I have always believed the following adage is true: “You don’t plan to fail, you fail to plan.” However, more recently in business, and especially fast-paced Silicon Valley, the mantra has been to “fail fast.”

Many Silicon Valley CEOs are on record saying they are trying to instill this concept into the DNA of their corporate culture. They want their employees to be able to develop products quickly to gain market share, like now, and profit in a high velocity and often fickle technology marketplace. It’s so critical from their perspective that they say they actually embrace failures as part of the process. (“You have to break some eggs to make an omelet” would be the appropriate folksy analogy for such a strategy.)

Much has been written in business and management journals about slight variations on this theme: Fail fast, fail often. Fail better. Fail smarter. Create room for the honorable failure. Iterate quickly and often. Whatever wording you decide on for your own organization, it should have these guiding principles:

Managing risk doesn’t mean risk avoidance: The sentiment here is pretty obvious, that risk taking is needed, especially in fast-paced technology businesses. Those who don’t innovate will definitely fall behind and might not survive in the long run.

However, I think all too often, in their need to be quick with solutions, employees may get the wrong impression. They may think the company doesn’t need a managed product lifecycle process, product development roadmap or even a reasoned method for providing limited resources to unlimited R&D demands. CEOs need to ensure that even with the “fail fast” mantra, people are still expected to give it their best shot the first time around.

This is not about just pushing stuff across the line and scrambling to fix it once it’s in the marketplace. You’d get stuck with a broken product, impaired customer loyalty, rotten PR, increasing warranty claims, and more.

Make sure that you walk the talk: In my experience working at both hardware, software and medical device companies in the Valley for over 25 years, I have seen my share of failed product strategies, product launches, marketing launches, and even accounting projects, ERP implementations and audits. However—and this may come as a shock to you—I have never been invited to a party to celebrate a failure. I have never listed a failure on my list of accomplishments at the end of the year during performance review time. And I have never seen anyone promoted immediately following a major (or even minor) failure.

The reality is that management has a tough challenge for ensuring that “fail fast” isn’t just a slogan that everyone pays lip service to while being secretly terrified of making a mistake. How mistakes are dealt with on an organizational and personal level will be critical for the “fail fast” mantra to become a natural part of the evolutionary cycle of any company.

Do the hard stuff first: I think this is the key message of “fail fast.” It’s all about having your eye on the right ball. All too often, human nature is inclined to do the easy (and cheap) stuff first to show some progress. But it might not be the most important stuff. In organizations, this manifests itself in glowing status reports that a project is on track and on budget when in truth the targeted release date is in serious doubt. It will look like a lot of the to-dos have been crossed off, but when you look closely only the easiest parts of the product development have been tackled.

Using an example, assume a product has two key projects that need to be completed for it to go to market. Project A is estimated to cost $1 million, and the engineering group figures it has a 90% chance of success. Project B costs $1.5 million and has a 50% probability of success.

Now let’s say the organization doesn’t have much of a product development strategy and immediately wants to focus on completing Project A, mostly to gain favor with management because of its lower cost and higher probability of success. This would be a mistake from both a cost and timeline perspective. Statistically speaking, the probability of success for the completed product before any work begins is really just 45%—the 90% multiplied by the 50% (which would be extremely high for most companies in Silicon Valley, but let’s just stick with this easy math). However, after completing Project A, the probability of success for the whole product increases only to 50% and the organization is still left having to spend $1.5 million to complete Project B.

Consider what happens if the organization took on Project B first. The results can follow only one of two paths. Either Project B is successful and therefore the probability of the entire product’s success skyrockets to 90%, or the company “fails fast” and avoids 100% of the cost of doing Project A. Doesn’t hurt as bad, does it?

Failing fast is still failing: Arguably, failing fast is only the third best choice in a four horse race. The only thing worse than failing fast—after succeeding fast or slow—is failing slowly and not vectoring quickly to take on the next attempt. Again, not to be overly simplistic, but the wording can matter here (I would recommend ultimately adopting some internal slogan that didn’t use “fail” as the first word—but I get the appeal of the slick alliteration of “fail fast” that’s easy to remember).

So what to do if the organization does making that losing lap? Make the most of it and move on. Try the next strategy. Some of the most successful companies had initial failures and got right up and tried again. It might not have been until their third or fifth product evolution that they hit it out of the ballpark. This work is all about absorbing key learnings and vectoring quickly to a new strategy that makes more sense.

Ensure that robust processes are in place for the organization and employees to review and understand what went wrong. Did the market shift? Economics? A major competitor threat emerge? Today’s business environment, especially here in Silicon Valley, is moving at warp speed. This is especially true of a large multinational organizations where visibility into all of its parts may not be that good. Internal communication forums like blogs and websites where employees can share lessons learned from product development and project management will help the organization move in a new direction—to succeed fast.

Not all risk is created equal: While the failing fast and vectoring quickly strategy is a natural part of the tech product development world, it is not a wise choice for the accounting and finance functions. It is important to take appropriate risks and move quickly but wisely. Taking too much risk in projects like a new ERP implementation to accommodate a newly issued revenue standard or other structural items probably isn’t the right approach. The cost of being wrong is too high. Compliance type functions like SOX, external reporting, tax, etc. demand extraordinary precision and careful processes. Fast growing companies upgrading systems should plan carefully and not shortchange the testing and deployment phases, or their project might crash and burn.

I believe inventor Charles Kettering was right when he said that “a problem well stated is a problem half solved.” Iterate fast and get that prototype out in the open for a test drive. If the reaction is horrible, live and learn. The final product will be better off based on those lessons learned.

John Cook is a member of the RoseRyan dream team. He is a CPA with over 25 years of experience working in finance and accounting organizations in Silicon Valley with a focus on operational finance and technical accounting.