Ultimately, data is everywhere

Data isn't just exploding in importance; it's literally exploding in volume all the time. Every day, every second of every day, more and more data is created. As of 2020, there was estimated to be 44 zettabytes of data in the world, and by 2025 there could be as much as 175 zettabytes of data.² What on earth is a zettabyte? To put it in context, a zettabyte is 1,000 exabytes, and an exabyte is 1,000 petabytes, and each petabyte is 1,000 terabytes, and each terabyte is 1,000 gigabytes. In other words, a zettabyte has 21 zeros on the end of it! It's insane to imagine that much data existing in the world.

Or is it? Because almost everything we do these days generates data. Every single interaction with your phone or computer. Every post and like on social media. Every time you walk down the street with a phone in your pocket, sending GPS signals on your location. Every time you pass by a security camera. Every time you tap and buy something with your contactless credit card. Every time you stream a movie or listen to a podcast. It's all creating data. And because this vast stream of data is only going to get bigger, it makes sense that more and more jobs are going to involve working with data in one way or another.

Explosive demand for data skills

Even for more traditional businesses, data is fast becoming the most critical business asset. Data is what allows companies to make better decisions, understand their customers better, and streamline business operations. Think about the average marketing role, for instance. Among many other things, data gives marketers valuable insights on customer demographics, so they can run targeted campaigns. Data isn't just “an IT thing,” then. It's a vital part of most modern business functions.

And this means employers increasingly need people with data skills, from the basic to the advanced. A study by the Royal Society in the UK found that demand for data scientists alone had tripled over the past five years, rising 231 percent.³ But it seems there aren't enough people out there with those much-needed data skills, since a government report on the data skills gap found 46 percent of UK employers had struggled to recruit for data-related roles in the last two years.⁴ So, on the one hand, data is the most valuable resource in the world. Yet data is also one of the main roadblocks to a company's success—largely because of a widespread lack of data skills.

The good news is that you don't need to be a data scientist to be successful. Because data isn't just for data scientists. (Although if you do fancy a career change, there's obviously a lot of demand for data scientists!)

Data touches so many roles in the average company. And this is why everyone should be able to use data to influence both their day-to-day activities and big-picture decisions. Used well, data can help you achieve your objectives at work, do a better job, and contribute to the company's overall performance. And, given the demand for these skills, and the data skills gap that exists in so many organizations, being data literate marks you out from the crowd.

Here are just a few of the ways data literacy can help you do a better job:

• You can solve problems more easily and make better decisions. Every job at every level requires some degree of problem-solving and decision-making. Whether you want to reduce waste, pinpoint key customers, increase sales, or whatever, data can help you do it.
• What's more, you don't have to rely on others for information. With modern data analytics tools, pretty much anyone can interrogate data and uncover useful insights. So if you can understand the basics of data, you can get on with more tasks yourself instead of having to wait for the analytics team to pull basic reports for you.
• Plus, you can make a more compelling case to stakeholders. Ever struggled to get buy-in for a new project or resourcing for a new initiative? Data helps you communicate your message and back up your arguments with hard evidence—meaning your pitch, whatever it's for, is more likely to be green-lit.
• You can communicate with technical colleagues more easily. These days, IT and technology functions aren't squirreled away in the company basement, doing their own thing. Cross-functional collaboration is the norm in most businesses, which means you need to be able to “speak data” with your tech-minded colleagues and ask the right questions.

Data speak 101

Bear in mind that I could write a whole book on the basics of data, and there's only so much I can cover in this brief summary. You'll need to do some self-study to learn more about the basics of data, and I give some recommendations for that towards the end of the chapter.

So what is data? Data is information, basically. Traditionally, data would be things like numbers and statistics. But these days, data can be any kind of information, including photos, videos, and text. Spoken commands to your Alexa device, social media updates, images you post online … it's all data.

Whatever the source, data always falls into one of two categories. There's quantitative data, which is anything that can be counted and measured (such as the price of a tub of ice cream, and how many tubs are sold). And there's qualitative data, which covers things like characteristics, perceptions, feelings, and descriptions (such as the flavor of ice cream, and how people feel about that flavor). In essence, quantitative data is the numbers stuff, while qualitative data is more descriptive.

Data may represent a one-off snapshot in time—such as a customer satisfaction survey—in which case, it's known as a cross-sectional study. Or it can be a longitudinal study, which means it's measured repeatedly over time to show how values change—monthly sales data being a good example.

Another term you might hear bandied around is data set. A data set is just a collection of data. And within the data set itself, each data point—whether ice cream sales, flavors, customer satisfaction scores, or whatever—is known as a variable.

Now that we've covered the basic terminology, let's look at some other things you need to know about data.

Data comes from a variety of sources

There are infinite sources of data out there to tap into. Broadly speaking, these sources fall into two categories. First, there's internal data, which is information gathered within your company—think sales and revenue reports, employee data, customer data, transaction records, business emails, and so on. And second, there's external data, which is any data collected outside the business. Some of these external data sources will be free (such as government data or Google Trends data), while others you have to pay to access (such as data from specialist providers).

Not all data is created equal

Data can be good quality or it can be bad, so you want to make sure you're working with high-quality data. (There's more on questioning and challenging data coming up later in the chapter.) In essence, good data is data that's:

• Accurate
• Consistent
• Current
• Complete (or as complete as possible). This may be as simple as understanding what “total revenue” actually includes in the context of a particular data set.

Data is meaningless without analysis

It's only when you analyze data that you can unearth interesting or valuable insights. Generally speaking, data analysis means looking for patterns and trends, and these hopefully tell a story that can inform future actions and decisions. Thanks to off-the-shelf AI-based platforms such as Amazon Web Services and IBM Watson, any business, big or small, can access intelligent tools to help them make sense of data—which means people right across the business, and not just analytics professionals, can get the best out of data. Your business may employ a range of different analytics tools, in which case part of being data literate means being able to select the most appropriate tool for the task at hand.

As I've said already, you don't need to be a data scientist to analyze data, especially with the rise of augmented analytics tools. With augmented analytics, data is automatically taken from data sources, analyzed, and communicated in a report using natural language processing (see Chapter 1) that nontechnical people can easily understand. To put it another way, augmented analytics looks for patterns and other valuable insights in the data, without needing data analysts to make sense of the data. This will drastically open up data analytics to a much wider array of businesses, helping to democratize data and turn all organizations into data-driven organizations.

(By the way, this doesn't mean the end of data scientists. Rather, the work of data scientists will move away from repetitive tasks that are easily given over to machines, and instead focus on more strategic and creative tasks, such as asking better business questions.)

Data should be at the heart of all decision-making

One of the things that makes data so powerful is that it can help you solve your biggest business challenges and answer your most pressing business questions. Therefore, a big part of getting the best out of data involves identifying the questions you most want answered, and then finding the best data to answer those questions. Maybe the data you need already exists in the organization, or maybe you need to gather new data with the help of others in the business. Either way, data can help you fill in the blanks, so you can make more informed decisions—both in your everyday work and for the big-picture decisions.

Let me put it this way: data for data's sake is pointless. You can have the biggest data set in the world, but if you're not using it to answer questions, solve problems, make informed decisions, and drive action, what's the point?

Data makes people nervous

Plenty of people love numbers, but there's a significant (arguably larger) portion of the population that really, really doesn't like numbers. As such, the word “data” provokes negative reactions in many people, ranging from mistrust and avoidance to outright fear and phobia. (There's even a name for it: arithmophobia, or the irrational fear of numbers.)

There are many reasons why people might not be thrilled about becoming data literate. Maybe they hated math at school (which is extremely common—one study found that six out of ten university students suffered from diagnosable math anxiety). Maybe they're worried about their job changing or becoming obsolete. Maybe they don't like asking questions for fear of looking stupid. Whatever the reason, fear stops people from trying new things, and it's the enemy of data literacy. Greater exposure will help overcome any fear of working with data, which is why it's a good idea to get used to your company's data and analytics systems sooner rather than later (more on boosting your data literacy later in the chapter).

Communicating insights from data is a valuable skill in itself

When you uncover insights within data, chances are you'll need to communicate those insights to others in the business (and potentially other stakeholders outside of the business). For example, you may want to use what the data tells you to support a new project, or make the case for more marketing spend, or propose a new product or service, or whatever. Data will help you make your case and get buy-in from others. But you must be able to present that data in an engaging, easily digestible way (especially considering how many people feel about numbers). Ultimately, the goal isn't to overwhelm people with how impressive your data is, but to ensure the data is understood. If the data reveals a story, your job is to find the best way to tell that story.

Data visualization is a great way to tell a story, because, as the saying goes, “A picture paints a thousand words.” There are plenty of data visualization tools out there—indeed, your company's analytics tools probably include some sort of visualization element, from simple graphs to trendy infographics. When it comes to presenting data visually you should aim to:

• Use benchmarks, such as a percent change, to help people easily see the difference between two numbers.
• Use colors—for example, red for a negative percent change and green for a positive percent change.
• Use pictures or graphics to convey positive and negative changes, such as check marks, pluses and minuses, or even weather symbols like sun, grey clouds, and storm clouds.
• But don't forget about words. Everyone digests information differently, and for some, the easiest way to understand the meaning in a set of numbers is through words, not visuals. So use simple headlines and short descriptions to highlight the meaning behind the visual, or to give a little context to numbers.

Data must be challenged

I believe critical thinking is another important future skill—so much so that I've devoted a whole chapter of this book to the topic. From a data point of view, it's really important to question and challenge data, rather than simply assume data is infallible, because, no matter how comprehensive a data set is, it's never going to be 100 percent complete—which means there will always be some level of uncertainty in what the data tells us. More than that, data can often be downright biased or skewed towards certain groups. Therefore, it's always a good idea to ask questions about the data you're working with—questions like:

• Where did this data come from? Did it come from a reputable source?
• Is this the right data for the task at hand? After all, different types of data are relevant to different tasks.
• How current is the data?
• How representative is the data? Is there potential bias within the data (or indeed, from the people looking at the data)?
• What is missing from the data?
• How has the data been analyzed?

Expensive mistakes can be avoided by simply questioning data. The Enron scandal, for example, was largely down to bad data. A simple audit would have identified that the data was fraudulent and saved shareholders the loss of billions of dollars. It's an extreme example, but it shows what happens when people don't question the data in front of them.

Bias in data is a particularly hot topic to be aware of. Data bias essentially means that certain elements within the data set—genders, races, etc.—may be more heavily represented or underrepresented than others. One of the great promises of artificial intelligence was that it would eliminate bias—after all, machines don't bring the same baggage as humans—but it turns out that AI systems can be just as biased as humans, largely because of the data that these systems are trained on. By some estimates, pretty much all big data sets are biased. And this is bad because it can produce results that are discriminatory and harmful. Amazon, for example, had to shut down a program that scored candidates for employment because it was penalizing female candidates.⁵ Solving the data bias problem is way beyond the scope of this book, but as a data-literate person, it's really important to be aware of the potential biases within data sets, and how these may affect your outcomes.

Of course, there are also biases in the way people respond to data. Research has shown that people can make completely different decisions based on identical information. Therefore, data-based decision-making can still be shaped by people's underlying beliefs and decision-making styles. This is why it's important to question decisions as well as the data itself. (Turn to Chapter 5 to read more about critical thinking, and Chapter 6 for more on decision-making.)

Correlation isn't the same as causation

It's a common mistake, thinking correlation and causation are the same thing. But just because there's a relationship between two variables doesn't mean one causes the other. To be clear, correlation means two or more factors tend to be observed at the same time. While causation means one directly causes the other(s), they're completely different things, yet they're often confused, especially in data science. But let me put it this way: just because there's a correlation between the divorce rate in Maine and margarine consumption—an actual real-world correlation, by the way⁶—doesn't mean Maine residents should avoid eating margarine to save their marriages!

Confusing correlation with causation can lead to poor decision-making, so always be wary of patterns in data and don't automatically assume that one variable causes another.

Data privacy and ethics are going to be increasingly important areas

Your company will already have data usage and security policies that you need to comply with. But going above and beyond compliance, data literacy also requires you to understand the ethical pitfalls around data. So much data contains personal information, and this is valuable stuff that needs to be protected and used responsibly. This will only become more important as regulators step up efforts to govern data collection and usage.

For me, good data governance means several things. For one thing, it means only gathering data that is business critical—as in, don't just gather data for the sake of it. It also means making people aware of what data you're gathering from them, why you're gathering that data, and how it'll be used. And it means giving people the chance to opt out wherever possible.

Of course, there's also a need to protect data from cyberattacks. (You can read more about cyber-threat awareness in Chapter 4.)

For individuals

First things first: everyone should be encouraging their employer to create a data literacy program (more on this coming up later). And in the meantime, to get comfortable using data, you can start delving into your company's data sets, using whatever management dashboards or business intelligence tools your company has. And if you don't have access to data in your role, ask for it!

There are also many self-study options online that will help you navigate data—covering everything from the basic data skills to advanced machine learning skills. A good starting point is to check out education platforms like Coursera, Udemy, and edX, as well as the excellent learning resources from the Data Literacy Project. You'll also find that there are specific data literacy courses for different industries, such as healthcare. (Coursera, for example, has a course on healthcare data literacy.)

I'd also recommend taking a basic statistics course, because this will help you understand the foundations of data and analytics, and a basic data visualization course, because this will help you communicate insights from data to others in the business.

Over and above taking an active learning approach, the best thing you can do for your data literacy is not to let fear or hesitancy around data stop you from becoming data literate. I get that there's a lot to be nervous about with data. But you really can't afford to be left behind. Data literacy will be one of the most vital skills anyone can have in the future, so try to acknowledge any fear and then find your way around it. For some people that'll mean reading whatever they can get their hands on until the topic becomes normalized. For others, it may mean simply diving in and having a go. Burying your head in the sand isn't an option.

For employers

Data literacy will be different for every business but, as a general rule, employers should look to create a baseline of foundational data literacy skills, and create a common language around data. As someone who's helped companies develop their own data literacy programs, here are my tips for boosting data literacy across your business:

• Step 1 is to understand your current level of data literacy. For example, how many people are actually using data on a regular basis to make decisions? Do managers routinely use data to back up proposals for new initiatives?
• Step 2 is to identify your fluent data speakers and data “translators.” You may already have data analysts who are used to speaking fluently about data. But you also need “translators”—data-literate people from different business functions who can bridge the gap between the tech folk and various business groups. As part of this, also look at gaps in your data communication—meaning where are the communication barriers that prevent data from being used to its full potential?
• Step 3 is to sell your people on the benefits of data literacy. If you can explain why data literacy is essential for the business's success, it'll be much easier to get people on board with data literacy training. Remember, this is an area that many people are wary of, so also try to emphasize the benefits to people's individual roles.
• Step 4 is to use examples and stories to demonstrate successes. Part of selling people on the benefits of data literacy is showing how it drives business success. At first, you can use case studies from other organizations, in or outside of your own industry. (There are plenty of use cases online, including on my website, and I also have a book of case studies called Big Data in Practice.) Over time, you'll be able to share stories from within the organization, showing how others in the business have used data successfully.
• Step 5 is to ensure access to data. It's vital that everyone is able to access, manipulate, analyze, and share the data needed to do their job well. There are plenty of management dashboard and visualization tools out there to make this easier.
• Step 6 is to create a data literacy program, but start small. There's no one single way to establish a data literacy program, and you may need different learning paths for different business functions. But it's likely to involve training on business-wide tools, plus job-specific skills. Whatever you do, start with one business unit at a time and use what you learn from that pilot program to make improvements for the next group. And try to make learning fun and engaging. Data training doesn't have to be boring.
• Step 7 is to lead by example. Ultimately, you want to create a data-first organizational culture, where data is prioritized at all levels of the business. To aid this, leaders need to prioritize data in their own work, for example, by using data to support decision-making.
• And finally, step 8 is to build a culture of continual learning. After all, this is an area that will continue to evolve, so you want to foster an environment where continual learning and curiosity is rewarded.