No Idea What to Build

Does anyone have any idea on what I could create that potentially could be marketable? I’m not looking for get rich quick, etc…just modest is fine. But something useful. But there is so much out there its overwhelming. Not sure what to start with.

Consider that for any hardware you put on the market you need to do CE/FCC/LoRaWAN certification testing which will set you back a fair amount. Are you able to spend $10k to get started? (You’ll need to spend that amount for every node you design.) Then you need funds to start a production run and spend money on marketing. And once you are selling you’ll need after sales support.

The other option is to use ready made hardware and create custom solutions for specific customer (groups). That requires you know what those customers need and are able to create an integrated solution for them. However I guess you wouldn’t be asking for suggestions if you were in this situation…

Basically you are at the point where many techies find themselves. We have solutions, now where do I find a customer to sell them to… (and that is basic mistake number one in commerce, you need to find a problem first and then sell the solutions)


Thank you for that detailed response. Much appreciated.

1 Like

Find a problem, obviously one that involves IoT, and find a solution.

Finding a problem involves asking around, but not like, “What IoT solution would solve a problem” but “Is there something you want to keep an eye on from afar”.

If you have no idea what to build then maybe you just shouldn’t…

1 Like

Hard to say.

Keep in mind Sturgeon's law - Wikipedia and don’t get discouraged by it, try more ideas! Also, keep in mind if you see someone else having the same idea, then it must be good :slight_smile:

I find that sometimes you can find a good synergy by combining two very different worlds. The chance is small that you’ll suddenly discover something that is an order of magnitude better than an existing thing. The best ideas IMO are the ones where you combine two things and think “why didn’t anyone think of this earlier?”


In the UK, that law is cited as “everything Westminster does is c7*p”.

Paging @Jeff-UK as I know he’ll appreciate the nuances in this humour.

What ever colour is in power! _ anyhow I thought it was Holyrood, Edinburgh these days…

…Jeeves, pass me another tin of caviar and the mini toasts please…!

Hi @Slashbrackets all of the comments are very valid.
Especially @kersing 's comments about certification.
Why don’t you start out with something that can be useful to yourself like e.g. a soil moisture meter for out in the garden that indicates when you should water the plants or grass? That way you experience yourself with many technical challenges that you can put to use when, and if, you ever stumble onto the fabulous product that everybody wants, but doesn’t exist yet.

What business do you want to be in?

Do you want to manufacture and market off-the-shelf sensors devices?

Do you want to customize firmware for existing devices, and maybe in a pinch customize the device hardware?

Do you want to develop a system and sell pay-to-use monitoring services to customers?

Very few organizations manage to be good at everything, especially when they’re imagining the need rather than responding to a real one.

To take an example, there are a number of companies that make low-cost LoRaWan sensor devices where the hardware itself is borderline usable (though would typically benefit from customer input to fix a few small unfortunate design choices).

The bigger problem is that they have the mistaken idea that what they are selling is software value. But their software is never right - effectively by definition it cannot be, because software can only have the opportunity to become right when there’s an iterative process of feedback corrections from trying to use it. Instead of doing a bad job of trying to provide software solutions, and then trying to lock their customers into that in the mistaken idea they’d jump to another hardware vendor, what they really should do is say:

  • Here’s the board we manufacture inexpensively
  • Here’s the github repo for a LoRaWan stack that works on it
  • Here’s a simple demo
  • You’re on your own.

Whenever they try to go beyond that, they inevitably fail.

The only way to actually succeed selling full solutions is if you really have engineering staff focusing on the actual problems of customers - and since that’s an ongoing expense that works out best when you sell services rather than try to price the support into the one-time cost of products.

1 Like