If you’re involved in market research in a technology company, or in a company that’s becoming more technology-focused (which is basically everyone), then you’re probably hearing more and more about the Internet of Things (IoT). In fact, an IoT project may have already come across your desk.

For a quick definition, the IoT refers to devices that send data over the Internet. This could be a consumer device such as a smartwatch that’s sending your heart rate and sleep cycle to the cloud, or a business device like a smart power meter that, every minute, sends data on your power usage.

The analyst firm Gartner, Inc. predicts that there will be 26 billion IoT devices by 2020.[1] Business Insider expects that, by 2019, it will be the largest device market in the world, “more than double the size of the smartphone, PC, tablet, connected car, and the wearable market combined.”[2] In short, there will be far, far more “Things” using the Internet than people.

In this article, we’ll walk through 10 “truths” to bring you up to speed on this megatrend so that you can lead more effective research in this rapidly evolving (dare we say, exploding) space.

#1 – “Old-School” Industries Will Matter

When people think of technical innovations, they often (rightly) assume that these will take hold first in Silicon Valley, but the truth is that many of the early adopters of IoT will be old-school industries such as manufacturing, natural resources and government. The reason is because these are capital-intensive segments in which efficiency is prized and they are ripe for “sensorification.” 

As an example, it’s expensive to inspect thousands of miles of oil pipelines. But pipelines can be instrumented with sensors that can instantly detect a pressure drop in one section, indicating the precise location of a leak. This allows faster response and limits the impact on local communities. As another example, a modern GE locomotive has “more than 6 miles of wiring and 250 sensors generating 9 million data points every hour to run as efficiently as possible.”[3]

And in government, the term to watch is “Smart City,” which involves monitoring of parking, roads and traffic, noise, crime, the environment and more. Already, cities are filling up with license plate readers.[4] Citizens and leaders will soon have more data than was previously imaginable.

So when you’re thinking about markets and segments, be prepared to look at industries that have historically been laggards in technology adoption. Don’t leave out manufacturing, mining, government, insurance, energy or agriculture.

#2 – All That Hardware Is Really About Software

While the nerve endings of the IoT will be hardware sensors, the real value is going to come from software. Each sensor will emit a steady drip of data. But, for the most part, sensors will be designed to be cheap, replaceable, and really “only so smart.” The drips of data they emit will aggregate into a raging torrent of information that must be stored and analyzed, sometimes in near real-time. That’s going to require staggering amounts of processing power and a lot of code. If you want to understand the use cases, you’ll want research to focus on the person cranking out that code (or the software architect they serve) more than on the person holding the soldering iron.

#3 – It Will Live in the Cloud

IoT will go hand in hand with cloud computing, Big Data, streaming analytics, machine learning and data visualization. Since the IoT sensors will be Internet-enabled, they’ll usually be sending data to, and receiving commands from, the cloud. Cloud vendors like Amazon Web Services, Microsoft Azure, IBM BlueMix and others will be huge beneficiaries of this trend. These services will let startups get off the ground without huge capital investments, let them “pay as they grow,” and let them hone their business models and algorithms.

#4 – There Will Be Three Enormous Markets

Much of the press regarding IoT focuses on wearable technologies, smart thermostats (such as Nest), home security, and (for reasons we don’t understand) smart refrigerators. But the consumer market will be dwarfed by business and government uses. Outside the old-school industries mentioned earlier, the possibilities for supply chain optimization, warehouse instrumentation, remote monitoring of in-home healthcare, energy-efficient buildings, smart retail, fleet management and many other enterprise use cases are enormous.

When constructing research, the place to start is to investigate where existing business processes could be greatly enhanced through instrumentation and intelligence. For example, how could a fleet of delivery vehicles provide intelligence and improve efficiency if they were “sensored up”? Then consider investigating completely new paradigms that will only be possible on an IoT substrate.

#5 – Look More at Landscapes than at Individual Players

Let’s face it: we’re in the “gold rush” days of the IoT. Players are going to rapidly come and go. Consolidation hasn’t even started. What’s most important at this stage is to understand the topology of the landscape. Think in terms of markets rather than specific players. The playing fields aren’t going to change as fast as the teams.

Much of the research opportunity, at this stage, will be focused on investigating entry into a market and thus a broader view will be critical.

#6 – Today’s Customers Are Not Tomorrow’s Customers

And today’s partners are not tomorrow’s partners. Case in point: John Deere has “APIs” in a program that software developers can join.[5] Let that sink in. That means that some of John Deere’s new customers are programmers. While we pine away for self-driving cars, self-driving farm equipment is already here. Drones will take over crop dusting.[6] Robots will replace security guards.[7] Do you think the corporate campus facilities department has ever dealt with a lot of robot vendors? Huge amounts of learning will be needed to figure out both the buyer and seller side as customers and vendors touch for the first time.

When constructing research studies, be prepared to throw out all assumptions of who customers, partners and vendors are. Historical norms will limit you greatly here.

#7 – Qualitative Research Will Be King

We are very much in the discovery phase of this trend. A quantitative survey instrument would be built almost entirely from guesswork, and respondents will likely not even understand the questions or interpret them the same way the researchers do. The high-value research will be qualitative. It’s going to come from in-depth conversations about use cases. What are people trying to do? How are they trying to do it? What have they learned along the way? What unexpected problems or benefits did they experience?

#8 – Partners Will Know the Most First

Customers aren’t that knowledgeable yet. Suppliers will hope to have guessed right when making products. 

Imagine you want to understand the market for the IoT for high-rise energy efficiency. You could conduct interviews with end customers who have had such systems installed. In your research, you could learn the triggers for making this investment, find out which vendors they considered and hear how the implementation went. That would all be great research, but each customer would only know their one and only case.

Now imagine including partners who had each performed a number of implementations in your research. They could go beyond the one customer experience and talk about what’s driving buyers in general, and what really resonates with them. They could describe how they’ve stitched products together from multiple suppliers to make a complete solution. They could talk about what’s lacking in products today and where they see opportunities. Because of their experience, they will have steered around pitfalls that the end customer was never aware existed.

In the early stages, customers are going to be looking for partners to be trusted advisors, and for a very good reason—the partners performing the implementation will actually be the most knowledgeable people in the value chain. 

#9 – This Stuff (Mostly) Isn’t Going to Be Secure

The market will be dominated by 900-pound gorillas like Cisco, Microsoft, Intel and GE, but also by startups that didn’t exist three years ago. The big vendors will do a good job with security.  Cisco has been building secure network devices, since, well, before there was an Internet. But the startups will be where much of the compelling innovation happens. And building something secure is much more expensive than just building something that works. Most of these startups will be deliriously happy to simply get the kind of market share that makes their device worthy of cyberattack.

As a result, the security aspect should be a component of many research projects. How much do buyers in a sector really care about security? And if they say they care, does that concern actually seem to carry over to purchasing behavior? These are all good questions.

#10 – Look for Quantitative Data that’s a Real Leading Indicator

There are times when you can run a quantitative survey and there are times when you can just ask the Internet. For example, consider the following graph:

This data comes from Indeed Trends, which will simply graph how frequently a term shows up in all the job postings that they index (millions). We’ve found that job postings are a good leading indicator for a trend, especially a technical trend. Here’s another—want to know which large companies are making the biggest IoT investments? Just ask LinkedIn:

This kind of data abounds, can be very cost effective (read: free), and provide substantial early insight. It doesn’t replace quantitative research, but it can augment quantity and quality in powerful ways.


If, in 1995, you had said that the Internet was going to disrupt most industries, many would have scoffed. In fact, many did. But, looking back, what do we see that remained untouched by the Internet? Virtually nothing. Blockbuster was blown away by Netflix. Kodak lost to the digital camera and instant photo sharing. Encyclopedia Britannica fell to Wikipedia. Amazon crushed bookstores, first, and then others in retail. Uber and Lyft are taking on taxis. How did you buy your last plane ticket? Your last insurance policy?

The IoT will cause a similar reconfiguration. The spoils will go to those who figure it out first. And it’s the trend that’s launching a thousand research projects. Hopefully, this article has helped put a few more things on the checklist for your next IoT research project.