To Build or to Buy: What to Consider When Implementing IoT

A key decision that will impact the overall success or failure of the implementation is whether to build a system from scratch in-house, seek a customized solution from third-party experts—or create some hybrid of the two.

Mnet 102666 Io T 0
Tor Jakob RamsøyTor Jakob Ramsøy

An estimated 60 percent of global manufacturers used analytics, data and connected equipment to optimize productivity in 2017, according to IDC. In addition, technology company Ericsson predicts that this year, IoT-enabled equipment and products will outnumber mobile devices. The Industrial Internet of Things (IIoT) is currently vaulting across industries. There is no turning back.

When I talk to CEOs of industrial companies, however, I find enthusiasm, but also a lot of hesitation. Something is stalling IoT efforts, even when leaders understand the importance of IoT. What’s keeping leaders from jumping on board?

One answer can be found in the complexity of connecting the hodgepodge of legacy IT systems, the intricacy of legacy physical assets and the convergence of info tech (IT) and operational tech (OT). A second answer is that few industrials possess the in-house knowledge to identify their IoT needs, much less the resources to develop and implement them.

A key decision that will impact the overall success or failure of the implementation is whether to build a system from scratch in-house, seek a customized solution from third-party experts—or create some hybrid of the two.

Why Build, Why Buy?

The primary benefit of building a system completely in-house seems obvious: you’re able to create a system specific to your needs and business. Unfortunately, a completely in-house approach may prove frustrating (and costly) for a number of reasons. First and foremost, most industrial companies lack the specialized software expertise and experienced data science talent to design and build the right system for their business needs. So, your in-house development would likely need to begin with an extensive and expensive human resource search, potentially setting any project back for many months.

Second, in an attempt to leverage and extend existing IT investments, you may fail to develop the functionality you truly need. Clouded by a desire to economize, an in-house IoT development process often fails to accurately identify important functions that users will require for effective implementation. It’s critical that any IoT implementation remains compatible with existing systems, while acknowledging any and all necessary changes required to create the most effective solution.

Third (and perhaps most important), very few businesses possess the expertise in-house to know IoT capabilities for their business, how IoT is being utilized in similar industries and how it can integrate seamlessly into their operations. The education alone may be required to even begin identifying IoT needs, and can be a major stumbling block—and a reason so many companies delay implementing a solution.

Solving for Build vs. Buy

As they weigh how to proceed with an IoT implementation, industrial executives should first do a deep dive on the available IoT capabilities in their industry, develop an honest assessment of their own current systems and conduct an extensive evaluation of their internal resources and expertise. A few guidelines to follow (and questions to ask) include:

  • Identify the highest impact: Where will IoT initiatives create the greatest effect on your business? What use cases are most critical to implement?
  • Address feasibility: What data will you need to implement these initiatives? What IT infrastructure and other systems will you need?
  • Ascertain capabilities: Are your current systems capable of managing these use cases? What additional technology will you require?
  • Categorize “low-hanging fruit”: Which use cases have the narrowest gaps between impact, feasibility and capabilities?
  • Assess: Can your business realistically execute on these use case implementations? Would a third-party provider be the most cost-efficient and effective solution?

With your prioritized use cases identified, you should be able to determine if your in-house capabilities can manage implementation. Most likely, you will be lacking resources in one or more areas. In such cases, partners with appropriate experience and capabilities may be able to develop a plan to quickly execute on your highest-priority use cases. From there, you can learn what worked, what didn’t work and what needs improvement to move forward on subsequent use cases.

The Bottom Line
IoT implementation is happening all around you. Businesses are scrapping homegrown systems and patchwork solutions for end-to-end point solutions and broader enablement capabilities. The intensified adoption of cloud technologies, the availability of more proven solutions and the overall market pressure to invest in this trend is increasing. One key learning is that an early business case win is a key factor for successful IoT implementation. If you learn to use the market to achieve this, the winnings will be clear.

Tor Jakob Ramsøy is Founder & CEO of Arundo Analytics.

More in Best Practices