Three Reasons 60 Percent of IOT Projects Stall at Proof of Concept

car stalled
I received some interesting notes back regarding the post (here) from SCPD, particularly on the 60 percent of IOT programs that don’t make it past proof of concept.  For those of you coming from outside the tech world, IOT refers to the “Internet of Things.”

I wasn’t surprised, as the Gartner Hype Cycle for IOT has been consistently showing various elements of IOT in the “peak hype” portion of the curve for several strategy cycles.  For those unfamiliar with this curve, it builds on the common observation of new-to-the-world technology undergoing a massive boom of expectations, before sinking into a deep trough of disillusionment, and then finally emerging on a solid platform of costs and benefits.

There are deeper reasons why this technology has had such a stubborn battle with moving from expectation to delivering on its promise.  And these reasons all pivot around one truth: the end customers don’t want IOT solutions.

Let’s unpack why this is the case.

IOT has been around for a couple of decades now in various forms.  The current wave was put in much stronger motion coming out of the ‘08 economic crisis primarily due to semiconductor firms needing to generate demand for devices that could put their huge fixed investment in fabs back to work.  What better way to do this than to go to the “bottom of the pyramid” where the number of nodes needing connection was nearly infinite.  By picking up on some existing trends in products and manufacturing, they were able to develop some early successes (smart cities, connected vehicles, smart home, retail tags, etc). Great marketing was created that pointed us toward the “future.”

The truth, however, is that these use cases had been along well before IOT was a thing, and the most successful of them are not looked at as IOT programs at all.  Smart cities have their roots all the way back to the mid 70’s and very high quality work has been going on in the transportation industry for decades, as illustrated here and here.  The same can also be said for Smart Grids, which you can read more about here and here.

What is really going on here is much bigger.

First, it is absolutely true that we have made huge strides in technology.  Moore’s law, cloud storage, sensors and open source software and hardware have opened an enormous amount of potential for connectivity and data gathering.  What we are just beginning to get our arms around is that our current view of supporting devices is inadequate for this huge explosion.  We simply cannot treat billions of devices in the same way we think in traditional enterprise IT security (see article here).

Secondly,  in many cases we are taking devices and architectures off the shelf that were purpose built for the previous applications, and mapping them to this new world.

Every significant wave of new technology has gone through this cycle and there are always some very strange artifacts on the path from one paradigm to another.  For example, consider how persistent the physical keyboard was in the evolution of the smart phone.  We consistently demanded that each device be a mini typewriter until 2009, when the iPhone smashed through that wall with a touch screen that changed how we all use devices and ended our need to move a mechanical key.  Do you recall the “hybrid devices” that had both? 

Whenever new business models are born, there is a predictable path that occurs.  I’ve roughed out a six-step process that captures this pathway:

  1. A hypothesis of benefits for a select group of clients based on a “new” application
  2. Build out crude prototypes using off the shelf elements from the last generation of technology
  3. Furious experimentation to solidify the reality of the benefits established
  4. Reference architectures emerging that provide a reliable platform for those benefits
  5. Customized elements of those architectures being built to improve cost, performance and refine the experience
  6. A long and robust cycle of adoption and scale

In many ways, IOT as a category skipped the first three steps and proceeded to step 4, i.e reference architectures and solutions looking for a profitable application.  What this means, is that we have things that worked well in one application, being sold as general purpose items for the “IOT” industry.  In doing this, the success ratio of new projects drops (in the case to 6/10 not working), and the whole cycle needs to start back at stage one.

Which leads us to the third key issue, the programs that are “failing.”  When you are undergoing a shift like we have described above, the temptation for firms is to create a scenario, rooted in their current business paradigm, where each project or program results in a financially contributive line of business.  This results in programs that are wrongly structured.  When these shifts occur, the business case needs to be looked at as a hypothesis to be tested, and the primary outcomes being the learning and insight that can be gained.  The reason this is so extraordinarily hard, is that the value narrative of these core businesses are shifting.  Just as value migrated from AOL as Google became popular, value migrates during successful IOT projects as well.  Programs need to be set up to derisk and discover these powerful new value shifts.

I’ve written about the STRIDE framework here and here.   IOT projects find themselves in the “R” trough, where the real learning occurs that is going to become the core intellectual property of the offering.  Recognizing this, savvy leaders structure programs with risk identification and reduction as the primary outcome, rather than initial financial success. (financial success is a lagging indicator in these transitions)  In this way, you can make prudent investments in risk reduction and learning, guided by the longer term vision of the business case outcome.  I can’t overemphasize how important it is to get this thinking “right side up.”  When taken to extremes, upside down thinking here will lead to catastrophic damage to P&L’s and careers (see article here)

So, to my opening premise, the shocking truth is that your customers don’t want IOT. Rather, what they really want is your core products and services benefits served in the most efficient way, which is highly likely to use all the technology bits combined with hard won IP that your firm develops while doing the hard work of hypothesis testing.

I’m working with several clients through this process right now, and it leads to some much deeper insights and breakthroughs in two areas.  First, by starting with the “Why” and the hypothesis, “if we do this, then the end customer will receive X, X and X benefits,” it leads to much crisper and specific understanding of the real needs, areas of risk and clarity in specifications.  Second, this allows us to look at building blocks that are much better suited to what the need is from a cost and quality perspective.  It allows us to spend value on the right elements to support learning and discovery on the path to real benefits.

I’ve covered a lot of ground with you today.  If you’d like to talk more about this first principles approach it all starts with a 20-minute virtual cup of coffee.   To get started, give me a call at 847-651-1014, or click here to set up a no-strings-attached phone call.

Related posts you can benefit from…

Did you enjoy this blog post?
Sign up to get access to Scott's monthly innovation newsletter and blog post.