stop a deal late, and a route to market through distributors, integrators or OEMs. A Sprint tests which plant function buys, at what price and through which route, before the vertical team is hired.

Who sits on an industrial buying committee, and who signs?

More seats than most sales motions brief. Operations, IT and operational technology (OT), environment, health and safety (EHS), procurement, engineering, finance and a plant manager with a site-level veto can each stop a deal, and each wants its own evidence. The economic buyer is usually the plant manager or VP of operations, a different person from the champion, so the two need separate interviews. The committee can be mapped before the sales motion is built.

Why does the demo go well and the deal still not close?

Engineers engage generously with good technology, which shows the technology is credible and says nothing about the purchase. The vendor built its evidence for the person who liked the product, and people who were never in the room decide the purchase. In our experience, deals rarely die at evaluation. They die in procurement, when a specific budget line has to produce the money, or in integration, when the OT team is consulted for the first time.

How do I tell a polite buyer from a committed one?

A polite statement costs the buyer nothing and can never turn out to be wrong. A commercial signal is specific: a named problem ranked against other priorities, a number for what it costs today, a budget line, an honest account of who else must approve, and a next step the buyer proposes. Score each interview from one to five on urgency, priority, budget, authority and openness. A strong signal averages 4.0 or above, declared before the interviews begin.