Duke Energy is using data to validate corporate strategies. You can, too.

March 4, 2019 Michael Coté

Nailing corporate strategy is hard. Normally, strategy is done on a yearly loop, as part of the annual planning cycle. It’s often driven by technical market analysis (i.e., “Where can we make more money?”), coupled with shiny ideas from the trends toybox. You know, like IoT, blockchain, artificial intelligence, and "social," “omni-channel,” or whatever "multi-sided platform" means.  

The problem with this process that it's thin on validation. Doing deep research in the field to backup and validate theories is often too expensive and time-consuming. And with an annual cycle, the feedback loop on testing strategic assumptions is much too long. Despite this, each year, brave strategy teams still manage to whip up a strategy smoothie full of new initiatives to put in place, existing businesses they want to kill, and optimizations (read: cost cuts) to make. The smoothie is always a bit thin and a disturbingly dark green.

However, we're starting to see how software-driven businesses can provide techniques to improve the success rate of new strategies. A product approach to software, marbled with insights from a rapid feedback loop, gives strategy groups the chance to actually test out their assumptions. Hopefully, letting them improve their accuracy in predicting new business models that'll work.

Proving that strategy knows what it's talking about

For several years now, software-driven businesses have relied on small-batch cycles to gather feedback on their product assumptions. Thanks to more reliable cloud-native technologies and lean product processes, these businesses can deploy software to production weekly, gathering feedback from how actual people use the software. Product teams at these companies can test out their assumptions of each feature, getting it right over several cycles.

For example, using this fast-feedback loop, Liberty Mutual sped up its agents' quote-to-sale workflow, contributing to a doubling of the average close rate. Orange improved the usability of its small business app, driving better customer satisfaction by streamlining common tasks like paying bills, looking at staff mobile plans, and diagnosing internet problems.

There's got to be a way to wrap this loop around the strategy smoothie, giving strategists a way to better test strategic assumptions. If the assumptions are invalidated, they can test and come up with new ones. That may seem like failure, but it means saving a year's worth of time and budget from being spent on a bad idea—and getting started faster on something that will actually work.

Duke Energy's 'journey teams' test out grand plans

I caught up with Duke Energy's John Mitchell recently and discussed one approach to small-batching strategy. As an energy company, Duke has plenty of strategizing to do around issues like: disintermediation from IoT devices, deregulation, power needs for electric vehicles, and improving customer experience and energy conservation. Duke has a couple years of experience being cloud-native, getting far enough along to open up an 83,000-square- foot labs building housing 400 employees working in product teams.

They're applying the mechanics of product teams to their strategy creation. "Journey teams" are used to test out strategies before going through the full-blown, annual planning process. "They're small product-type teams led by design thinkers that help them really map out that new [strategic] journey and then identify [what] are the big assumptions," Mitchell explained. Once identified, the journey teams test those assumptions, quickly proving or disproving the strategy's viability.

Mitchell gives a recent example: Labor is a huge part of operating a nuclear power plant, so optimizing how employees spend their time can increase profits and the time it takes to address issues. For safety and compliance reasons, employees work in teams of five on each job in the plant, typically scheduled in hour-long blocks. Often, the teams finish in much less than an hour, creating spare capacity that could be used on another job.

If Duke could more quickly, in near real-time, move those teams to new jobs they could optimize each person's time. "So the idea was, ‘How can we use technology?’" Mitchell explains. "What if we had an RFID chip on all of our workers? Not to ‘Big Brother’ check in on them," he quickly clarifies, but to better allocate the spare capacity of thousands of people. Sounds promising, for sure.

Not so fast though, Mitchell says: "You need to validate, will that [approach] work? Will RFID actually work in the plant?" In a traditional strategy cycle, he goes on, “[You’d] order a thousand of these things, assuming the idea was good.” Instead, it’s "Let's order one, let's take it out there and see if it actually works in plant environment." And, more importantly, can you actually put in place the networking and software needed: "Can we get the data back in real time? What do we do with data?"

Key to all this, of course, is putting these journey teams in place and making sure they have the tools needed to safely and realistically test out these prototypes. "[T]he journey team would have enough, you know, a very small amount of support from a software engineer and designer to do a prototype," Mitchell explains. "[H]opefully, a lot of the assumptions can be validated by going out and talking to people," he goes on, "and, in some cases there's a prototype to be taken out and validated. And, again, it's not a paper prototype—unless you can get away with it—[it's] working software."

Once the strategic assumptions are validated (or invalidated, the entire company has a lot more confidence in the corporate strategy. "Once they ... validate [the strategy]," Mitchell explains, "you've convinced me—the leader, board, whatever—that you know you're talking about."

Big strategies, small experiments, huge results

As ever, doing things in smaller batches has huge benefits. This approach gives the executive leadership team much more confidence in selecting new strategies. It also helps organizations increase innovation because the new structure allows them to try out many, smaller initiatives without all of the heavy ceremony and delay of the annual planning process.  Thales uses an innovation process that gives teams 3 months to test out strategic assumptions and product viability.

However, you can't always put something like journey teams in place. Some strategies require big bets and time to play out. But corporate strategy can always do with more data-driven theorizing and decision making. If the business relies on software, try out this journey team approach. As Duke proves, some lightweight, quick processes can result in much more confidence in your strategic choices.

If you want to learn more about how Duke Energy and other large enterprises are transforming their corporate cultures and approaches to software, download Michael Coté’s free ebook: Monolithic Transformation: Using DevOps, Agile, and Cloud Platforms to Execute a Digital Transformation Strategy

About the Author

Michael Coté

Coté works on the advocate team for Pivotal Cloud Foundry®. See @cote for more.

More Content by Michael Coté
Previous
My cloud costs what?!
My cloud costs what?!

Debates about the true cost of cloud computing have been happening for a decade, and are becoming more comp...

Next
Microservices are not the destination
Microservices are not the destination

Popular definitions of microservices omit the part about why they exist in the first place: To let develope...