Data and analytics is delivering on its promise. Every day, it helps countless organisations do everything from measure their ESG impact to create new streams of revenue, and consequently, companies without strong data cultures or concrete plans to build one are feeling the pressure.
Some are our clients—and more of them are asking our help with their data strategy.
Often their ask is a thinly veiled admission of overwhelm. They struggle to even articulate their objective, or don’t know where to start. The variables seem endless: data — security, science, storage, mining, management, definition, deletion, integration, accessibility, architecture, collection, governance, and the ever-elusive, data culture.
But for all that technical complexity, their overwhelm is more often a symptom of mindset. They think that when carving out their first formal data strategy, they must have all the answers up front — that all the relevant people, processes, and technologies must be lined up neatly, like dominos.
We discourage that thinking. Mobilising data is more like getting a flywheel spinning: it takes tremendous effort to get the wheel moving, but its momentum is largely self-sustaining; and thus, as you incrementally apply force, the wheel spins faster and faster, until fingertip touches are enough to sustain a blistering velocity. As the wheel builds to that speed, the people, processes, and technologies needed to support it make themselves apparent.
In this article, we offer four things you can do to get your flywheel spinning faster, and examine each through the story of Alina Parast, CIO of ChampionX, and how she is helping transform the company (which delivers solutions to the upstream and midstream oil and gas industry) into a data-driven powerhouse.
Step 1: Choose the right problem
When ChampionX went public, its cross-functional team (which included supply chain, digital/IT, and commercial experts) avoided or at least tempered any grandiose, buzzword-filled declarations about “transformations” and “data-driven cultures” in favour of real-world problem solving.
But also, it didn’t choose just any problem: it chose the right problem — which is the first and most crucial step to getting your flywheel spinning.
At the time, one of ChampionX’s costliest activities in its Chemical Technologies business was monitoring and maintaining customer sites, many of which were in remote parts of the country. “It was more than just labour and fuel,” Alina explained.
“We had to spend a lot on maintaining vehicles capable of navigating the routes to those sites, and on figuring out what, exactly, those routes were. There were, and still are, no Google maps for where our field technicians need to go.”
Those costs were the price of “keeping customers’ tanks full, not dry” – one of ChampionX’s guiding principles and the core of its value proposition to improve the lives of its customers. “And so, we wondered, ‘how can we serve that end?’”
The problem the team chose to solve — lowering the cost of site trips — might appear mundane, but it had all the right ingredients to get the flywheel moving. First, the problem was urgent, as it was among ChampionX’s most significant expenses.
Second, the problem was simple (even if its solution was not). It was easy to explain: It costs us a lot to trek to these sites. How can we lower that cost? Third, it was tangible. It concerned real world objects—trucks, wells, equipment, and other things people could see, hear, or feel.
Equally important, the team could point to the specific financial line items their efforts would move. Finally, the problem was shared by the enterprise at large. As part of the cross-functional leadership team, Alina didn’t limit herself to solving what were ostensibly CIO-related problems. She understood: if it was a problem she and her organisation could help solve, then it was a CIO-related problem.
IT executives talk often of people, processes, and technology as the cornerstones of IT strategy, but they sometimes forget to heed the nucleus of all strategy: solving real business problems. When you’re getting started, set aside your concerns about who you will hire, what tools you will use, and how your people will work together — those things will make themselves apparent in time.
First get your leaders in a room. Forego the slides, the spreadsheets, and the roadmaps. Instead, ask, with all sincerity: What problem are we trying to solve? The answer will not come as easily as you expect, but the conversation will be invaluable.
Step 2: Capture the right data
Once you’ve identified a problem worthy of solving, the next step is to capture the data you need to solve it.
If you’ve defined your problem well, you’ll know what that data is, which is key. Just as defining your problem narrows the variety of data you might capture, figuring out what data you need, where to get it, and how to manage it will narrow the vast catalog of people, processes, and technologies that could compose your data environment.
Consider how this played out for Alina and ChampionX. Once the team knew the problem — site visits were costly — they quickly identified the logical solution: Reduce the number of required site visits. Most visits were routine, rather than in response to an active problem, so if ChampionX could glean what was happening at the site remotely, they could save considerable time, fuel, and money.
That insight told them what data they would need, which in turn allowed ChampionX’s IT and Commercial Digital teams to discern who and what they needed to capture it. They needed IoT sensors, for example, to extract relevant data from the sites.
And they needed a place to store that data — they lacked infrastructure that could manage both the terabytes pouring off the sensors and the coupling customer data (which resided within enterprise platforms such as ERP, transportation, and supply & demand planning). So, they built a data-lake.
Each of these initiatives — standing up secure cloud infrastructure, the design of the data lake, the sensors, the storage, the necessary training — was a major undertaking and is continuing to evolve. But the ChampionX team not only solved the site-visit problem; they provided a foundation for the company’s data environment and the data-driven initiatives that would follow.
The data lake, for example, came to serve as a home for an ever-growing volume and variety of data from ChampionX’s other business units, which in turn led to some valuable insights (more on that in the next section).
Knowing what data to capture provides the context you need to start selecting people, tools, and processes. Whichever you select, they will lend themselves to unpredictable ends, so it’s a taxing and fruitless exercise to try and map every way in which one component of your data environment will tie to all others — and from that, to choose a toolkit.
Instead, figure out what you need for the problem — and the data — in front of you. Because you’ll be making selections in relation to something real and important in your organisation, odds are, your selections will end up serving something else real and important.
But in this case, you’ll be able to specify the names, costs, and sequencing of the things you need — details that will make your data strategy real and get your flywheel spinning faster.
Step 3: Connect dots that once seemed disparate
As you begin to capture data and your flywheel spins faster, new opportunities and data will reveal themselves. It wasn’t long after ChampionX’s team had installed the IoT sensors to remotely monitor customer sites that they realised the same data could be applied elsewhere.
ChampionX now had a wealth of topographical data that no one else did, and it used this data to move both the top and the bottom lines. It moved the bottom line by optimising the routes that ChampionX’s vehicles took to sites — solving the no-Google-Maps-where-we’re-going problem — and it moved the top by monetising the data as a new revenue stream.
The data lake, too, took on new purpose. Other business initiatives began parking their data in it, which prompted cross-functional teams to contemplate the various kinds of information swirling around together and how they might amount to more than the sum of their parts.
One type was customer, order, and supply chain data, which ChampionX was regularly required to pull and merge with site data to perform impact analyses — reports of which and how their customers were affected by a disruption in supply chain networks. Merging those data used to take weeks, largely because the two data had always lived in different ecosystems. Now, the same analyses took only hours.
There are two takeaways here. The first is that it’s okay if your data flywheel spins slowly at the start — just get it going. Attracting even a few new opportunities or types of data will afford you the chance to draw connections between things that once seemed disparate.
That pattern recognition will speed up your flywheel at an exponential rate and encourage an appropriately complex data environment to take shape around it.
The second takeaway is similar to those of the first two steps: Choose wisely among the opportunities you could pursue. Not every insight that is interesting is useful; pursue the ones that are most valuable and real, the ones people can see, measure, and feel.
These will overlap significantly with tedious and banal, recurring organisational activities (like pulling together impact reports). If you can solve these problems, you will prove the viability of data as a force for change in your organisation, and a richer data culture will begin to emerge, pushing the flywheel to an intimidating pace.
Step 4: Build outward from your original problem
The story of ChampionX that we’ve examined is only one chapter of a much larger tale. As the company has collected more data and its people gleaned new insights, the problems that Alina and her business partners take on have grown in scope and complexity, and ChampionX’s flywheel has reached a speed capable of powering data-first problem-solving across the company’s entire supply chain.
Yet, most of the problems in some way trace back to the simple question of how the company might spend less on site-checks.
ChampionX’s team has not hopped willy-nilly from problems that concern the supply chain to those that concern Marketing, or HR, or Finance; the team is expanding outward in logical progression from their original problem. And because they have, their people, processes, and technologies, in terms of maturity, are only ever a stone’s throw from being able to tackle the next challenge — which is always built on the one before it.
As your flywheel spins faster, you will have more problems to choose among. Prioritise those that are not only feasible and valuable but also thematically consistent with the problems you’ve already solved.
That way, you’ll be able to leverage the momentum you’ve built. Your data environment will already include many of the people and tools you need for the job. You won’t feel as if you’re starting anew or have to argue a from-scratch case to your stakeholders.
Building a data strategy is like spinning a flywheel. It’s cyclical, iterative, gradual, perpetual. There is no special line that, if crossed, will deem your organisation “data-driven.” And likewise, there is no use in thinking of your data strategy as something binary, as if it were a building under construction that will one day be complete.
The best thing you can do is focus on using your data to solve problems that are urgent, simple, tangible, and valuable. Assemble the people, processes, and technologies you need to tackle those problems.
Then, move onto the next, and then the next, and then the next, allowing the elements of a vibrant data ecosystem to emerge along the way. You cannot will your data strategy into existence; you can only draw it in, by focusing on the flywheel. And when it appears, you, and everyone else, will know it.
This article was co-authored by Duke Dyksterhouse, an Associate at Metis Strategy.