The Right (and Wrong) Way to Be Influenced by Data Great data analysis might just give you a leg up when it comes to pitching your ideas.

We're living in a data-centric world. Data-influenced decision-making is increasing everywhere, and for good reason: By 2025, Network World reports that 49 percent of the world's data will be in public clouds and ready for use in data-driven decision-making, and PwC reports that companies that let data drive their decisions are three times likelier to note "significant" improvement in their decision-making efforts.

Still, that doesn't mean organizations and managers are used to incorporating data in their decision-making. In fact, Business Application Research Center notes that 58 percent of people say their organizations rely on "gut feel or experience" when making at least half of important decisions — leaving significant room for error.

Related: How to Use Real-Time Data to Fine-Tune Your Business Decisions

You can harness all of that available data to influence others. Imagine the next time you propose a new course of action armed with data to back up your points. Faced with facts and data that support your ask, your audience will have a harder time relying solely on gut checks.

Related: Why data is the world's most valuable resource today

To ensure you're using data in the right way, you must understand a few things: what the data-driven decision-making (or DDDM) value chain is, where your issue fits in and how to best leverage your place in that chain.

Understanding the DDDM value chain: Where do you fit in?

All other things being equal, when you stand up to present data (in the hopes of persuading you to take one action or another), you'll find yourself in the middle of a rigorous DDDM effort that happens in seven steps.

The first steps in the process begin by answering a series of three questions: What is the decision I want to influence? What new information do I need to influence that decision? And finally, what kind of analysis will get me that information? The process ends with the following four actions: gathering, preparing and analyzing data; communicating results; applying judgment and then taking action.

Look closely, and you'll notice two perspectives in action. Steps one through five are often taken by the person hoping to influence a decision, and the decision-maker takes steps six and seven.

Getting the most you can from data: "storytesting' versus storytelling

By the time we get to step five, you'll see the results of a data analysis, perhaps even beautifully rendered in images, charts and graphs. If the presenter has attended a course on the topic, they will likely use storytelling skills to animate their message, too. Rather than listen passively, you're now ready to get the most out of your analysis. Here's how:

1. Clarify the question

To begin, make sure you understand what question is being answered so that you can work out whether the analysis matches the question. It can also be useful to remind yourself of what type of question the presenter is trying to answer — if it's descriptive, predictive or prescriptive — to make sure the completed analysis matches the question at hand.

For example, if half of a survey's reviewers are meant to be predictive (indicative of future results) while the other half think it was merely descriptive (describing what had happened under unusual circumstances due to Covid-19), then you would waste time and emotional energy arguing across purposes, with each half not understanding why the other doesn't "get it."

It's generally good to clarify this before a presentation. If a mismatch exists, the presenter can go back and rework the analysis. At this stage, however, if you discover yourself stuck with a mismatch, the correct move is to stop, reclarify the question and start the conversation again with active listening in mind.

2. Know the "n'

In statistics, "n=" refers to the size of the samples used in data analysis. In this sense, "knowing the "n'" means understanding where the data came from, what type of data was used and how much of it was used. In many instances, your presenter will cover results from a sample of data taken from a larger data set. Your job is to ask: What was the source? What was the size? Do we have the right amount of data from the right places? Remember: It's important to check the "n" before accepting any conclusions derived from the "n."

This issue can be solved in advance by clarifying expectations around what an acceptable sample size should be. But if you find yourself downstream of that step, the onus is on you to both know the "n" and then to ask whether the n is sufficiently large to inform the decision you are looking to make. Do your research: Various industries have different standards for confidence levels and margins of error.

3. Check the vector, Victor

In data science, a "vector" is a quantity that has a direction and magnitude. I like to use a classic saying from the movie Airplane! to keep it top of mind during data analyses: "What's our vector, Victor?" For our purposes, this saying involves thoughtfully interrogating both direction and magnitude during a presentation.

Checking the direction of your results is usually straightforward: It's a positive, negative or neutral association in the data. (For example, "What is the association between marketing spend and revenue?") Checking the magnitude is different. This means measuring the size of the effect. The questions to ask here are: Is the size meaningful? Does it make sense? What impact would it have on my business if it were true? To get this right means that leaders must consider the data in the context of the business. The context is important because it incorporates the totality of the evidence available.

4. Be certain about uncertainty.

Chances are, your presenter created their analysis by taking a sample of data from a population of data and then using that sample to create a reasonable model of the total population. Because your presenter is using a sample of data but not all the data, you will have uncertainty in the analysis. Uncertainty is a fact of life, but it's important to know just how much uncertainty there is around the analysis before you allow it to influence big decisions.

In this way, "be certain about uncertainty" means you always ask for the measure of uncertainty in the presentation and then judge whether that level of uncertainty is acceptable. Of course, your determination of what's "acceptable" will depend on context: When planning out a three-year growth agenda, the leadership team at a China-based logistics operator my company worked with divided the growth projects into categories according to the expected value of each project, which calculates the possible outcome for each project by the probability of it occurring. Projects in high levels of uncertainty were staffed and managed differently from those with less uncertainty. This brings us to our fifth and final point.

5. Use your contextual judgment.

Ultimately, great data analyses are done to help inform decisions for you and your team. How much data you need will depend on the outcomes you want, how you weigh costs and benefits and (most important) your values. Using your contextual judgment means considering the implications of your decision through a wide lens, so make a list that includes how this decision will affect people, processes, profits and purpose. Good data analysis can give you interesting insights into the relationships between things, but it can't tell you how to best maintain those relationships. Here, good judgment is required.

Related: Why Business Leaders Need to Take Data Seriously in 2021

Case in point: For many firms, data shows that their productivity during Covid-19 has improved dramatically. But this data, taken without contextual judgment, misses the human cost of an exhausted workforce — an outcome that lies beneath the numbers. Microsoft compiled data on that human cost through a survey and Microsoft Teams information; it found that around 40 percent of workers say their companies are asking too much from them.

The future will arrive, quite literally, at the speed of light. With 150 trillion gigabytes of data needing analysis by 2025 according to SharesPost (and companies seeing respectable profit increases and cost reductions by harnessing it), data synthesis and data-influenced decision-making will continue to increase in popularity. And as they do, it will increase the importance of human judgment from you and your teams. You'll need more of this data — and faster.

Understanding how analyses are done, how they're presented and the best tool kits for interrogating them will ensure you're using data to influence the right decisions. When used correctly, it will make decision-making increasingly better, cheaper and faster for you and your teams. Armed with the right insights to make the greatest influence, go forth in the data-centric world.