A few months ago, I shared that I’d be working on an early warning water quality IoT device (image above). It’s been an enlightening few months and I thought I’d share some learnings and updates on the project. I’m sharing as much for myself documenting this process and also to share my mistakes, as lessons for the thousands/millions of people working on hardware products.
The summary of the four lessons
- Does the data support your gut reasoning for building the product?
- …know that the process is messy.
- First figure out the function before the form and
- ...be comfortable making decisions with incomplete information.
It’s a humbling experience and, Quoting Austin Kleon in Show Your Work ‘In all kinds of work, there is a distinction between the painter’s process, and the products of the process’. Here are some lessons from the process.
1. Does The Data Support Your Gut?
We decided to work on this project based on the amazing response to an article on empathetic IoT products. But the decision to continue is based on the sad data on water contamination from North Carolina in May 2015, Silverton Colorado in 2015, West Virginia in 2014, Chicago, Hinkley (of Erin Bronkovich fame) is still worrisome and 4 other states. And we haven’t even talked about Flint. Having a problem to solve is serious motivation and the data proves this is a serious problem. But be sure that you’re not just using data to justify a decision you’ve already made to build the product.
“In order to get to new solutions, you have to get to know different people, different people, different scenarios, different places.” Emi Kolawole, Editor-In-Residence, Stanford University d.school
2. …this hardware development process is messy.
Despite this being an integration play — we are building a water quality IoT device using off-the-shelf sensors — but we are finding that there might be the need for some invention here. The lead sensor has proven difficult to source. We’ve ordered products, that never got delivered, off Amazon. We’ve searched for products off Alibaba and (I kid you not) I believe, like the real-time demand driven business model of combatant gentlemen, the bots on Alibaba generate product splash pages based on search terms for products that do not exist.
Getting an ugly but functional MVP out led to donations from friends and family, conversations with experts in the water and IoT space and with heads of non-profits. Regardless of how ugly your MVP looks, it takes people away from dismissing your idea. Because it is no longer just an idea. The MVP makes your conversations real due to the form you’ve given the product.
“We want to give ourselves the permission to explore lots of different possibilities so that the right answer can reveal itself” Patrice Martin, Co-Lead and Creative Director, IDEO.org
3. First focus on function over form…
..but give consideration to the form. We’ve built and rebuilt several iterations of the Minimum Viable Product/MVP (see below) trying out different sensors and calibrating/recalibrating to ensure the product detects and communicates. Despite the missteps and progress, we’ve never lost sight of the fact that our final product has to ease data collection and early detection of the top 4 causes of water quality issues with the technology we now have at our disposal. Regardless of how ugly the MVP looked it had to serve the core need; detect water quality issues earlier and more quickly than traditional water quality detection methods.
4. Be comfortable making decisions with incomplete data
Once we had the function down, excluding the lead quality measures we have not been able to find or develop, we had to start thinking about the form. We could not continue fiddling with sensors even though the device wasn’t measuring our initial list of quality measures (Turbidity, Lead, Total Dissolved Content/TDC, Temperature). The black box is the partially functional prototype of the ugly but awesome MVP. We’ve compacted the MVP in a container to enable testing. The package, while black, also starts to move folk away from questions around the quality of the product itself. The irony hasn’t been lost on me that a quality measurement device should look like a quality device. With any startup product decision making speed is one of a few things that you actually control, don’t waste that advantage.
So where are we? It’s been great working with Waz (he’s the Wozniak on this project) and Dan (who owns and runs Axis Design and Forge Proto). We’re learning a lot about ourselves and the process. It’s honing my IoT product management chops and knowledge of the water utility space. We all have the work we do to keep the lights on (I have two clients with Asha Labs and 3 with Harper Jacobs) and we’ve come to a decision making point that you can have some input into; should we attempt a crowdfunding campaign or take money from a friend who’s offering to invest for a return like a traditional VC? Or should we hand the project off to someone else? We want this product to exist! Under a month ago I read this article stating that ~13% of kids in a particular zip code in Fresno California had 3X the lead levels of Flint! The problem isn’t going anywhere and someone has to solve it…
And all this came from writing a blog post about empathetic IoT. It’s amazing how, whenever I’m reading Dr Seuss’ ‘Oh The Places You’ll Go’ to my son these days, it takes on a certain level of heft for me..
Please like, tweet, share and heart the article if you enjoyed reading it. Sign up for my Polymathic Monthly Newsletter here, you’ll love it. Also, check out HarperJacobs for compelling content creation & Asha Labs for innovation/strategy consulting.