This month, our CEO Pilgrim Beart gave a talk at Smart IoT London 2019 on his experiences building Hive, the growing pains in scaling IoT projects and best practices to solve them. Here is the talk, but in handy blog form.
In any IoT project, you'll find that your mindset completely changes once you hit a certain number of devices. You'll begin to care more about quarterly goals, processes and metrics, plus your customer support, product and organisisation - whereas previously it was all about proving "can we even do it?"
"Once you're in growth stage with 1000s of devices, your mindset will change... it will be all about process + measuring stuff. Your customer support must become more structured, and your key people skills will be more around product + organisational management." - @pilgrimbeart
— DevicePilot (@devicepilot) March 13, 2019
As you grow, you'll begin to hit certain pain points and have to reframe your thinking in order to solve them. Here, we list Pilgrim's 7 step plan to becoming a faster, sleeker connected product company to ensure IoT success:
1. Measure to improve
Metric |
At creation stage? |
And now at growth stage? |
| Speed | Sprints | User base growth |
| Iteration | Jira tickets | Resolve customer tickets |
| UX | Features and crashes | Engagement, churn and TTKD |
| Risks | Feasibility and "fit" | Profit and loss |
| Costs | R&D | Recurring, proportional to user base |
| Revenue | What revenue?! | $ per user |
2. Find your monthly recurring revenue
One of the joys of IoT is the realisation that you can turn any product into a service with its connected offering - for example, the way Hive has done from offering what was once a simple thermostat to a connected device which is paid for as an ongoing service.
Monthly recurring revenue (MRR) will become crucial for revenue sustainability but is only one way to get there. You could also consider multi-sided market and bundling.
3. Understand your customer
Did you know that IoT gives product managers superpowers?
Here, you can see the circular lifecycle of IoT product delivery. Being able to understand your customers better, constantly reviewing how they're using the product and if they're happy with it is imperative to offering a product that will have a far less reduced "shelf to kitchen drawer" life. Using the right tools, you can take the data coming from your device and turn it into customer insight.
4. Flatten costs with tools and automation
It's simple - connecting a product drives ongoing costs. But your biggest cost will be people, and doing things manually is slow, unreliable and expensive.
Rather than hiring more and more people to throw at issues, consider how you can better equip your current staff - are there tools that can help them do their job better and with more ease? Are there processes that could be automated?
Don't forget - R&D amortises with scale, but operations doesn't.
5. Eliminate these three bad practices
1. Being reactive instead of proactive
2. Allowing teams to remain disempowered
3. Doing things manually
6. Outsource non-core capability
Rather than building systems for every issue you come up against, see if the system has already been built by someone else. You should ideally be buying in everything except your unique selling point (USP) - your area of domain expertise - freeing you and your team's time to focus on your value proposition.
Spend more time specifying, selecting, testing and accepting, and if you've already built too much, plan to replace it.
7. Assign clear ownership for key metrics
There are three clear drivers of IoT happiness:
1. Your customers' experience
2 Your costs
3. How you're scaling
But you should always ensure it's clear who is responsible for these three things:
