2-Day Build (Part 3: Prototype vs Product)

December 27, 2017
Written By:
Evan Reese

In Part 1 of this series, we built a working prototype of our IoT Plant Sensor concept in 2 days at the Nashville Maker Faire. In Part 2, we took it to the next level of polish. So if we can do that in about 2 days, why aren’t we ramping up production and auditioning for Shark Tank? Well, building a business around a product is a lot different than just making cool stuff. Here’s our list of what we’d do next if we chose to get serious about producing and selling our Plant Sensor. Hopefully this can highlight some of the big differences between making stuff for fun and making stuff for a living, and maybe even help some people make the transition.

Focus

Almost every married person that came by our Maker Faire booth suggested that their spouse could definitely use a sensor to keep them from killing the plants (cue vigorous winking and elbowing). In reality though, there are lots of markets that could use more data about plants; agriculture, home gardening, plant nurseries, hotels, groundskeeping companies, parks departments, and so on. We would probably choose only one of these use-cases to design for (at least to start out). That way we could really wow one group of people, instead of being maybe slightly noticed by all of them. It’s so tempting to try to design for every category at once, but just because they all have dirt and plants doesn’t mean they need the same thing.

Image for post

User Research

Once we pick an overarching use-case to target, we would want to make sure we could actually provide meaningful value there. So, if we decide to make a product for homeowners, not farmers, we’d talk to homeowners and test prototypes with them. In fact, we would have started doing this before even making the kind of prototype we made at the Maker Faire, and would continue doing it long after we’re shipping product. It would be our business to know our user’s problems, maybe even better than they do. In the end, if nobody wants it enough to pay for it, it’s just a hobby. It’s a solution in search of a problem.

Design a Business, Not Just a Product

For many companies, choosing the right business model is probably even more important than making a really nice product, though companies that last must have both. There are a number of interesting ways that a Plant Sensor business could create value. Let’s look at a few.

Data

When a sensor connects to the internet, a whole new universe of opportunity opens up. Our device’s primary function might be to let owners know how to better care for their plants via phone notifications, but that’s just the surface value. Picture this. Thousands of Plant Sensors all over the world are collecting data on all kinds of plants (and user behaviors). This one-of-a-kind dataset will eventually be big enough to be mined for useful insights about how people interact with their plants. From there, we could use machine learning to help users take better care of their plants by providing tips that vary by person and by plant. Armed with data, we could improve the sensor software over time, so the more people that have our product, the better it is for everyone else that has one. Almost every IoT product has a network effect like this.

Selling to People

If we wanted to try a B2C approach (Business to Consumer), we might try to connect our users into a social community to help them all improve their plant care skills together. They could share custom plant profiles, post stats, graphs, questions, or anything else they are interested in. In doing this, we would try to build a fanatical fan base for our business, which would give our company a great advantage over new competition. We could even explore a subscription model where we automatically ship people the right nutrients, seeds, or supplies at just the right time in the plant’s life cycle, which would allow us to sell more product, reduce customer buying friction to almost zero, and make our revenue predictable (recurring revenue is more valuable than the same revenue received a “lumpy”, unpredictable way.)

Selling to Companies

Another possible route would be to sell B2B (Business to Business) and serve customers like nurseries or small farms. To do this, we would explore a HaaS (Hardware as a Service) model, where we come and rig up all of their soil with tons of sensors, and charge a flat monthly fee for analytics. Our customers would borrow the devices from us, instead of owning them, and would pay monthly for the data we collect and parse. We might do the device install for free or even at a loss if the monthly revenue would break even within a reasonable amount of time, and if the LTV (lifetime customer value) was high enough.

The Point

We know that all of this is easier said than done, but these are the sorts of thoughts we’d be having. In all likelihood most of these ideas would flop for one reason or another, but hopefully we would be testing early, often, and cheaply so that we could evolve our way to a working business model with our early customers and try to scale that up.

The biggest thing to take note of here is that each of these business models needs a different product. We’d want to know which one we’re designing for from the start. We can always pivot later as we learn more. We would not just build something that we think is cool and then try to find other people who also think it’s cool to give us money for it.

A Note on Competition in Tech

To many of the people who stopped by our booth, our concept was new and exciting, but people have been observing plants since the beginning of humanity. We’re just doing that with a twist. We’re pretty sure this idea has been around in various forms for decades. In fact, after we chose a project direction, a quick search turned up the Edyn sensor, which is very similar to our concept (They’ve done a great job at it too!). We didn’t know of it ahead of time, but that’s how ideas like this work. It’s nothing to be surprised by. Google “smart (anything)” and it probably exists at least in conceptual form. After all, we bought an existing soil moisture sensor from Sparkfun.com, so it’s safe to assume there are quite a few people interested in this data. We wouldn’t let it scare us off, but we would want to know what we’re up against. Early Twitter and Uber investor, Chris Sacca, said, “ideas are cheap, Execution is everything”, which makes so much sense. Do you have a Palm Pilot or an iPhone? This is also why hardware startups–the ones that survive–have to keep innovating forever. Fitbit didn’t release their first device and kick back to watch the cash roll in. To survive, they have to defend themselves every day against lots of new companies with very similar ideas, and they can‘t stop. Ever.


Iterate. Iterate. Iterate.

Once product/market fit is at least partially validated, we would turn our attention to making the best product possible. Hopefully our research will turn up ways to make the product better. For example, a product like our Plant Sensor is just begging for solar power and waterproofing. Some other features we would explore adding would be ambient light sensing, detecting nitrogen content in the soil, detecting PH balance, responding to weather data, or even automated watering. No matter what we decide to make, we’re going to have to figure out how to make a lot of them. So far we’ve made a half-baked idea into a working thing. To succeed, we have to make a whole-baked idea into 1,000,000 or so working things.

Wrapping Up

Hopefully, you can see how much thought and research work was missing from our 2-day sprint. Our prototype could teach us a lot and be a great tool, but it’s really got a long way to go to be a viable product.