Several years ago, an international utility in North America with millions of customers set out to become more digitally adept. On paper, it was about as big and as ambitious a “digital transformation” as a CFO could face. It was going to be expensive — and as it turned out, frightfully risky.

The goal was to move beyond simple tasks like online bill pay and offer customers new ways to manage their accounts, adjust their energy usage, and sign up for new products and services.

And that, of course, would require investing in countless back-office changes. It would mean creating the ability to monitor customers’ energy use in real time and providing installation and repair crews with complete and constant transparency into the grid’s operations and into each end-user’s quality of service. The “to do” list went on and on.

What could possibly go wrong?

Plenty, apparently. In fact, digital transformations have a notoriously high failure rate — as high as 86%, according to a McKinsey study last year. Companies also often perceive the cost of the transformation program to be the cost of bringing technology in, but they fail to account for the far-reaching impact these programs have. So the risks grow.

By next year, businesses around the world are expected to spend nearly $2 trillion on digital transformation projects, according to IDC.

What went wrong in the case of the utility company is it took a moon-shot approach to the project. There was a $250 million budget. Hundreds of people were brought for a five-year digital overhaul. And the project roadmap showed little noticeable benefit until the entire project was complete.

It was, in short, an old-style “Big T” — a massive, unwieldy big transformation project. And as an all-or-nothing endeavor, it was fraught with risk, because its projected ROI didn’t account for the extent of change the organization could absorb or for the constantly evolving business landscape the transformation had to address.

A couple years into the effort, the utility realized it was burning through a lot of time and money. It knew it had to stop and, unfortunately, start again. Recognizing the signs of potential failure, the company went back to the drawing board to completely change its transformation approach.

It realized that with the rapid pace of technological change, successful digital transformations should not rely on make-or-break moon-shots. To succeed, the new approach had to treat the digital transformation project as a continuing series of smaller, bite-sized projects.

This results in a more agile approach, but agility in this context is about more than being fast and nimble. It is about recognizing that you can take the take the risk out of such a typically bloated capital expenditure program by adapting the concept of minimally viable product (MVP) to digital transformation. The concept is popular among start-ups, where essentially only the most basic features are in place before a product launch, and new features are then added incrementally.

Applied to digital transformation, the idea is to break down what otherwise would be a massive project into small incremental changes across the enterprise, and then determining how they deliver outcomes in the context of the larger enterprise ecosystem.

Companies can evaluate the ROI delivered through the increments and decide whether to proceed or pivot. Put another way, by breaking massive transformation programs into smaller more digestible projects, the MVP approach will de-risk the Big T beast.

So, instead of thinking of digital transformation as a bet-the-whole-company proposition, it’s better to think of it as a continual process — a “perpetual transformation.”

The MVP approach gives CFOs more flexibility and latitude as chief change agents, as they can more effectively manage the financial and strategic risks of the digital transformation.

That’s the beauty of today’s digital technology. Now that the cloud, IoT, automation, AI, and agile methodologies have made organizations less beholden to a single vendor’s approach compared to a few years ago, enterprises are empowered to become more flexible and experimental in their transformation programs and taking advantage of the opportunities they create.

By embracing a mindset of risk acceptance, and by using an MVP strategy, companies can more easily learn from small setbacks or failures, adapt quickly, and run with a revised approach.

In fact, when that big utility changed its approach to get its digital project back on track, it was the CFO, along with the senior vice president of business transformation, who embraced the perpetual transformation approach to accelerate the process.

But it’s not a matter of racing to the finish line, because there is no finish line. Once a company can master the perpetual transformation model and make it its own, it can continually refine its digital journey. No moon-shot required.

The company learn from what works and what doesn’t, and enjoy a constant cycle of greater innovation with lower risk. No moon-shot required.

David Jordan is vice president and global head of consulting and services integration for Tata Consultancy Services.