Skip to content

Insights

The mistakes we have already paid for.

What we have learned building connected hardware, ourselves and with clients, rather than from a content calendar. Written for founders deciding what to build next, and for investors deciding what a hardware plan is actually worth.

Start here

A Hardware Founder's Glossary for 2026

Seventy-eight terms a hardware founder meets in quotes, inspection reports, customs entries and compliance emails, defined plainly.

17 min read

Read the glossary

Three things founders get wrong

Funding and validation

Hardware is fundable again, but under different labels and against different evidence than in the last cycle. These pieces cover how the money actually moves: which demand signals a serious investor treats as real, why the deep tech rebrand changed who gets funded, and why a round sized to reach the next gate tends to stall at it. The recurring argument is that evidence and money have to be sequenced together, because a hardware programme cannot pause cheaply.

3 articles

Building and shipping

The work between a signed spec and a pallet of sellable units. These pieces cover the requirements document as the cheapest prototype you will ever build, what running a hardware programme involves once someone has to own the calendar, and where the difficulty in hardware moved now that prototyping is nearly free. The through line is irreversibility: hardware decisions harden long before anyone tells you they have.

4 articles

Compliance and economics

What it costs to be allowed to sell, and what it costs to keep selling. These pieces map EU and US certification for connected hardware with the 2026 deadlines that are already live, count the two thirds of the company that is not the device, and read the post-mortems of the hardware companies that died between 2023 and 2026. Most of those deaths were commitments hardened before the evidence arrived.

3 articles

Everything, newest first

Funding and validation

12 min read

Fund the Whole Phase, or Don't Start It

Why programmes that raise only to the next gate stall there, what a development phase costs, and how factories treat stop-start customers.

  • funding
  • hardware startups
  • npi
  • manufacturing

Compliance and economics

13 min read

The Device Is a Third of the Company

The app, the returns desk, the channel, the packaging register and the liability all outlive your launch. Budget for the two thirds nobody demos.

  • hardware startups
  • operations
  • go to market
  • unit economics

Funding and validation

12 min read

Ideas Are Cheap. Buyers Are Expensive.

How to test demand for a connected product in 2026, which signals count, and why a working prototype stopped being evidence.

  • hardware startups
  • validation
  • demand testing
  • pre orders

Building and shipping

14 min read

Someone Has to Own the Calendar

What running a hardware programme actually involves once the spec is signed: gates, cost walks, remote factories and a calendar you do not control.

  • programme management
  • hardware startups
  • manufacturing
  • npi

Building and shipping

12 min read

The Cheapest Prototype You Will Ever Build

Why the requirements doc is a hardware startup's cheapest prototype, and what regulation, security and sunset policy now force into its first draft.

  • product requirements
  • hardware startups
  • connected products
  • regulatory compliance

Building and shipping

14 min read

The Hard Part Moved

Prototyping a connected product is nearly free in 2026. Tariffs, memory allocation and working capital are where hardware breaks now.

  • hardware startups
  • manufacturing
  • supply chain
  • trade policy

Get started

Tell us what you are building.

A short conversation is usually enough to tell whether we are the right team for it. We will say so if we are not.

Start a project