Don’t fail to plan - Plan to fail (and fix it)

If you’ve ever worked in a hardware startup, you’ve probably seen this pattern:

- Engineering begins to execute a robust multi-phased development and validation plan.

- Before the development cycle is completed, the business needs to put products in paying customers’ hands to attract new investment.

- These early units are prone to failure due to incomplete validation

- Engineering is forced to pivot to field support to keep customers (and investors) happy

- Progress on formal validation slows to a crawl

- We spend the rest of our days playing whack a mole with issues on poorly documented prototypes

Isn’t there a better way??

Maybe we should stop planning for perfection (a complete, robust validation cycle), 

Engineering could collaborate with finance and marketing to create a development timeline that includes early customer deliveries.

Based on this timeline, we could prioritize critical testing before shipping anything.

We could preemptively create a service team whose sole focus is supporting our immature product in the field, allowing the development team to remain focused.

We could document our prototype designs as if they are in production (because they are!), ensuring our service team knows what they’re working on.

We could design our early prototypes with service in mind, designing in modularity, access to critical components, etc.

We would make sure we’re capturing as much data as possible from our early sales units - this is some of the best testing possible!

If you plan for failure, you just might succeed!

-Brian Schoolcraft

Selling prototypes - a necessary evil? 👿

If you’ve ever been in a hardware startup, you’ve felt this pressure. 

Investment has given you a runway, and you’ve got to start generating revenue (or attract more investment) before it runs out. One of the best ways to do this is by selling something.

In a perfect world, we’d work our way through each phase of the development cycle before we field our product. We’d have good confidence that it works, and high expectations for its reliability.

But that basically never happens right? As soon as the prototype works well enough to hand to a customer, it’s extremely hard to resist selling it.

Engineering pushes back, but most of the time the business team wins, and the prototype jumps straight to production. It’s in the hands of the customer now, and it’s everyone’s job to make the best of it.

It’d be easy to just sit back and say this is the wrong way to develop a product, but we have to face that it’s necessary sometimes. 

So what’s an engineering team to do? I’ll dive into my thoughts more next week, but I’d love to hear your experience in the meantime!

-Brian Schoolcraft

Old tech builds new tech

Your startup is doing something new, applying a new technology in a new way that’s different from everyone else.

This is your sole focus at first. We have to make this new thing work. All our effort and resources goes into becoming experts at the new thing. You can’t hire experts, because no one is an expert in this yet. You have to grow them from the inside.

And it works! You’re now the best at what you do, probably the only company doing what you’re doing. 

But a new technology is never completely new. It has a lot of new tech, sure, but I bet it still has a bunch of stuff people have been doing for decades.

Wiring harnesses. 

Bolted joints.

Gasketed seals.

Control software.

Sheet metal enclosures.

Injection molded parts.

You name it. No matter how new your idea is, it’s built up of dozens if not hundreds of existing technologies. 

Technologies that you can hire an expert for. Maybe you should!

-Brian Schoolcraft