IoT is a business problem first

If we follow the Gartner hype cycle from July 2016, we can see that ‘IoT Platforms’ is still not quite at its peak. Therefore, I expect to see more platforms in the coming months.

Since there are many players in the IoT ecosystem, I want to clarify the definitions used in this article. I’m using vendor for companies who provide the ingredient components to OEMs (Original Equipment Manufacturer). OEMs create the device, products and services. Customers are the consumers of the IoT products.

IoT Vendor Approaches

The IoT platforms I mentioned earlier try to solve various issues, from making hardware devices easy to create to moving data easily to the cloud. If we take five broad, technical areas (namely hardware, firmware, networking stacks, cloud and machine learning) where someone could create value, then we see two approaches emerging:

1) Focused (Specialised) Approach: In such a case, a vendor tries to create the best solution in one technical area. For example, a hardware vendor would create the best price/performance hardware chip possible. They would then partner with others to ensure that the chip supports the best RTOS/OS, networking stacks and cloud APIs. Sometimes the vendor goes a step further and creates an ecosystem by supporting many partners in each of the above technical areas.

2) Platform Approach: Vendors using this approach try to bring together and supply hardware, firmware, networking stacks, cloud and machine learning libraries – all in one solution. They believe that by integrating these technology areas together, they are creating a more optimal solution which is usually better than an OEM picking their own options. This approach has been gaining momentum lately. 

While the above approaches create value in the marketplace, when we surveyed OEMs to find their biggest problems in IoT, the two things that stood out were:

1) Business models: In essence, how does an OEM make money in IoT? At the end of the day, companies have to increase revenue, decrease cost, or reduce risk. Without that, we have solutions looking for a problem. We will cover this topic in more detail next.

2) Security:  Since an IoT product, by definition, is connected to the Internet, security is a big concern. It opens up the product to many more vectors of attack. Historically, an OEM didn’t have to deal with this and therefore doesn’t have the knowledge or expertise with this level of security. We will cover this topic in a separate post. 

Business Model

Customer Expectations Have Changed

Creating a product and following the basics of sales, marketing and distribution made sense in the past.  An OEM created something and pushed it down an existing channel. In the brave new IoT world though, we are seeing IoT products from incumbent OEMs struggle to gain traction. For example, with the existence of incumbent, market leaders like Honeywell, Trane and Carrier (who know more about thermostats, heating and cooling than almost anyone else out there), a company like Nest should never have become so prominent. Even if they did, it should have taken a long time to do so. However, in the world of IoT, Nest was founded in 2010 and got bought in 2014 for $3.2bn. Meanwhile, the incumbents’ connected home automation product didn’t generate as much buzz (and I suspect, not as much sales either). The incumbents continued following their existing channels, whereas Nest re-envisioned the thermostat, the sales channel and their relationship with their end customers. With this they established immense loyalty from their customer base.

Today, more than ever, we are seeing IoT customers (consumer / user of IoT devices, products and services) wanting a closer and more direct relationship with the IoT-enabled OEM. They expect the OEM to:

1. Sell them the product more directly (less intermediaries)
2. Service/Maintain the product for the life of the product
3. Provide on-going operational support with things like Analytics, Control & Monitoring, Upgrades, Diagnostics
4. Help shift costs from a CAPEX to OPEX model.

Data/Insights is the New Currency in IoT

In the previous example, the data Nest collected re-enforced and strengthened their algorithms. It was no longer just a product in the home. It was a system of products which spanned many homes. Learning’s (through machine learning) from a group of homes could be applied to other homes. With every improvement in insight, customers got more reliant and loyal to Nest.

We are seeing this network effects story play out in almost every IoT industry – Autos, Industrial, Smart City, Smart Building, Energy, Healthcare and even Aerospace. However, we are also noticing that incumbent players in an industry are slow to embrace a “data” strategy for their company and products. They are usually too busy worrying about how to make their existing products better, faster or cheaper (usually concentrating on cost avoidance and reduction in OPEX). However, if they embraced “data”, they would unlock tremendous value, and get a competitive advantage like never before. They would be able to increase their ASP (average selling price) and increase the customer base (revenue enhancement).

Minimal Executive Support

Executives hear the prediction of IoT devices that will be online by 2020, which ranges from 20 billion to 200 billion. The numbers make them realise the strong potential of IoT for their business. They believe that they need to play in IoT and therefore allocate funds to do so. However, the process that they follow is business as usual.

Executives throw the problem at their engineering and innovation teams. They expect a magical product to come out on the other side. While such a product may emerge (usually it wouldn’t due to other reasons covered below), the business model to go along with that product certainly will not emerge from that team. And since this team has no ability and mandate to create new business models, the rest of the organisation would force this team to conform to the existing business models. Thus the end result is a failed programme.

Even the most well intentioned teams get into two roadblocks – the review/refine cycle with executive management; and the selling/manufacturing cycle (sales wouldn’t sell something that isn’t real or is relatively new, and manufacturing wouldn’t produce something that is not selling). Meanwhile, even if it gets produced, the product sees minimal impact because the customer expectations have changed.

Value Proposition of Connectivity

While Engineering is busy creating and enabling connectivity, no-one is creating the new business model and the new value proposition associated with the new, connected products. Even after the company emerges from the review/refine cycle, the customer doesn’t see a value of a connected product and therefore refuses to pay for it.

The sales struggle ensues at all levels of the customer organisation from IT (who doesn’t want the headache of another device connected to their network), to operations (who may be happy with the current solution and doesn’t see value with connected products), to finance (is the extra cost of connectivity worth it? Is there a positive ROI?). And since no-one is in charge of the business model, no-one has thought through how to overcome these objections. And since the OEMs’ sales group is worried about meeting their quarterly results, they often ignore the new product/system until there is a clear path to winning deals.

Fundamental question like “Why does the product need connectivity?,” “What benefits gets enabled with a connected product?” and “What is the ROI of connected products?” need to be answered by the OEM. Only after that, can they hope to convince their customer to accept and buy the new, connected products.

Finding the Right Talent

OEMs in the past have gotten away with creating products themselves. In fact, many of the IoT platforms created are domain neutral. The OEM is expected to use the IoT platform and layer on their domain knowledge on it. This usually means a brand new engineering cycle, which requires new skillsets. However, an Intel Labs survey, found companies are struggling to bring on board the right coding talent.

Let’s consider the problem across the five technical domains we discussed earlier. We will add one more dimension for business. As you can see in the chart below, each dimension has a lot of options (the list is not exhaustive. Not all options are shown in the picture). The options have exponentially increased over the last few years.

So if you are an executive who launched an IoT program, should you rely on one of your existing engineering leads or should you go hire someone? If you go hire someone, how do you even write the job requirement? Should you pick a generalist or someone with in depth knowledge? The same Intel Labs survey found that an OEM needed a team with breadth and depth at the same time. As you can imagine, finding such an individual (or team) is very difficult.

IoT Design Center

Wind River noticed that many of the OEMs didn’t have the appropriate IoT business model for success. It resulted in 'we made this system, but we are seeing limited success,' even though other competitors may be succeeding. In some cases, the OEM followed existing channels to market, which didn’t work for their new IoT product. Lastly, we noticed that finding the depth & breadth of talent, at the same time, was very difficult.

To overcome these problems, the Wind River IoT Design Center was formed with a mission to 'Help customers harness the potential of IoT through connected, secure, turn-key solutions.' Through this centre of excellence we provide IoT consulting services around:

1. Business models
2. Product strategy
3. Technology strategy
4. OT/IT  system integration
5. Project delivery
6. Cloud solutions, which includes UI/UX, Helix Device Cloud and device management
7. RTOS, Linux Operating systems
8. Machine learning
9. Security

For more information on Wind River’s IoT services, visit: http://windriver.com/services/.

Courtesy of Wind River.