Executives gamble on “radical” innovations all the time. If they expect to be in business for the long haul, they must make big bets to sustain leadership – or risk obsolescence. Even when they refrain from putting money on a bold idea, they’re still placing a bet of omission.

Radical innovations tend to disrupt established markets or create new ones altogether. They contrast with the safe, modest enhancements companies regularly make to existing products and services.

When Amazon ventured outside the province of online retail and launched Amazon Web Services (AWS), a cloud infrastructure as a service offering, it was a daring move. But it paid off handsomely. AWS now accounts for 13 percent of overall revenue for the multi-billion dollar company.

Safeway, a century-old supermarket chain now owned by Albertson’s, experienced the opposite.

Safeway and Theranos: Innovation gone bad

In 2011, Safeway CEO Steve Burd bet big on a Silicon Valley startup called Theranos. The latter’s spellbinding founder Elizabeth Holmes claimed to have developed a revolutionary blood-testing device that was sure to upend incumbent diagnostics firms.

Burd hoped to use this groundbreaking technology to set off an equally dramatic rise in Safeway’s revenues, which had been in decline. The plan was to install wellness clinics, with Theranos blood tests as the centerpiece, inside hundreds of Safeway supermarkets. Unfortunately, Burd’s audacious bet failed.

By the time Theranos had been exposed as a sham, Safeway had dumped $350 million into the partnership – nearly half of its net income in 2012 – and built clinics in over 800 stores. Burd, like scores of other smart people, had been duped by the Bay Area unicorn.

If nothing else, the Safeway debacle reminds us that the waters of radical innovation are treacherous. But executives need to brave the weltering surf to stay relevant. The question then is, how can they increase the odds of pursuing radical innovations that yield profit, not pain?

To avoid placing catastrophic bets, executives must cultivate a discipline of innovation. This means instituting a consistent process, culture, and playbook aimed at uncovering novel ideas before competitors.

To start, as a part of this regimen, executives should devote 10 percent of their budget to radical innovation.

Obey the 10 percent rule

The 10 percent number stems from the 70-20-10 investment approach, which urges companies to put 70 percent of their resources into “sustaining” innovations, 20 percent into “adjacent” ones, and 10 percent into those of the radical sort. For Safeway, a sustaining innovation might have entailed selling more organic produce to better compete with Whole Foods. In an “adjacent” scenario, the grocer might have chosen to package and deliver ready-made meals to consumers to preempt Instacart’s encroachment. The shift from selling groceries to blood tests qualified as a radical innovation.

Though he didn’t invent the 10-percent  idea, Eric Schmidt popularized and employed it to great effect during his time as CEO of Google. I’ve also seen the benefits of it first-hand as an executive leader at several enterprise software companies.

Allocating 10 percent of your resources to exploring, developing, and even operationalizing big, new ideas is a wise approach, particularly when done over a long period of time. To be fair, this percentage will vary depending on the context. For example, new startups, which can only break out and topple reigning titans of industry with big innovations, have little to lose by putting most of their eggs into the radical basket. But for larger and more established leaders – with storied brands, beloved products, loyal customers, and public financials – there’s much to consider and protect, and thus a 10 percent allocation to radical innovation is fitting. What matters is that you consistently invest some amount of money in transformative ideas.

A big reason this is important is it gets you in the habit of looking for both opportunities and threats in your industry. By staying vigilant, you ward off the lethal complacency that so many business leaders fall into when times are good. What’s more, you boost your chances of finding a disruptive idea first.

You may lose money by following the 10 percent rule. After all, by definition, radical innovation success stories are few and far between. But these short-term losses pale in comparison to those you’ll incur if you get disrupted or make an oversize investment in a flawed venture.

To align with the discipline of innovation, financial investment is only part of the equation. You also need to develop a culture of innovation.

Build a culture of innovation

My vision of a culture of innovation rests on the assumption that some people are good at operational execution, others are good at ideas. You shouldn’t assign operations people to abstract idea work, nor should you rely on idea people for day-to-day operations. With that in mind, building a culture of innovation means hiring idea people and assigning them to radical innovation projects.

I’m painting with broad strokes here. Obviously, great ideas can come from anywhere – and anyone. But certain types of people thrive in experimental, creative, highly unstructured environments. These are the sorts of folks you want on your innovation squad.

It’s often best to sequester these innovators so they can probe and test and create without being hampered by operational demands or core business constraints. Depending on the size of your organization, you might consider setting up an innovation lab, perhaps off-campus and away from current organizational influences.

To a certain extent, these free spirits should be viewed like internal consultants, academic researchers, or scientists. As such, they need a high degree of autonomy. Otherwise, they’ll struggle to unlock the kind of breakthroughs that impact the long-term prosperity of your business. They also need permission to fail. This, of course, requires you to adapt your palate to the bitter taste of investing in dead-end experiments.

But bear in mind, a failed experiment is not the same as an innovation disaster. If you employ well-designed procedures for vetting radical ideas in their nascent stages and beyond, you’ll hedge against such calamity. I’d suggest implementing the gated pass-fail process for this.

Follow a ‘gated’ vetting process

A gated process entails setting well-defined gates, or benchmarks, through which an innovation must pass as you gauge its market viability. For each gate, you set specific, detailed pass-fail criteria. If an idea fails to satisfy the criteria of a particular gate, then it does not move forward. In fact, the idea might be worth scrapping at that point.

A gate must be clear but needn’t be complex. Safeway, for instance, could have set a gate during the Theranos trial run stipulating that all the blood tests must yield accurate results within a fixed margin of error.

Without a doubt, gated processes help mitigate risk when making big innovation bets. But they are wholly ineffective if you ignore the results they generate. This is arguably what happened at Safeway.

Before going live with the wellness clinics, Safeway did something akin to a beta test. As a part of this, Theranos performed its blood tests at an employee health clinic at Safeway’s headquarters. Kent Bradley, the Chief Medical Officer (CMO) of Safeway Health, oversaw the beta run. (As an aside, Safeway Health was a subsidiary spun out of an initiative Burd led to reduce the company’s skyrocketing employee healthcare costs.)

While observing the trial, Bradley noticed some alarming patterns that suggested the Theranos devices were seriously flawed. Perhaps the most disturbing example was when a senior Safeway executive took a Theranos test indicating he more than likely had cancer. As was the case for several other employees, when the executive sought a second opinion from a conventional lab, his results came back normal.

Bradley collected these red flags and eventually presented them to Burd. But it was too late – Theranos had already ensnared the long-time CEO, and he ignored the warning.

If the preliminary test at Safeway represented a pass-fail gate, then the radical Theranos innovation clearly failed. In which case, the wellness project should have been halted.

Gated processes make it easier to measure an idea’s feasibility. But they can only help you make good decisions if you set aside your cognitive biases and obey the process. Easier said than done.

Adhering to a discipline of innovation

It’s impossible to predict the future. You can use every tool and procedure for evaluating radical innovations and still end up making the wrong move. Moreover, you can be self-aware about the cognitive biases you tend to harbor and slip into them all the same.

The 10 percent rule, culture of innovation concept, and gated pass-fail process reduce the chances of that happening. And for that reason, they’re worth practicing.

In the end, Burd’s fatal mistake was in assuming that Theranos’s Holmes operated with the same level of integrity as he did. Still, a rigorous discipline of innovation could have snatched him out of the lion’s jaws. Or at least kept its teeth from sinking in so deep.

James Markarian is chief technology officer at SnapLogic, a Silicon Valley data and application integration company. He is former CTO of Informatica and previously held leadership roles at Khosla Ventures and Oracle.