I recently interviewed “Freakonomics” author Stephen Dubner for the Alter.Next Virtual Summit: Accelerating Analytics Maturity to Win. Stephen Dubner has not only written the best-selling “Freakonomics” book series but is also the host of the “Freakonomics Radio” podcast and an award-winning journalist. He has broad experience investigating businesses of every size and type, having spent most of the last two decades studying “the hidden side” of hundreds of organizations. Stephen shares his learnings — from esoteric academia to the best practices of leading global enterprises — in engaging, simple-to-understand terms.
During our interview, Stephen and I discussed the challenges and value created by data and analytics in business. He shared guidance worth following in any modern organization. Stephen identified surprising places where analytics drives successful enterprises, creating competitive advantages and building resiliency in the business. Plus, Stephen shared how everyone can take greater advantage of the data that surrounds them. Below I’ll share my takeaways from the conversation, and I encourage you to check out the full interview at Alter.Next.
The analytics revolution is an elephant
I asked Stephen how he’s seen data and analytics change over time. He told me he believed the analytics revolution is like the allegorical elephant from “the blind men and an elephant” fable. To paraphrase Stephen:
The analytics revolution is totally different things to different people. Those who generate, analyze, and try to make actionable sense of data see most of the analytics ‘elephant’ and know what they’re doing. But typically, the C-suite of a firm don’t have a deep understanding of analytics and may misunderstand it. For customers of the firm, the analytics revolution can mean something different still — for many, it can mean an invasion of privacy.
Indeed, the analytics revolution has unsettled many people’s relationships with data. In his story, Stephen points out a modern truism: no one can completely opt out of data. In a world where businesses run on big data and the average person’s day-to-day decisions are influenced by machine learning algorithms, saying “I’m not a data person” puts you at a distinct disadvantage — whether you’re a consumer or a business leader. As the conversation continued, Stephen conveyed his belief that we can all become “data people” of a sort, and that doing so is easier and more fun than many expect.
Analytical and technical are two different things
Stephen discussed his observations of real people and organizations responding to the analytics revolution, and he highlighted a common misunderstanding: people confound being analytically minded with being technically minded. Stephen asserted that one doesn’t have to be a “math-y” person to succeed with analytical thinking and that when it comes to choosing between the two, he prizes the analytical mindset. He shared (again, paraphrasing):
People are realizing that data make a fantastic tool, although a key component to success with data is asking the right questions of them. People are also realizing that analytics is an incredibly useful way to tell stories from data. Stories are critical because most people process information better when it’s in story form.
Stephen went on to emphasize how critical it is to be able to read, understand, create, and communicate data as information—a skill set commonly referred to as “data literacy.” It’s what allows us to approach ideas, challenges, and the world as a whole in an analytically-minded way. Note that developing data literacy doesn’t require learning a specific technology, framework, or coding language. Rather, data literacy is similar to critical thinking.
What does data literacy look like in a business context? A professional with high data literacy:
Understands which data are relevant for their team and organization. For example, if you’re trying to understand customer behavior on your Australian website, you’re unlikely to benefit from looking exclusively at data regarding your German customers’ offline product preferences.
Asks questions about the output of algorithms. If you have two prospective customers, and one of them was marked by your CRM with a higher purchase propensity than the other, why was that? What behaviors, demographics, and other factors did the algorithm look at? Why is one potential customer more promising than the other?
Separates lagging indicators from leading indicators. If you monitor safety at a construction site, there’s a difference between the percentage of your workers wearing hard hats (a leading indicator — something you might be able to directly control) and the number of accidents on the site in a given month (a lagging indicator — a measure of how effective your interventions were).
Designs and conducts structured experiments to test theories. If you run a website, you can leverage A/B testing — two different configurations of your website which are randomly offered to different site visitors — creating a “control” group and a “treatment” group. When you observe the groups’ different behaviors (how long they stay, what they click on, whether they make a purchase, etc.), you learn whether the treatment or the control version of the site creates the result you’re hoping for.
Communicates data to business leaders via stories and business cases. If you’re designing a new product and want to share user testing results with leadership, you can point out that 80% of users liked design C and 50% of users become frustrated with design D. When you add the stores of individual users — how someone uses the product in the course of their day, why they’d tell their friends to buy it, and what they said while using the product — you create a detailed picture that helps your leaders make better-informed decisions.
As you can see from the above, data literacy has many valuable applications in business. Plus, these behaviors don’t require a technical background or education.
Why a more numerate world is a better world
Stephen discussed the practical value of data literacy in daily life and in business. He also mentioned the world might be a better place if we were all more comfortable with data. He said:
“A little bit more numeracy in the world is good, because — if for no other reason — you can see a politician’s campaign ad that is demonstrably full of garbage and understand why it’s full of garbage. For example, if something doubled in the last year, but it started from a base of one, that’s not necessarily a significant change.”
As discussed above, data literacy enables objective problem-solving. In a data-literate organization, business decisions are based on demonstrable facts of a situation rather than on intuition, stereotypes, anecdotal experience, or impressions.
So, it’s no surprise that data literacy has become a standard business skill. Data propels innovation in the modern market. Companies leverage machine learning and AI to gain insight and keep pace with ongoing changes in technology. Indeed, machine-learning algorithms can be found in nearly every business system. As organizations continue to amass data, all employees — not just “the data people” — must become data literate to perform their roles and sharpen their organization’s competitive edge in an aggressive global economy. According to an MIT Sloan professor:
“In a world of more data, the companies with more data-literate people are the ones that are going to win.”
If you can critically examine information, its limitations, and what conclusions you can draw from it, you’re harder to mislead with incomplete data, or data that’s been manipulated to serve someone else’s agenda. This matters in business, as knowledge workers who understand how data was created and can challenge the output of algorithms can better protect and grow the business than those who assume a decision made by some “system” is always right.
Learn more about how leading enterprises worldwide prioritize data literacy in the “Future of Intelligence” IDC Survey: Importance of Data Literacy.
Every business is a data business
I asked Stephen how his work on Freakonomics Radio relates to data and analytics. He told me:
“What I’ve tried to do with Freakonomics is tell stories that are true. It’s journalism, nonfiction, plus a lot of data from interviewing academics and other experts.”
Since his original book, “Freakonomics,” illuminated obscure economic phenomena, Stephen expanded on the topic, saying (paraphrased):
To me, economics is usually misunderstood as the study of money. Economics is really the study of incentives and how people get what they want and need, especially when in competition with others who may want and need the same thing. The best version of it is: how can we be optimal? How can we optimize so we can get what we each individually want, but do it in a way that makes society stronger? That’s the big challenge.
Seeing economics as a specific sort of optimization problem was eye-opening. Indeed, across industries and business units, we’re all attempting different kinds of optimization. In supply chain, the optimization question may be: how can we get goods to where they need to get, quickly and inexpensively, without sacrificing safety or auditability? In the office of the CFO, the optimization question may be: how can we make the highest return on assets, capital, and equity while controlling costs and maintaining financial compliance?
Although the phrasing varies across businesses and functions, all business optimization problems must be addressed with data — and specifically, by subjecting data to analysis. So, does having more and better data, and more and better analytic capabilities, lead to better business outcomes? Experience says: yes. According to the International Institute of Analytics, the more well-developed an organization’s analytics practices are, the better its performance.
This concept is called “analytic maturity,” and it’s commonly measured in five stages. Organizations at Stage 1 of analytic maturity may struggle to organize and understand their data, so they make critical business decisions based mainly on the opinions and instincts of their leadership teams. On the other hand, organizations at Stage 5 of analytic maturity have optimized business decisions with data and analytics, leverage analytics to gain competitive advantages, and may even disrupt their entire industry.
Having a clear view of where your business lies on this spectrum of analytic maturity is crucial to competing well and planning the right near- and long-term investments. If you’d like to benchmark your business, visit the Alteryx Analytics Maturity Assessment. In 15 minutes or less you’ll receive a custom report outlining your organization’s data and analytics strengths and challenges, plus resources you can use to accelerate your business to the next stage of maturity.
How to own your personal analytics journey
Businesses aren’t the only entities that take analytic journeys. Stephen pointed out how everyone can be quickly empowered to better understand and leverage data with a few simple questions. He said:
“There’s a set of questions I like to ask — and anyone can do this — starting with, ‘What’s the best evidence that the argument you’re making is true?’ That’s a great way to get a sense of your data set. ‘How big is your data set?’ ‘How representative is your dataset?’ ‘Does your data represent enough people, enough time, et cetera, et cetera, for us to draw a worthwhile conclusion?’”
Stephen dove deep into this topic, sharing questions about missing data, the way in which data was acquired, types of data, and how data gets used. If you’re interested in discovering all of Stephen’s critical data questions for everyone, I encourage you to watch the interview at Alter.Next.
Improving your data literacy starts with asking questions, but there’s more you can do to develop your analytic skills. First, explore what’s available in your own organization. Perhaps your company offers data and analytics training programs? Are there learning and development funds to take courses outside of work? Are mentors available who have analytical wisdom to offer?
Even if your company doesn’t have in-house analysts or data scientists you can talk to, connect with someone who works with data. Check networking websites, like LinkedIn, or local networking groups. Building contacts in the data world opens opportunities to collaborate on a shared analytics project or have someone guide you on an analytics project of your own. You can get a jump-start with the Alteryx Community — a space full of analytic superheroes happy to connect, share stories, and lend a hand. Check whether there are Alteryx ACEs and Alteryx User Groups in your area. With Alteryx User Groups you’re welcome to join and learn — even if you’re not an Alteryx user. After all, any analyst will tell you the great joys of data and analytics are collaborative problem solving and learning.
If you’re curious about the technical side of analytics, consider exploring open-source sites like Stack Overflow and GitHub. They get you free access to developer communities that are happy to share their expertise and insight into what code can offer you.
Finally, explore MOOCs offered by Coursera, Udacity, DataCamp, and more. Data literacy programs abound in the online world! Learning this way offers flexibility to start and study at your own pace while engaging with and receiving feedback from fellow learners. For example, Udacity partnered with Alteryx to offer a nanodegree program: Predictive Analytics for Business.
How to lead your organization’s analytics journey
If you can direct your organization’s policies and priorities, you can choose from many tactics to quickly improve data literacy in the business. First, define data literacy goals. As different roles carry different desired levels of fluency with data and analytics, have this important conversation with other key stakeholders and leaders.
Once data literacy goals are agreed upon, assess your employees’ current skill levels. Develop an assessment customized for your organization or leverage the Alteryx Analytics Maturity Assessment to get a clear picture of capabilities across the business.
Next, lay out appropriate learning paths based on your data literacy findings and goals. Developing an in-house data literacy program can be driven largely by employees, or you can engage outside help as well. In this context, many analytically-minded leaders rely on the Alteryx Analytic Automation platform to educate and empower teams to deliver high-impact business outcomes with data.
Then comes implementation. Offer data literacy trainings within your organization, ideally in multiple formats. Everyone learns differently; some of your employees will excel with hands-on exercises, while others will thrive taking self-led courses. Ensure the program’s success by pairing data literacy training with hands-on, valuable projects that carry measurable performance indicators. That way you can drive results and gain a clear view of the training’s value-added outcomes.
Discover how leaders at organizations large and small have successfully driven data literacy initiatives, by visiting 4 Tips for Growing a Data-Literate Workforce.
Finally, turn advancing your organization’s analytics maturity into an ongoing practice — not just a one-time endeavor — by fostering a learning culture within the business. Create an environment that rewards curiosity rather than punishes a lack of data literacy. In this way, you’ll head off poor data practices, such as hiding data or manipulating data via vanity metrics. Also, be sure to recruit support from other leaders, as all top executives should be onboard and modeling the desired behaviors and results.
Get the deep dive into successful strategies that improve your organization’s use of data and analytics and drive superlative business outcomes with Automating Analytics: A Human-Centered Approach to Transformative Business Outcomes. In it, you’ll find dozens of business case studies and examples of how deploying analytic automation with a data-literacy-first philosophy quickly leads to measurable business results.