Your job is not what you think it is.

What is your job?

Let's start with the obvious: your job is not to make your company money. Your job is not to "drive revenue growth." Your job is not to "cut costs." Your job is not even to "drive profitability."

No, these various jobs are just branches on a tree. And focusing strictly on one or the other guarantees you'll miss the bigger picture.

Your job is to make decisions.

These important areas—revenue growth, profitability, cost reduction, etc.—all have three shared traits that render them ineffective as performance metrics: they're measurable, they're subjective, and they're lagging indicators. That is to say, these markers are the product of decisions. Once they turn south, you're long overdue for corrections. Corrections that may not have been needed if better decisions were made in the first place. Corrections you may have run out of time to make.

I can only hear my friends in accounting and finance cringing at this take. How can numbers be 1. Poor performance metrics and 2. Subjective? Bear with me.

With the natural limits on time and resources we all experience in roles across every level, numbers never tell the whole story. And numbers, as any good CFO knows, are only as useful as the story they tell. And that story depends entirely on who is interpreting them.

Fact is, revenue may be accelerating. But it could probably accelerate faster. Profitability could be healthy, but it could be healthier. Costs may be low, but they could be lower. Or maybe they could be higher, but these higher costs may be more than offset by accelerated profitability. Revenue may be growing today, but is it due to gimmicks or promotions that destroy your brand value tomorrow? And that new partner you landed after months of wrangling details and hammering out contracts? The ensuing sales of your partner offering may be so soft it doesn't offset the huge sunk cost of the time spent building it.

See what I mean? Cherry-pick any number you want from these scenarios and you could tell a story of victory. Only with the complete set of data—and with the undefined and immeasurable variables that will play out into the future—can we tell the true story.

If we're riding a bus down the freeway into the future, success is guided by which on ramps and off ramps we take—and which we ignore—and navigating this journey increases in complexity every day as distractions to innovation accelerate.

At the crux of all of these performance indicators is a decision. But it's not one decision, and it's not one time. If we're riding a bus down the freeway into the future, success is guided by which on ramps and off ramps we take—and which we ignore—and navigating this journey increases in complexity every day as distractions to innovation accelerate. The fact is, we rarely know if we've been successful. We only know if we're on the right path, and if we're making the right decisions. A decision that benefits you today at the cost of tomorrow is a bad one. But it may not look that way in the books ... yet.

So the $10 million question is: how do we make good decisions? That's a topic worthy of a novel, but here are some guiding tips:

What to do:

1. Gather all of the meaningful data points you can. Quantitative and qualitative.
2. Identify gaps in your knowledge. Write out the assumptions you have to make to bridge them.
3. Write out your hypotheses. Layperson terms is fine: "If I choose this, then this will happen. If I choose otherwise, then this will happen."
4. Depending on the weight of the decision, test your hypotheses. Your results may not need to be statistically significant. If you're looking for directional feedback, aka a "gut check," run the decision by a random handful of customers, coworkers, peers, even family members who might understand it. Look for red flags. Generally speaking, your testing plan should be commensurate with the impact(s) of the decision.
5. Look to your company's vision. A documented vision should always be the trump card when everything else is fuzzy, but a decision needs to be made. At Intuit, I know exactly how to make well-informed, defensible decisions, because when everything is on the line I can ask "Is this in the best interest of our customers?" If yes, I know no matter how the chips fall, I can defend the choice, using shared language guided by our vision.

What to avoid:

1. Analysis paralysis. Don't get lost looking for an answer in data you don't have. Have a bias for action. Write out those assumptions, and write up a testing plan that stabs them in the heart as quickly as possible.
2. Indefensible decisions. Every decision we make is with limited data and limited foresight. So long as you can defend the approach and the choice, you're doing it right. If you can't, review the tips above.
3. Blame. The goal of this whole process is to eliminate bad decisions. But that doesn't mean something won't happen in the future that will turn a good decision into a bad one. When that happens, blame doesn't need to be assigned. So long as you've learned something—and you've documented the learning for your organization to digest—you're doing the best anyone can do.

No matter the role, from sales to customer support to executive, the decisions you make on the customers you pursue, the products you launch, the people you hire, the things you do with your free time -- these are the decisions that will drive success for you and your organization. And when you make good decisions, the numbers will follow.