I have a friend who had a house built to his specifications. He bought the property, hired an architect and then made sure that he had highly experienced contractors and project managers working together to build the house. There were a few cost overruns and some surprises along the way, but the experience of the contractors and the architect paid off, and he is now very happy in his house. Best of all, the quality that went into the design and execution resulted in very few issues after he moved in.
When it comes to experience, embarking on digital transformation through the internet of things (IoT) is no different from building a house: It is complex to put together and significant know-how is needed. As the co-founder of an IoT software company, I know you can’t solve a lack of experience simply by throwing great technology at it. IoT requires so much orchestration of various elements that the potential surface area for failure is massive. In most cases, I’ve found it is not the technology that results in failure.
Among the biggest pain points resulting from lack of experience in IoT are:
• Missed launch windows: I’ve noticed that many companies miss their IoT launch deadlines — and they miss them by a lot. The biggest culprit to remedy is a lack of ability to appropriately scope the technical challenges and dependencies of an IoT project.
• Lacking in-house expertise: Companies that embark on IoT projects often find that they lack critical experience on their teams. If this is the case, you may have to hire several new employees and then plug gaps with contractors. It takes time for these groups to form up, get trained and begin working together as a team.
• Under-budgeted support costs: Companies that do IoT without the proper experience are always surprised by the ongoing costs they incur. Decide (and plan for) how you will maintain a high-availability service, keep mobile applications up to date and ward off security threats.
• Inexperienced project management: The groups working on an IoT project are often diverse. They may include, for example, third-party developers, manufacturing facilities located in various geographies and certification bodies in different regions. The less experienced the project management, the greater the chances for wasteful spending, so it’s important to choose a project manager who can balance these groups.
• Not being flexible enough: Research and development teams are good at building to a specification, but they aren’t always so good at guessing what customers will ask for next. Especially in IoT, it is critical to be “future-proofed” and to move fast with changing market and business needs.
• It all just breaks: Just as you need a strong foundation for a house, you need the same for IoT. Success will mean billions and possibly trillions of bits of critical data being processed to power your applications — with resulting risk surrounding platform availability, the products or assets connecting consistently to the cloud or security breaches. Using prefabricated elements that have already been tested and are functioning at scale is one way to reduce this risk.
Based on these items above, it doesn’t always make sense to build an IoT platform in-house.
Still, there are cases when it does make sense to build. So when should a company “build its own house?” And how can a company avoid the pitfalls above?
Like most business investment decisions, the decision to build should be driven by the ability to have a strategic advantage as a result of building in-house. A few key reasons for doing an IoT build are listed below. In the case of a build decision, the best way to improve your chances of success is to focus on the common theme of “experience” above: If it makes sense to build a platform, then it’s important to have a highly experienced team on your side to help you make it happen. Sometimes this can be drawn from your own staff, but often it means recruiting externally or bringing on a system integrator that has proven their capabilities in similar past projects.
Here are a few key reasons why it might make sense to build a platform in-house:
• Unique “real-time” performance requirements: The requirements for an airplane wing, as one example, are much different from those of a typical smart-home product. Certain markets require significant capability at the edge and demand real-time processing of massive amounts of data. Typical IoT platforms may not be built to meet the specific requirements, which could render them less useful.
• Specialized data sets: IoT platforms can provide significant flexibility but may not support highly unique data sets or be able to process the data sets as efficiently as a purpose-built capability. If a large part of the two-way data consumption is different from what most other companies need, then the “buy option” of IoT platforms is less likely to effectively support your needs … and if built out, the result may be a significant competitive advantage over companies that use “off-the-shelf” platforms.
• Special certification needs: Platforms are typically built to meet broad certification requirements for data privacy, security, and other ISO standards in vertical markets. But there may be times when special certifications are required that are more difficult to get from the standard platforms. That is most likely the case in niche markets such as military and aerospace in which it doesn’t pay for the platforms to go through the cost of the certification processes.
Building IoT platforms is complex. Because the concepts and goals for broad digital transformation are still new and emerging, the platform leverage via a build or buy decision needs to be very configurable and able to meet changing market and company needs. Regardless of which choice a company makes, it is imperative to review the skill set of the team involved and technology they’re deploying and make sure it is the right decision in the near term and scaling into the future.