So far in this blog I have touched on many areas relating to the technology and leadership challenges of working in high-speed tech environments. While these musings were shaped by my experiences, it is perhaps time to look closer to what I have been doing in my day job.
I lead the product development at Drayton, a Schneider Electric company focused on building temperature control products for homes. Until recently, our products consisted of thermostats, boiler controls, and valves used by anyone who has a central heating system. For many years this was a stable, comfortable industry that we understood well. However, the apple cart was well and truly up-ended by new entrants exploiting technologies derived from smartphones. While this could be seen as an existential threat of the ilk that Nokia and Kodak faced, this upheaval also represented a great opportunity. The value we could bring to end users, and hence the area we could stake out in the value chain was now significantly greater than ever before. This gave our products the potential to be relevant to a much greater range of customers – from utilities to insurance companies. Just as importantly, we now had the opportunity to be much more interesting to our users. What was clear however, is that staying still was not an option.
Staying Still is Not an Option
This is a brief story of some of the lessons we learnt along the way. Earlier in the autumn, we launched Wiser Heat, our second generation connected product. Wiser Heat gives users control of the heating of their home on a room-by-room basis, and integrates with many Smart Home platform. It represents the direction we believe home heating should take and as such is our own little version of what the future should look like.
Jump to content Introducing Wiser, the beautifully simple heating system from Drayton. Easy to install from app to thermostat, Wiser is an altogether better, smarter, easier way for you to control the heating in your home. Wherever you are, whenever you need it. With Wiser, you are always in control.
Lesson 1 – It’s not easy
No part of the business remains untouched
Making the jump to becoming a connected products company is not an easy undertaking. It impacts not only the product development arm of the business, but all facets of the organisation. You now have a much broader set of technologies to master. You also need to get familiar with a different supply chain, develop new customers and channels, overhaul its service operations business, and develop new business models.
There is no part of the business that remains untouched or unaffected. There is no point learning how to create a new generation of products, if you cannot market, sell, source, support and service them. This is where many start-ups go wrong – creating exciting new products but having no means of getting them to market. This is also the reason why many companies embarking on digital transformation projects establish them as independent organisations at arms’ length from the existing operation. The culture and processes required can be radically different to what is already in place.
New skills will be required. Indeed, in our experience, the limiting factor that determined the speed at which the business could be changed was the ability to absorb and internalise new skills and competencies. For example, creating Smart Home services such as integration with Amazon’s Alexa voice control system and IFTTT (If This Then That)’s automation platform was done with a newly created cloud services team.
The flip side to all this, is that most companies that undertake this venture will fail. They will be lured by the siren call of the untold value of the data they will produce, and the clamour of investors to play in the IoT space. In doing so, they risk biting more than they can chew. The economics and complexity of IoT development means that fewer will succeed.
Lesson 2 – You are no longer a product organisation
The most fundamental change when producing consumer IoT products, is that you no longer remain just a product company, you become a supplier of services. This means becoming a 24-7 operation. Drayton had built its business through its reputation with installers of heating equipment, who in turn bought products through specialist merchants and distributors. This was very much an ‘office hours’ business. Although manufacturing has long operated around the clock, all other engagements with customers were carried out in office hours only.
Now we needed to support and maintain our cloud-hosted connectivity on a 24-7 basis, paying particular attention to early mornings and late evenings, when homes are cold, users are awake, and interaction with heating controls is at its peak. Companies that are starting out on the connectivity journey will struggle to operate their systems around the clock fully in-house. For this reason, it is essential to design a system that can be easily monitored and supported around the clock by specialist companies. Similarly, as many end users are now installing or setting up their systems themselves, they now need assistance late into the evening and into the weekend.
Lesson 3 – Your current business model will not survive the journey
Becoming a services company means that the customer journey no longer ends at purchase, but becomes an ongoing relationship that lasts as long as the product is used. This is a big commitment. You not only need to commit to supporting a product for many years, but also to persuade would-be customers that the service will still be supported many years from now. For services such as heating, lighting, security, this is a tough hurdle to overcome, and one for which brand credibility and trust is essential. When Google closed down the Revolv platform following its acquisition, the products were disabled. This is not an option available to companies evolving their core product to the connected space.
There are two main implications on the resulting business model. On one hand, the cost of producing the product does not end when the physical product is shipped. Instead it needs to be supported, maintained and updates for as long as the last customer is using the product. However, long term success depends upon realising some of the value that connectivity brings. While physical IoT products have an edge over pure-online products in the sense that customers expect to pay for physical goods, the long-term viability of connected products and the companies that produce them lies is in capturing the value of connectivity brings. There are a multitude of ways in which business models can be adapted, which I will not delve into. Options include using data insights to lower support and maintenance costs, using such data as the basis for creating long-term partnerships, using the new-found relationship with end customers to sell related products, or creating new services that generate a revenue stream either from end users or third parties. Either way, don’t expect your business models to remain unaffected.
Lesson 4 – Don’t try to build everything yourself
There is simply too much to master
When developing the Wiser Heat concept, it became obvious early on that this was a massive undertaking. The range of technologies involved was far greater than anything Drayton had worked with before. Although most of these were not new to Schneider Electric, they had never been brought together to create smart heating products for the European market.
The key proposition of Wiser Heat was to be the best multi-zone smart heating product for the mass market. This meant doing two things –make smart heating accessible to all, while allowing users to control the heating of each room individually in the easiest way possible. This had large implications in terms of the technologies to be mastered.
Each room was controlled by battery-powered Smart Radiator valves which needed to operate over a low power home network to a hub. In turn, this connected to the Internet and the Wiser Heat cloud platform via the home WiFi network. All this had to be secured against malicious or unauthorised access locally or over the Internet. Additionally, while Drayton has decades of experience in temperature control, managing temperatures of each room in a house individually required the development of sophisticated learning algorithms which modelled the thermal characteristics of the home and took into account the forecast weather.
The development complexity challenge was addressed in a number of ways. First, wherever possible, we used ready-made components, from using Microsoft’s cloud-based IoT Hub platform to standards-compliant Zigbee modules. The use of standards-based solutions also greatly reduced the cost of development, as well as facilitating integration into third-party products and systems. Additionally, wherever we needed to up our game in an area of expertise, we sought experts to partner with. This ensured that not only were we adopting best practice, but it also allowed our teams to develop and learn from the best people, normally from outsiders to our industry.
Lesson 5 – Don’t forget where your core expertise lies
This is the reason you exist in the first place
While we were keen to outsource where we could, this was, and remains a fine balancing act. Fully outsourcing software means that you lose control of the overall architecture of your product – which is essential for managing its quality and planning its evolution. In our experience, software outsourcing worked well where the problem was well-defined and constrained by clear interfaces. In an evolving IoT environment, where physical devices, apps and cloud-hosted services were being developed simultaneously, this was hard to achieve.
The risk of working with new technologies is that it is easy to lose sight of why customers come to your company in the first place – your expertise in your line of products. As such, it remains clear within Drayton that while we work with and recruit experts from across many industries, we remain first and foremost a temperature controls company. Our customer, market and product insights, accumulated over decades is what defines us as a company, and is essential that this is not lost as we broaden our technical know-how.
Lesson 6 – Agile is not just for software
Hardware. Why it is called the Internet of ‘Things’
You would be hard-pressed to find a software development organisation today that did not subscribe to at least some of the principles and practices of agile development. There are many good reasons for this that we needn’t go into here. However, developing Internet of Things products does bring one complication – namely the fact that you are building ‘Things’ and not pure software.
Hardware development differs from software development in two key areas. First, defects cannot easily be fixed through an update. At worst case, it will require a product recall. This places a very strong onus on robust quality practices. Secondly, the manufacture of hardware entails lead times in supply chain and manufacturing that can take several weeks or months. Creating IoT products is therefore a balancing act of bringing together the very different worlds of software and hardware development.
Through trial and error and after many false starts, we adapted agile approaches into an approach that worked, at least for us.
First, we grouped the hardware and software engineers developing a single product together into the same scrum team. We followed the principle of grouping people into teams that best reflects where most communication needs to take place. This meant creating cross-functional teams of electronics, software and mechanical design experts. This was a founded on a realisation that most challenging engineering problems were best solved by bringing expertise from many backgrounds.
Secondly, despite the advent of rapid prototyping and 3D printing, a reality of much of mass-manufactured hardware is long lead times for the sourcing and production of parts and manufacturing tools. It was therefore essential to ensure that long-term milestones, more associated with waterfall project management were captured within the agile planning process. In software, to an extent, the order in which work is done can be changed around. There is howeverlittle such flexibility in producing physical goods. What we ended up with is a hybrid delivery methodology that allows us to marry the constraints and strengths of both the hardware and software worlds.
Lesson 7 – Iterate, iterate, iterate
The most important lesson of them all
The most important lesson I have learnt while on this journey is that you will never get anything right first time. Your assumptions on how customers may use a product, what user interface will be more popular, which technical solution will work best will be wrong more often than they will be right. It is therefore crucial to build in a certain degree of modesty in your planning, an assumption that it will take a certain amount of trial and error to get things right.
This is where the value and wisdom of techniques such as lean development really hold their own. By encouraging building and testing usable prototypes as early as possible, lean development is really the only way to create IoT products. The uncertainty that often accompanies the proposition, usability and development decisions can only be properly de-risked by building prototypes and testing them as early as possible.
Wiser Heat was not our first connected product. That honour lay with miGenie, a fully connected variant of our high-end home thermostat system. We deliberately did not attempt to develop what we thought might be an ‘end-game’ product, but take small, incremental steps, from which we can learn.
In this context, the definition of ‘testing’ is very broad. It covers anything required to reduce the risk of a decision needing taking and covers both user testing as well as technical testing. The example that best brought this to life was the design of the Smart Radiator valve. Launched this autumn, it featured a design that is distinctive and different from any other valve in the market. No complex buttons or displays – just a simple twist cap to operate. We were initially surprised to see how positively the product was received by users and reviewers, and how intuitive it was perceived to be. With hindsight, perhaps we should not have been surprised. This was by no means our first, or indeed, intended design. We had honed into our current design through several iterations, most of which involving user testing of some form or other.
Where to next
The journey we have been on very much illustrates the old adage that strategy is useless without good execution. To mis-quote Thomas Edison, success is ninety-nine percent perspiration and one percent inspiration. It is undeniable that IoT has been over-hyped, not least by the over-inflated numbers relating to connected devices. Nevertheless, for many businesses and industries, there is no turning back. Once users give their seal of approval to a new way of doing things, this expectation cannot be undone. The convenience of controlling your Home through smartphones is now well established. Voice control is now becoming more popular. We will continue to move forward, in small steps, following the lessons learnt thus far, and hopefully learning new ones on the way.