First Contact

Demand for measurement keeps compounding. Electrification, battery and EV validation, aerospace and defense test programs, grid modernization, and the steady creep of predictive maintenance into nearly every rotating asset all want the same three things: more channels, faster sampling, and numbers you can actually trust. The global market for data acquisition systems is tracking toward roughly $3.7 billion by 2030 at high-single-digit growth, with testing the dominant application and North America still the largest slice.¹ The complicating part is that the box itself is quietly commoditizing — a hard-won, low-noise 24-bit front end is increasingly judged not by its noise floor but by its software, its API, and how cleanly it hands data downstream.²

What owners are navigating right now

The pressure in this corner of instrumentation isn’t really about demand. It’s about where the work — and the margin — keeps migrating: away from the analog front end the founders built their name on, and toward the software and integration layer wrapped around it.

• A maker of precision acquisition modules for aerospace test labs has a healthy backlog and a noise floor competitors struggle to match. Lately, though, every RFQ leads with the analysis stack, the cloud handoff, and the API — and the 24-bit converter that took three years to perfect is treated as table stakes. The team is world-class at analog. The customer, increasingly, is buying software.

• A regional builder of data loggers and sensor front ends serves customers running mixed fleets — Modbus here, CANopen there, a decade of Profibus nobody intends to rip out.³ Buyers now expect edge filtering so they can ship insight instead of raw terabytes, and the founder keeps pulling analog engineers onto firmware and integration work, quietly wondering when he became a software company.

• A hardware-heavy manufacturer with meaningful offshore board and component sourcing is weighing 2025 tariff exposure against a genuinely hard call: reshore final assembly, absorb the margin hit, or pass price through to customers who are themselves watching every basis point.⁴

When the box gets soft

The question worth sitting with is who captures the value as instruments become software-defined — reconfigurable platforms that change function in firmware rather than fixed-purpose boxes. Hardware doesn’t stop mattering; a clean measurement is still a clean measurement, and no amount of software rescues a bad front end. But the recurring revenue, the switching costs, and the customer relationship increasingly live in the software and data layer, not the chassis. The consolidators have noticed: Emerson absorbed National Instruments for $8.2 billion and folded it into a dedicated Test & Measurement segment, betting squarely on the software-connected side of the category.⁵ What that leaves for focused independents is the open question — and, we’d argue, the more interesting one.

What Watershed is watching

We’re drawn to businesses that sit on the boundary between the physical and the digital, where decades of domain knowledge are hard to replicate and the measurement simply has to be right. A focused DAQ maker with real analog depth, a defensible niche, and a software story that’s earnest rather than aspirational is exactly the sort of thing we like to understand slowly and hold for a long time — no clock running. We’re learning here, not pitching, and we’re glad to compare notes with anyone else paying attention to this stretch of the trail.