By clicking “Accept”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Cookie Policy and Privacy Policy for more information.

From Industrial Data to Business Value - Why Data isn’t the New Oil

From Industrial Data to Business Value - Why Data isn’t the New Oil

Starting a data analytics initiative is unfortunately still a bet against the odds. Here you'll learn why and what you should invest in to succeed.

Few people realize how unlikely it is that their industrial analytics initiatives are going to deliver real bottom-line value. In this post, I’ll share some statistics you need to be aware of. I’ll then talk about some of the reasons why the odds are so poor. Finally, I’ll share my thoughts on the one capability you should invest in to succeed.


In our work in industrial IoT analytics “evangelist”, we meet a lot of very impressive companies across the Oil & Gas, Maritime and Energy ecosystems. They all have great people who are leading ambitious programs to digitalize and transform using data and analytics. We’ve been exposed to a few hundred different initiatives already.

However, hardly any of these initiatives have gone beyond the pilot stage. In fact, most of them didn’t survive beyond that cool PowerPoint presentation supported by a mock-up demonstration to the management team.

Don’t take our word for it. Ask around. Google it. The most frequently cited source is probably Gartner who predicts a mere 20% of data analytics initiatives will deliver business outcome through 2022.

If you’re more into studying past track record - researchers at the Oxford Saïd Business School found that any technology program with a decent budget was a wild bet against the odds. Two-thirds of all studied programs (roughly 1,500) had budget overruns. The average budget overrun on those projects was close to 80%. Even more disconcerting, hardly any of those projects delivered any measured business value.


We've played a variety of roles on data initiatives (and tech-enabled initiatives more broadly): Executive sponsor, project manager, external advisor, independent auditor to name a few. We’ve seen (and been responsible) for successes and failures. We've analyzed and discussed these at length with people way more experienced than us. All in all, it seems there are at least three behaviors making these things risky:

  1. Possibility trumps Utility or “Death by PowerPoint”: People start and stick to the wrong initiatives; Initiatives that look good on paper often should never have left that stage. Typically, the eager business developer or digital ambassador will choose initiatives that are possible to do (i.e., there’s good data and tooling) overestimating their utility (i.e., is there an important business problem to solve). Conditions will change over time, rendering an initiative more or less possible. However, the business problem stays the same. Initiatives must have utility.
  2. Novelty trumps Practicality or “Death by Prototype”: Investment decision makers tend to prefer novel solutions (e.g., modern user interfaces and applications using latest “machine learning / AI features”), while equipment operators (often the real users in the case of industrial applications) need practical solutions (i.e., simple, convenient and informative alerts). If you think about it, the last thing an equipment control room needs is yet another screen with graphs and lights. Even the best intended solutions and designs end up unused.
  3. Standardization trumps Speed or “Death by Plumbing”: You’ve got a great solution. You want to roll it out across the company. You quickly realize that it won't scale unless you connect to dozens of proprietary control systems, and hundreds of disparate data sources. Before you know it, the pace at which your business case will materialize depends on the pace of several large and never ending technology standardization programs (e.g., the “data lake / platform program”, the “IT modernization program”, the “Agile organization program”, to name a few favorites).


Let’s get one myth out of our way. Data is not the new oil. Oil has clear utility, is expensive to produce and is valued in relation to its scarcity. On the contrary, data is cheap to produce, limitless in volume and has no default value. As long as there are organizations treating their data as a scarce, expensive and invaluable asset, we can forget about the promises of a networked, collaborative digital ecosystem.

So how does data turn into value? It’s simple. By giving it purpose. Not structure. Not context. Not standard. Not tooling. But purpose.

Here’s an example to illustrate my point. If we show you two pictures:

Sack of beans
Wooden log

Sack of beansWooden log

A sack of beans and a wooden log. You’ll think “something to sit on”. You process information through a filter of utility. Can you think of any structure, standard, tool or context you can use to organize those two objects by their pure nature (organization of their real-life molecules, or indeed the pixels of the images) and come up with the same conclusion? Ping us if you do.

That’s barely half the truth, though. Data doesn’t really turn into value purely by giving it purpose. It turns into insight. In order to complete the path to value, you need people to turn insight into decisions and actions. For instance, when Arundo’s Pump Toolkit suggests you could run the pump 15% faster (and increase daily production) without incurring significant risk of problems, you need to somehow trust that suggestion and decide to turn up the pump RPM.

Where are we going with this? We're suggesting you stop thinking about data as a source of competitive advantage. Rather, your competitive advantage is the pace you’re turning new data enabled insight into operational change. Put differently, it’s the speed you’re able to find and scale possible, useful and practical solutions.

To achieve that speed, there’s probably one capability that’s more important than any other for you to invest in: Lower the barriers to experimentation. More specifically: Invest in the tooling, people and culture that will increase speed, lower costs and reduce the risk of experimentation. This allows you to:

  • Maintain and progress a larger portfolio of business-centric initiatives, quickly abandoning those turning out to be impossible or useless
  • Iterate very early versions of solutions with equipment operators (or whoever the user is) to get their feedback on how to make the solution practical and useful
  • Move fast with “non-standard” solutions until value is proven, and then selectively standardize (i.e., “plumbing on demand”)


Starting a data analytics initiative is unfortunately still a bet against the odds. Maintain a portfolio of business-centric initiatives and work through them fast. Invest in lowering the barriers to experimentation. Your future source of competitive advantage is not your data, it’s the pace you can turn your data into insight and operational change.