The All Analytics Podcast: Recent Episodes
All Analytics
Big data and analytics are critical to the success of an enterprise today, but doing analytics right isn’t easy. The expert guests on All Analytics Radio share their advice and opinions on what it takes to succeed with big data and analytics initiatives. Learn about strategies for customer intelligence, the Internet of Things, predictive analytics, data privacy, visualization, the analytics talent shortage, and other challenges and opportunities that face the data science community.
Join InformationWeek and AllAnalytics Radio as we welcome top quantitative recruitment expert, Linda Burtch. She is founder and managing director of the quantitative executive recruitment firm, Burtchworks, and has worked in the space for decades. Burtch shares her advice for quantitative professionals looking to keep their careers on track -- and properly paid -- in 2018 and beyond.
Why are some retailers outperforming the competition? What secrets do the top Omnichannel retailers share?
Join AllAnalytics Radio as we welcome retail experts Brian Kilcourse and Paula Rosenblum to share the results of their research on how some retailers are outperforming the market, even in today's tough environment. Kilcourse and Rosenblum are analysts at RSR Research, specializing in insights and advice for the retail industry.
Learn what factors you should focus on to make your business one of those overperformers.
Lean Analytics author Alistair Croll joins AllAnalytics radio on Monday, Nov. 27 at 1 pm ET/10 am PT. In the 4 years since Lean Analytics was first published, the book has helped companies determine their business models and growth stages and identify the One Metric That Matters. Following this path can help companies reduce risk and move faster in product development and customer satisfaction, following the Lean Startup methodology.
Lean Analytics grows out of another book, Lean Startup, which spells out a methodology for organizations to reduce risk and improve product development by creating a minimum viable product and then engaging in iterative product development that incorporates customer feedback. A number of companies have implemented this methodology, including Intuit, GE, and Dropbox.
Building on that idea, the book Lean Analytics helps organizations measure their progress and make better decisions along the way. A preface to the book describes it this way:
"Lean Startup helps you structure your progress and identify the riskiest part of your business, then learn about it quickly so you can adapt. Lean Analytics is used to measure that progress helping you ask the most important questions and get clear answers quickly."
Croll is a frequent speaker at industry conferences around the world on the topic of analytics and lean startups.
In this show we'll talk about:
- What is Lean Analytics and why do you need it
- What companies are using Lean Analytics, and how does it work for them
- What has changed in the field of analytics since the book first hit the market 4 years ago
- What mistakes and best practices are there for Lean Analytics
What do black holes, hedge funds, and glaucoma have in common? They are all subjects that Intel Chief Data Scientist Bob Rogers has worked on over the course of a long career in data and analytics. And they are all examples of the varied problems where advanced analytics can be applied.
These days Rogers brings his experience with data science, AI, machine learning, and related technologies to a broad range of applications as he helps enterprises with work on their big important projects. It's part of a program in place at Intel to help spread the knowledge and implementation of advanced analytics to businesses and other organizations.
AllAnalytics Radio welcomes Bob Rogers as our guest on Wednesday, Nov. 1 at 1 pm ET/10 am PT. We're excited to hear about what he's working on today. We are going to ask him about AI, Machine Learning, and talk about some projects that apply these technologies to solve some of society's problems.
As Intel Chief Data Scientist, Rogers has a unique view into the issues, challenges, and trends across multiple industry verticals when it comes to these technologies.
We'll ask him about the trends he sees in the field, the most interesting projects he has encountered, and some of the mistakes he has seen as organizations work to implement these technologies.
We'll also ask him for his recommendations on how to tackle these problems. What are the best practices? What are good choices for initial big-impact projects?
Students of data science who are entering universities tend to have two things in common. First, they excel at and/or love math. They are mathematicians. Second, they have a passion for fixing a particular problem. Maybe they think younger students could be able to do better in school. Maybe they are looking for a cure for a particular disease. Maybe they see issues with financial reporting that they know they could fix if only they had skills and tools.
Now consider the people who did not pursue a degree in data science. They have other careers. But they still want to be able to use data and analytics to help solve problems. These people are so-called Citizen Data Scientists. It's not a formal job title. These people may have other titles, such as sales representative, marketing director, social media associate, or something else. Data may not even be an official part of their job descriptions.
But they want to use data to make a difference. Maybe they have taken some data or statistics-related courses here or there. Maybe they use their analytics skills in their jobs, and maybe they use them in a hobby or some other way. They are not full-time data scientists. But they are passionate about solving a problem, asking the right questions, and finding answers. They care and are willing to learn and work.
These individuals can bridge the gap between mainstream self-service analytics by business users and the advanced analytics techniques of data scientists, according to market research firm Gartner.
Citizen data scientists in an enterprise organization can strengthen the commitment to analytics adoption. These individuals possess the domain knowledge so that they know the right questions to ask. And these individuals are already embedded within functional departments of the organization. They are on the front lines, dealing with the every day challenges that analytical insights can help influence and resolve.
On Thursday, Oct. 26, at 1 pm ET, All Analytics Radio will assemble a panel of our expert bloggers to talk about citizen data scientists. We'll discuss who they are and how they can help. Are you a citizen data scientist? Do you want to reach out to potential citizen data scientists inside your organization in order to strengthen your organization's data analytics practice? Then you won't want to miss this show.
Among the industries disrupted by the internet and growth of the internet, retail and entertainment are two of the big ones.
Retail has experienced countless online-only upstart companies that let consumers search for what they want and shop on price. And on the other side of the online upstarts are the giants like Amazon. It's turned what used to be a destination experience into a commodity transaction experience.
Similarly, the TV entertainment industry now has competitors in the form of streaming services such as Netflix, Amazon Prime, Hulu, and more.
QVC sits at the crossroads where the entertainment industry and the retail industry intersect. This online shopping channel has grown to much more over the years, with mobile and web channels in addition to its TV channel. And as so many in retail and entertainment are looking to leverage their data for insights, QVC is doing the same.
David O'Toole, Director of eCommerce, at this hybrid entertainment and retail company, will join AllAnalytics radio as our guest on Tuesday, September 26, at 2 pm ET/11 am PT to talk about the analytics experience at QVC.
O'Toole will tell us about why QVC has created a new real-time data collection and analysis program. Join us as we ask O'Toole:
- Why analytics are so important to the retail and entertainment success of QVC
- What unique challenges QVC faces as a retailer without physical stores
- Why real-time data is key to success for QVC
- And what lessons other retail and entertainment firms can learn from QVC's experience.
We hope you'll join us on Tuesday, September 26, at 2 pm ET/11 am PT for this fascinating view into the world of analytics at a fascinating company.
How do you transform and prepare a big organization to be ready for opportunity and success? Many companies are facing that challenge right now as they prepare to compete in a new era of analytics-driven business. And who knows better about a job like that than the CTO of the US. Megan Smith was the third Chief Technology Officer (CTO) of the US in the Obama Administration, and left the role this year. She joined a special edition of AllAnalytics Radio on Monday, September 18, to talk her experience transforming a large organization and more.
Smith calms those fears about everyone losing their jobs to automation and offers innovative ideas for growing a new generation of tech-savvy citizens. For more about Smith's presentations at SAS Analytics Experience, visit AllAnalytics.com
Privacy is on the docket today, but not in the usual way, where we focus on how to protect the data your company holds or our personal data. Our topic is the European Union's General Data Protection Regulation, or GDPR.
Don't think that because your company is based outside of Europe that it doesn't matter to you. There's a very good chance companies in the US, Canada and elsewhere will be subject to GDPR simply by holding data about EU citizens.
Failure to comply with GDPR could put your organization at risk for millions of dollars in penalties.
Our guest today is Todd Ruback, who is Chief Privacy Officer and Vice President of Legal Affairs at Evidon, which is a firm that helps other companies with data compliance efforts.
Todd will outline some of the things that you need to know about GDPR and how to prepare for its May 2018 kick off.
Welcome Todd. How are you today?
Before we get started, I have a note for the audience. You can share your own comments and questions for your peers and our guests in the text chat box appearing on the streaming audio page on All Analytics. If you enter a comment and you don't see it in a few seconds, please refresh your browser and it should appear.
Also I want to bring to your attention that we have additional resources related to GDPR on All Analytics.Com. Whether you are listening live on All Analytics or through the All Analytics Podcast on iTunes, you can access a directory of GDPR-related content by going to AllAnalytics.com/GDPR
Now, on with today's show.
As data and analytics have gained more attention as key components of successful modern businesses, it's not a surprise that we are seeing more c-level executives charged with driving analytics and data initiatives.
CEOs are recognizing the strategic importance of putting data at the center of the business, business decisions, and even communications with customers. And as firms now have more respect for the critical importance of their data, they also need executive-level workers to oversee the efforts.
Ambitious analytics pros looking to make their mark may want to consider the chief analytics officer job as a place they want to be in 5 or 10 years. But what are the day-to-day duties? What are the career paths from CAO?
To find out, we've invited Dun & Bradstreet CAO Nipa Basu to AllAnalytics radio to talk about her job, her duties, and what an CAO does every day. Consider it your opportunity to score an information interview with an insider who can answer your toughest questions about the job. And be sure to bring those questions to our live show on Wednesday, July 26 at 1 pm ET/10 am PT. You can register for the show now at this link.
In her role as CAO at Dun and Bradstreet, Basu not only oversees her own company's analytics strategy, but she also has an insider view into how CAOs work inside other companies.
She's served as CAO at a time when the role has risen in importance.
A study of 180 chief data officers and chief analytics officers by market research firm Gartner in in November 2016 found that CAOs and CDOs are more likely to report to the CEO than they are to anyone else (30%). The second most likely executive for them to report into was the CIO (16%).
(Gartner VP Mario Faria told me last year that the shift in CAOs reporting to CEOs from CIOs is moving at a much faster rate than Gartner had expected.)
CAOs and CDOs have increased in numbers, according to Gartner, from just 15 in 2010 to 1,400 in 2016.
And if your ambition extends beyond CAO, there's a path for you, too. Gartner predicts that 15% of successful CDOs and CAOs will move into other C-level positions, including CEOs, by 2020.
Amid a turbulent time for the news media and a groundbreaking time for the advancement of data analytics, Bloomberg is positioned to apply the benefits of AI and machine learning to news gathering, information gathering, and more. Bloomberg CTO and head of data science Gideon Mann joins All Analytics Radio. Mann's work touches a wide range of issues, from how data science can be used for news gathering and reporting and fact checking.
There's a popular W. Edwards Deming quote among data professionals: "In God we trust; all others bring data."
But can we even trust data anymore? It seems like data is often misinterpreted or misrepresented. Every day we are bombarded by headlines about new surveys and studies that tell us what we need to do to be healthier, have a better career, be happier, be smarter, and be successful at dating or in a relationship. Yet, as data professionals, we know that correlation doesn't mean causation. Right?
So when we see a headline that says "Grilled cheese lovers have more sex and are better people, according to survey," we probably won't out and order a grilled cheese sandwich for lunch.
Indeed, according to John H. Johnson, who looked at the data behind this headline and survey, the headline should really have read: "Grilled cheese lovers SAY THEY have more sex and are better people, according to a DATING WEBSITE survey for NATIONAL GRILLED CHEESE DAY."
Johnson, a statistician with a Ph.D. in economics from MIT, is the author of the book Everydata: The Misinformation Hidden in the Little Data You Consume Every Day. He joins us at AllAnalytics radio on May 31, 2017 at 2 pm ET/11 am PT.
As the whole world grapples with the concept of "fake news" and looks for insight into what information to trust and not to trust in the growing flood that drowns us every day, Johnson will bring his expertise to bear during this radio show.
We'll examine:
- Common ways data is misrepresented or misinterpreted,
- The limitations of forecasts and predictions,
- How to recognize "Tells" in data and headlines,
- And we'll answer your questions.
UBM All Analytics bloggers Jessica Davis and James Connolly examine the progress that analytics has made in the business world over the past couple of years.
Learn more about the rapid growth of enterprise interest in leveraging data to build out artificial intelligence (AI) and machine learning applications. We discuss the factors that are driving the ever-increasing acceptance by business managers of data-driven decision making. In fact, many of those managers are demanding data in their decision processes.
Learn about the demand for data scientists on the hiring front and the emergence of citizen data science. We examine the state of concepts such as the Internet of Things, predictive analytics and prescriptive analytics that guide managers to the best solutions, and more.
You have the data, you have the tools. What now? Today's voluminous data and highly sophisticated tools can deliver better insights to end users and significant competitive advantages to organizations. What's more, these insights can be embedded in the tools that end users access every day.
But where do you start? How do you begin to decide what to do first? Real world use cases can provide inspiration if you are looking for a place to start, or a way to get unstuck with your analytics program.
This week we are taking a look at some real world use cases that can do just that. From healthcare, to child welfare, to establishing a Center of Excellence, our guest on A2 radio this week has seen them all. Marie Lowman's new book is a collection of case studies of analytics that can be applied across any number of organizations, although these specific case studies are have been implemented in government.
Lowman told me one of her goals with this new book, A Practical Guide to Analytics for Governments: Using Big Data For Good. She said she really to transform the word "analytics" from a word that some users may find intimidating into a word that is part of everyday conversation within organizations. The transformation happens because users become comfortable with the word analytics and understand what analytics can do for them.
Lowman demonstrates this approach -- embedded analytics in the real world -- through this series of case studies. Each case study is written by people who have years of experience in government.
In the forward to the book Lowman writes, "It's not [a book] written by a bunch of PhD statisticians and data scientists." Rather, the book is written by the people with the domain knowledge to turn data into insights. It is written to show how analytics that are embedded inside an organization's processes can radically improve what can be accomplished by the organization. It's written to show how this voluminous data can be used for Good with a capital G.
We're excited to have Lowman join us this week and take us through a few of the many case studies included in her new book, and look forward to hearing about some best practices she can share with us about how to create the type of analytics practice that offers these kinds of results.
Did psycho-demographics help Donald Trump win the US presidency? Several recent articles, including this one, have chronicled the use of this type of consumer profiling and related marketing tactics as the secret sauce behind Trump's winning campaign against rival Hillary Clinton.
But just what is a psycho-demographic profile? This kind of profile looks at what you like and don't like to determine what you might want to buy, who you might want to vote for, and what marketing messages would most motivate you to take a particular action. For instance, instead of segmenting this middle-aged female suburban resident as a soccer mom, it might profile her according to whether she liked the TV show Glee on Facebook and draw conclusions about her politics and consumer preferences based on that.
Michal Kosinski is an assistant professor at Stanford, and a doctor of psychology and holds a Masters degree in psychometrics. He created a method to profile consumers according to their Facebook likes. Recent reports say that a consulting firm unaffiliated with Kosinski, Cambridge Analytica, leveraged that research to deliver the White House victory for Trump. (Kosinski will be our guest on AllAnalytics radio on May 23, and you can register here now to join us for that show.)
UBM Tech, InformationWeek, All Analytics and Interop ITX conducted research into how enterprises are using data analytics, the progress they are making with key analytics and data science concepts, and the benefits they are achieving.
Consultants Lisa Morgan and Jen Underwood join All Analytics Radio for this first glimpse into the survey results.
As more organizations today are looking to leverage data and data insights to gain competitive advantage, a laser-focused spotlight is being shined on that data. Just how good is it? Are we all using the same definitions? Are we paying attention to consumer privacy concerns and regulations?
Can your data stand up to that kind of scrutiny? What if it's your C-level executives who are looking that carefully at your data?
As enterprises have recognized the real value contained in their data, suddenly concepts such as data governance, data quality, and data management have gained more of the respect that they deserve. That's because the data and insights are only valuable when organizations pay attention to these important foundational data programs.
Maybe that's why in our end-of-year quick poll on AllAnalytics.com, the greatest number of you -- 60% -- said that your Top Analytics Priority for 2017 would be "ensuring data quality and data trust."
So how is that going for you so far in 2017? Got questions?
Author and consultant Donna Burbank says that the value organizations have placed on data has conferred a new level of respect for programs such as data management and data governance. Burbank joins AllAnalytics radio as our guest on Thursday, April 27, at 1 pm ET/10am PT to talk about data quality and trust for 2017.
Burbank is an author of the book Data Modeling for the Business: A Handbook for Aligning the Business with IT using High-Level Data Models, and she's also a principal for the consulting firm Global Data Strategy. She'll be talking about Navigating Data Governance in an Era of Global Uncertainty at the upcoming Interop ITX event in Las Vegas in May. But you can get an early preview of some of the things she'll discuss during this A2 radio show.
You can register here for the event, and once you register you can add any questions you have for Burbank in the comments section.
Small analytics projects that can lead to the biggest returns for organizations can also fall by the wayside when they must compete for attention with other priorities inside the business -- for instance other analytics organizations, roles, and data efforts.
That's according to Andrew Wells and Kathy Chiang, authors of the new book Monetizing Your Data. They should know. Both have worked with organizations for decades, helping them shepherd their analytics programs to turn insights into value.
"When we first started tackling this problem, one of the key challenges we noticed was the siloed approach to development and distribution of analytic information," the authors wrote in the preface to their book. These days more self-service efforts are helping improve the negative impact of some of these gaps. But gaps remain, and that means that organizations are not getting the full value out of their analytic investments.
So how do you fix it? Wells and Chiang compiled their approach in their new book, and will be talking about both the book and the approach on AllAnalytics radio on Wednesday, March 29 at 1 pm ET/10 am PT. (Register now for the show: Monetizing Your Data: Turning Insights Into Action.)
They bring their experience with real-world challenges in the field to their approach to analytics.
Wells is CEO of Aspirent, a management consulting firm focused on analytics. He's spent 25 years building analytics solutions for a wide range of companies, from Fortune 500s to small non-profits. Chiang is an established business analytics practitioner with expertise in guided analytics, analytic data mart development, and business planning. She currently serves as VP of business insights at Wunderman Data Management, and has also consulted with Aspirent on several projects with clients including IHG and Coca Cola.
Wells and Chiang's book arrives at a challenging time for analytics programs and projects.
It's funny how one little bit stands out from a complicated work. Consider the one brief scene that you remember from a movie, or maybe one lonely line from a book. Some of us remember, "It was the best of times, it was the worst of times." Others recall, "It is a far, far better thing that I do, than I have ever done." But we might forget everything that happened in the 500-plus pages between those opening and closing lines of Charles Dickens' A Tale of Two Cities.
Granted, the 2011 McKinsey report Big data: The next frontier for innovation, competition, and productivity might not have the literary legs of a Dickens work, but it did set the stage in terms of defining the role and the importance of a data scientist.
Six years later, however, most of us remember just one paragraph, the doomsday prediction of a data science talent gap, which we immediately paired with the Tom Davenport declaration that data scientist is the sexiest job.
It's time to look at just how sexy the data science role turned out to be, and whether the data science talent gap really will put on chill on corporate analytics initiatives in 2018, just a few months from now. All Analytics Radio will host an interview with John Reed, senior executive director for recruiting firm Robert Half Technology, Thursday, March 23, at 1 pm (EDT).
A long time ago when I was a student at a small liberal arts university, the university made a bold move. For the first time ever all incoming freshmen would be required to purchase personal computers. It was a big deal, particularly for a liberal arts school. But this was the revolution that was coming, according to the administrators. This was a tool that would be useful, regardless of your major or field of study or your career. In the future, everyone would use computers. So this new technology would be baked into the curriculum. You would learn to use your PC if you were a computer science major. And you would learn to use other, different tools, if you were a sociology major, or an English major.
Fast forward to a few years from now. When my children head to college, I expect that the cutting edge technology that is baked into the curriculum that everyone uses, regardless of their field of study, will be analytics. We're not quite there yet, but we are getting closer. And it won't be just aspiring analytics professionals who will be learning these skills. Some specialists may learn considerably more about analytics and the underlying principles and tools. But everyone will learn how to apply analytic insights.
Today's Demand for Analytics Pros
Let's get back to today for a moment and not get ahead of ourselves. The recent evolution of analytics and data science education has led to more analytics masters degree programs sprouting up at universities across the country to meet the demand for skilled data professionals. The development follows a 2011 McKinsey report forecasting a shortfall in talent available. Career-minded technology professionals sought out programs to help them join this specialty and gain the best work-life balance and some of the highest salaries available in technology.
The demand has also spawned data science and analytics MOOCs, boot camps, and other programs, but quantitative recruitment specialist Linda Burtch said in a blog post these programs will begin to thin out, eliminating "ineffective or overpriced options that can't deliver on their promise of creating career options. However, the demand for these alternatives will not recede, so the ones that are found to be worth the time and investment will flourish."
Meanwhile, across many organizations, business users seem to be gaining greater analytics literacy. It used to be that businesses would seek that so-called rare data science unicorn -- the technology pro with a background in statistics, coding, and a specific business domain. It's no surprise that finding a person with that combination of skills was no easy task. That's why that vision may be evolving as organizations look to create a team that includes all three skill sets from three different people rather than trying to find one person who possesses them all.
Which brings me back to the idea of baking analytics into all college and university curriculum. Analytics is poised to expand from a discrete department within organizations to something that is embedded in the life force of an organization. Business users from every department will use analytic insights to test their ideas, discover new products and services, and predict customer demand.
As we move to that new generation of business, many more workers will need basic analytics literacy if not actual analytics skills. Analytics can provide feedback on every action taken, whether you are in retail merchandising or government or healthcare.
Dr. Alice Louise Kassens, professor of economics at Roanoke College in Virginia, is blazing the trail on bringing analytics education to a wider audience, beyond the statistics and predictive analytics Masters degree students. Kassens has worked to embed analytics into the undergraduate economics curriculum at this liberal arts school. The training, which includes a class Kassens teaches in econometrics, gives students an edge as they enter the workforce.
Kassens will join me as a guest on AllAnalytics radio on Wednesday, March 8, at 1 pm ET/10 am PT to talk about how analytics education has evolved, and how education has evolved to embrace analytics. Register now to listen in and ask questions during this event as we all get ready to welcome a new generation of analytics pros and analytics literate workers to the workforce.
One of the first forms you are asked to sign when you see a new doctor is a HIPAA form -- a form that tells you your privacy rights under the Health Insurance Portability and Accountability Act of 1996. The idea is that your medical records are "protected health information" that stays private.
Yet your health information can be "de-identified" -- your name, address, social security number, and other identifiers removed -- and then sold as data to other companies.
The collection and sale of prescription information by pharmacies to third parties has been going on since the 1940s. Back then, pharmacies weren't required to de-identify the data, although they often did. The rise of digitally stored information created more opportunity for data collection and sales.
The HIPAA law now requires that information to be de-identified, but it can still be sold. And advances in analytics are now enabling data collection clearinghouses to re-identify medical data to specific consumers. So if you are a patient and your doctor prescribed an antidepressant to you, your pharmacy has sold that information and removed your name and identifying information. But a data clearinghouse company that purchased that information may very well be able to use analytics to figure out who you are. The technology has made it possible.
Join us on All Analytics radio February 21 at 2 pm ET/11 am PT as we explore the history and current practices around medical record data in the wake of privacy laws and evolving analytics tools.
Our guest is an expert on the medical record de-identification and re-identification. Adam Tanner investigated the topic and is author of the new book Our Bodies, Our Data: How Companies Make Billions Selling Our Medical Records. He is also the writer in residence at Harvard University's Institute for Quantitative Social Science.
It's invisible and we take it for granted, but the moment that it's unavailable all the things we rely on to keep us connected, informed, and comfortable disappear. It's power and electricity, and we need it for light, heat, Internet access, and more.
Like everything else, it comes with its costs, monetary and otherwise. Can analytics help?
Is it possible to use analytics to identify ways to cut power use and costs?
Recently my household's power company has been sending me emails and letters showing me how I compare to my neighbors in terms of energy consumption, and the news is not good. In spite of the many LED lightbulbs we've installed in our home to replace incandescent bulbs, our home still consumes more power than our energy-efficient neighbors and most of our other neighbors, too.
In spite of their offer to conduct an energy audit to help me discover why my energy use is higher than that of my neighbors, I confess that I have not followed up with the power company to investigate. Part of me wishes that I could just have the answers right there in the email itself, because a power audit would take up time, and I'm busy and impatient.
But I have my own theories about why our power use is higher. Perhaps it is because I am a telecommuter who uses a computer at home during the day. Perhaps it is because I have two sons who play video games. Maybe it's because I'm one of those people who sometimes leaves the computer on rather than shut it down at night.
Chances are I'll spend time thinking about these things, apply slap-dash fixes to them, see no results, and end up wasting more money and energy than I would have if I'd just had the power company come in to do their audit in the first place.
That's what happens when you rely on your gut without the analytics to back it up. By applying analytics to the problem, you are more likely to identify the efficiencies and apply solutions in a timelier fashion. Analytics can truly save time and money.
Now multiply the energy savings in my home by hundreds or thousands of homes. That could really make a difference. Now consider how much businesses could save through greater energy efficiency. Energy savings means less money burned and going up the chimney and more money going to the bottom line.
To help you learn how to apply analytics to power usage and improve your company's bottom line, we welcome Robert K. Kaufmann to All Analytics radio on Feb. 14 at 1 pm ET/10 am PT. Kaufmann is a professor of Geography and the Environment at Boston University, and separately has founded a company to help businesses use analytics to find energy cost savings. Join us (register here) to find out how to make energy analytics work for you and your organization or company and bring more of that revenue down to the bottom line.
The year 2017 brings multiple challenges to the data arm of many organizations. More data and more types of data are collected than ever. New companies are entering markets and leveraging data to disrupt those markets
But some things stay the same. Data management and data quality tend to fall at the top of the to-do list for organizations that rely on analytics for a strategic advantage. Without that single source of truth, an organization's departments can find themselves working at cross purposes. What's more, they may be wasting time and resources as they focus on the wrong tasks.
This week on All Analytics radio we will focus on what's new in terms of trends in data management, data quality, and analytics. What are the new ways that your organization can improve its processes? Do those change whether you are a digital native or an organization with a years' worth of data?
All Analytics radio is welcoming Karen Lopez to talk about these trends, big and small, this week. Lopez is a senior project manager and architect for the consulting firm InfoAdvisors, whose clients have included many household names such as BJ's Warehouse Club, H&R Block Financial Services, and The National Retail Federation.
More importantly, she is a self-descripted data geek. She is a data architect, speaker, and trainer.
She also was a Datanaut in NASA's 2016 program, serving as a beta tester for data engagements designed by the space administration.
Lopez brings real world experience in the field at multiple organizations to her assessment of data trends for 2017.
How many times have you heard this quoted or paraphrased (usually inaccurately) in the past six years?
By 2018, the United States alone could face a shortage of 140,000 to 190,000 people with deep analytical skills as well as 1.5 million managers and analysts with the know-how to use the analysis of big data to make effective decisions.
Of course, it's from the breakthrough report on data science, published by McKinsey Global Institute back in 2011. But it's just one sentence out of a 156-page paper, one sentence that has been used to support the marketing claims of hundreds of organizations, some of which have only minimal ties to data science.
Talk about soundbite marketing!
McKinsey Global Institute's Michael Chui shares insights from MGI's update on its famed 2011 data science report.
Most of those citations missed the rest of the key points in the report, particularly about the potential opportunities that data science and analytics were going to present to enterprises in the coming years. Hey, we're in those years now. In fact, that 2018 doomsday for data science talent is less than 12 months away. Make sure the underground bunker where you keep your data scientists safe from poaching competitors is well stocked with sugary foods and energy drinks.
A couple weeks ago, All Analytics hosted a discussion among A2 bloggers about the relationship between data and humans. The consensus was that as we move through 2017, we will see more organizations trying to strike a balance when it comes to utilizing data and experience in decision making.
We've seen anecdotal evidence that more people who are on the analytics team or on a business unit leadership team are moving off some firm positions where they sat a couple years ago. You know those people. There are those data analysts and data scientists who are determined to bump "experience" out of the decision equation. Then there are the business managers who adhere to the "I know best" or "No data, no way" philosophy.
In reality, the sane approach to decision making in the business world is to leverage both data and experience. The right mix of the two? Well, "it depends" seems to be the answer. It depends on the company, the issue at hand, the amount of good data available, and probably a dozen other situational factors.
That balance between experience and data is being put to work in retail. It's happening not only during this holiday shopping season but over time, as so many familiar retail names disappear and new retail models emerge.
What's interesting is that the changes in retail -- not just the death of once-huge names and the growth of Amazon and the evolution of omnichannel marketing -- are highly visible. As consumers we see it every day.
Analytics are a driving force in most of the changes, and our guest on All Analytics Radio today knows about the role that analytics play in bringing about change and helping retailers succeed.
Britanny Bullard is a Solutions and Analytical Consultant in the Retail and Consumer Packaged Goods Practice at SAS. So, she knows analytics.
However, she also started her career in retail and advocates for experience, what she refers to as "style." Experience in retail is about identifying good products and pricepoints, recognizing when a hot product has peaked, and understanding which products and prices are a fit for a given market.
Blend that experience with analytics and a retailer can test their intuition with data or use data to uncover new opportunities. If they find the right balance, the retailer has a good chance of achieving success, whether it is carving out a niche with a boutique or holding their own in the big world of Amazon and Walmart.
Join us today at 2 pm (Eastern US) when Bullard joins A2 Radio to discuss her new book, Style & Statistics, and she shares advice not only for retailers but thoughts on how non-retailers can learn from the consumer goods market. Register here.
It's easy to sit back and look at any technology concept and take the curmudgeonly approach, saying, "It'll never work."
Consider Digital Equipment Corp.'s Ken Olsen in 1977: "There is no reason anyone would want a computer in their home.”
How about Ethernet inventor Bob Metcalfe in 1995, "“I predict the Internet will soon go spectacularly supernova and in 1996 catastrophically collapse.”
I'm sure that someone once said we will never need a phone, computer, music machine, camera, and navigation system that fits in our pocket. Hey, that might have been me.
When we talk about the Internet of Things, it's really easy to be a curmudgeon. Some of the IoT uses that people propose are pretty frivolous. I would be proud to be remembered as someone who said we don't need a high tech refrigerator that tells us when to buy milk.
However, we do need some intelligence in our home heating and energy management systems, and the ability to turn our lights on and off from a cell phone -- making home appear occupied when we are away -- isn't a bad idea.
Yet, the IoT in our personal lives is only a thin slice of the IoT concept. The real benefits are in the business world, where that smart refrigerator is a walk-in cooler in a restaurant, letting management know when it's time to reorder meats or which produce should be served first. The data that the home energy system sends back to your utility provides the business intelligence that the company needs to guide others in smart energy use and ties into the power grid. The IoT can route delivery trucks, smooth airline travel, and keep manufacturing lines moving.
That business angle to the IoT is what we will be discussing this week.
On Tuesday, November 29, at 2 pm Eastern time, Why the IoT Matters to Your Business will be the topic on All Analytics Radio, when our guest if Dima Tokar, co-founder and CTO of MachNation, joins us to provide a snapshot of how far we have progressed and where we still have to go with the IoT in business.
Companies are launching IoT initiatives where technology promises to address real business problems. Some of those companies have found success with IoT in a relatively short period of time. Others are still in the planning or exploration stage.
IoT holds promise in operational efficiency, identifying new business opportunities, and improved customer intelligence and support. It also is a technology concept that both serves and connects the enterprise and consumer markets.
Join us for this Internet of Things status report and learn what the IoT can and maybe cannot do for your organization.
Then, next week, on December 6, A2 Radio is on the air again, this time looking at the use of analytics in the retail sector. Brittany Bullard is is author of the new book Style & Statistics, which will be featured at January's National Retail Federation's Big Show in New York.
Bullard, who launched her career in the retail sector, is a Solutions and Analytical Consultant in the Retail and Consumer Packaged Goods Practice at SAS. Her book details how a retail strategy today has to be based on a combination of data analytics and real-world management experience. Have you ever wondered why a product with what seems to be "style" doesn't sell? Often it's because it's the wrong product for the wrong market or at the wrong time. Understanding products, buyers, pricing, timing and other factors calls for that mix of style (market knowledge) and statistics, including big data.
Take a moment and register for both shows now.
Has anyone else noticed what may be a warming trend? No, not global warming, but a warmer reception to data initiatives within the general public and the business community.
It just might be that our acceptance of data as part of our personal and business lives has grown stronger over the past 12 to 18 months. My observations about a stronger embrace of data and analytics are largely anecdotal, and could be blown away by a couple of informative All Analytics articles from the past two weeks. Tricia Aanderud raised some good questions about our love-hate relationship with data after the presidential election in The Election Aftermath: Is Data Dead?. Then Ariella Brown delivered an informative analysis of a KPMG survey on trust in In Data Analytics We Trust? .
Yet, we hear about how interest in data-driven shopping aided by digital assistants and tools such as the Amazon Dash button is growing, and plenty of people are high on the autonomous car concept.
When my faithful home PC went into digital hospice last week I ordered a replacement. My first task with my new laptop was to uncheck the dozens of default "sharing" options that would have let Microsoft, Hewlett-Packard, and other vendors collect data from every keystroke, every application, every website visited, and every location where I used the PC. While I wasn't in a sharing mood, I'm sure that a million or more people buying similar machines this year will gladly feed such data to the companies based on the promise of "a better user experience."
In business, have we passed the tipping point where the majority of managers increasingly trust, or even demand, data as part of the decision process? Plus, more CEOs are recognizing the value of data by having their chief data officers report directly into the corner office. No marketing or product initiative will move forward without some plan to collect performance metrics. The Internet of Things isn't going away.
The All Analytics team is looking at the human-data relationship on Tuesday. Bloggers Lisa Morgan and Pierre DeBois will join me on All Analytics Radio at 2 pm (Eastern) for Will Data and Humans Become Friends in 2017. (If you miss the live broadcast, you can listen on demand using the same link at any time).
We will be discussing whether data has expanded into our lives by stealth, as we rely upon and also share data without consciously recognizing the role that analytics and data play in our decisions, our actions, and our entertainment. Do people who actively avoided data a couple of years ago now at least tolerate it, even if they don't embrace it? What role do the millennials -- I know, I heard that we aren't supposed to call them that any more -- play in our changing attitudes toward data? Will there be a backlash against data because of the election and the never-end data breaches?
When we grade technology initiatives on the basis of real benefits to people there is a natural tendency to concentrate on what the tech delivers today and what level of convenience is offers to us as individuals.
We can claim that we value the smart phone because it provides us with a means for communication and delivery of information in the event of an emergency. That sounds good, but let's admit that the benefits we most care about are the multiple ways it allows us to connect with friends, family, and business associates. Then there are factors such as games and entertainment, and easy access to news and trivia. We simply love our phones.
However, when it comes to the raw potential for delivering real benefits to a massive number of people, you have to think about smart city initiatives.
Consider how even a single project in a smart city strategy can deliver a life-altering benefit to a few million people, even those who aren't using the technology themselves. Think in terms of healthcare, urban transportion, delivery of electrical service, improved housing, and food.
We will be discussing the potential for smart cities on All Analytics Radio this Thursday at 2 pm ET, when my guest will be Tim Herbert, senior vice president of research and market intelligence for the Computing Technology Industry Association (CompTIA).
CompTIA recently completed a report on smart cities, looking at where progress has been made, but also evaluating the interest the municipal officials around the globe have expressed in implementing technologies such as analytics and Internet of Things (IoT) capabilities.
Credit: Pixabay While only a fraction of cities surveyed by CompTIA (11%) have formal IoT initiatives underway at this point, 25% have pilot projects in place, and three out of four have more favorable impressions of the IoT today versus two years ago.
There's plenty of interest in smart cities, and plenty of potential. While some cities, like Dubai, are far along the road in their smart city initiatives, most are still at the stage of virtual groundbreaking; early implementation of a handful of applications.
They might have launched a digital crime fighting initiative, or a smart energy management system, perhaps a traffic optimization application. Not many communities have a broad portfolio of advanced applications.
But we're getting there. Progress isn't just about working through government bureaucracy and tight budgets. It's also about bringing multiple technology concepts together, technologies that are still maturing. The sensors in an IoT network are only part of the job. Smart cities are also being driven by advances in predictive analytics, cloud computing, and mobile devices.
Cities are gradually making the transition from paper-based transactional systems to an environment where they can apply data science to big data and improve delivery of services.
Then there is the massive organizational and operational change that has to take place. As any corporate project manager knows, collecting and analyzing data is only part of the process. Putting that data to work through action by workers or devices in the field is the critical next step. The traffic management system that was designed for yesterday's traffic flows has to adapt and reroute traffic when the unexpected happens today, say, when an accident or weather condition shuts down a key thoroughfare.
Why should you listen to Tim Herbert's interview, and learn more about smart cities? Because you probably live or work in a city, or in the immediate vicinity. More than half of the global population lives in cities today, and that percentage is growing. If you don't live in the city, there's a good chance that you are among the workers who live on the outskirts but help to double the population of a city on any given work day. Or, perhaps you live in a state whose budget and political climate are dramatically impacted by what happens in the big cities. Our cities are important to all of us.
So, join us on All Analytics Radio, Thursday at 2 pm, and learn more about smart cities.
I suppose there are two ways an analytics pro can look at visualization. One could view it as window dressing to satisfy business unit managers who just don't understand data. Or, someone could view visualization as the unveiling of results from their hard, smart work.
I'll guarantee that the first approach will result in a sloppy presentation, leave the business manager confused, and doom the valuable but underutilized data to the trashbin. Worse, a business problem won't get solved, and that business manager is going to think twice before calling on the "numbers guys" in analytics for any future help.
What could possibly go wrong with delivery of data? The business manager could get a 20-tab spreadsheet simply loaded with numbers. An analyst could present a screenshot showing 87 colored, overlapping bubbles of varying sizes that the business manager sees as just a bunch of bubbles. You could display an animated map of the globe, turning and flashing lights representing cities all over the world. With no accompanying explanation, the business manager will simply see a bunch of lights.
In a recent blog, Misunderstood? Try Data Storytelling, Lisa Morgan highlighted the importance of storytelling as a means to deliver data.
Sometimes visualization alone isn't enough. And, sometimes the visualization itself simply is lacking.
So, what makes a good visualization?
Scott Berinato has some ideas, and is ready to share some best practices in visualization. Berinato is a senior editor with the Harvard Business Review and a self-described dataviz geek. He also is author of the book Good Charts: The HBR Guide to Making Smarter, More Persuasive Data Visualizations, in which he shares best practices in dataviz.
Berinato joins All Analytics Radio on Thursday, October 13, at 2 pm EDT to discuss those best practices.
In preparation for the Thursday show, he has shared some examples of poor visualizations that he improved. Check out these three before and after samples that he will reference in his chat.
Before:
After:
Before:
After:
Before:
After:
Register now. Then join us on Thursday for my interview with Scott Berinato, and bring your questions about best practices in visualization.
The McKinsey prediction about a painful data scientist shortage still lives and breathes. Five years after the fact you probably see it referenced on a weekly basis, often by people who couldn't spell "big data" back in 2011.
However, it's easy to forget two aspects of the McKinsey report, Big data: The next frontier for innovation, competition, and productivity, starting with the fact that the dire prediction of data science talent shortages took up just a little over three pages in a 156-page document. The bulk of that report focused on the benefits that big data analytics could bring to various types of organizations.
The second fact is that the report came out in 2011, and plenty has changed in five years. Will the supply of deeply trained and experienced data scientists -- projected by McKinsey to come in at a 150,000 shortfall just in the US by 2018 -- really stagger analytics initiatives? Will the 1.5 million shortfall of managers prepared to use analytics in decision making really happen? The safe answer to both questions: Maybe. I guess we will know for sure in a year and a half.
However, even now the data science/analytics landscape is far different from how it appeared in 2011. Back then, the managers who embraced analytics clearly were exceptions. Today, we aren't close to 100% adoption of analytics in the management ranks, but anecdotal evidence says that we have a lot more managers who accept data as a tool than we did five years ago.
On the data scientist question, a few trends have emerged that address the shortage. It's clear that employers are getting creative in how they fill data science roles. The purist's definition still demands that data scientists have advanced degrees in math, that they know data tools, and that they have real world business experience. Yet, I am encountering more employers who are using alternative approaches, such as pairing a technical expert an experienced business person. That means using two people to do what might have been done by an individual in the ideal data science scenario, but it gets the job done. Others are using crash courses in math and analytics to give mid-career business people the tech skills that they need to be data scientists.
Then there is the longer term approach, where colleges are stepping up with new data science and analytics programs. Some schools are new to the data science space, while even those that have offered data science programs for years have expanded their programs with more advanced offerings, dedicated recruitment, and corporate partnerships. There are an estimated 150 data science programs now being offered by US universities.
With the changes in data science education in mind, Tuesday's All Analytics Radio program features a discussion with Robert J. McGrath, an associate professor and director of graduate programs in analytics and data science at the University of New Hampshire. UNH is one of the schools that has worked to keep its analytics programs relevant in terms of preparing students for the job market.
I'll be speaking with McGrath at 11 am EDT as we discuss themes such as the skillsets data science pros will need moving forward; how educators can balance the need to teach technology versus business and problem solving; schools can work with potential employers to shape curriculum and give students real experience; and the job outlook for the next few graduating classes of analytics professionals.
There are places in the tech space where we cease to stare in amazement about what the tech can do. Instead we whine that the tech can't do more.
Take the case of handwriting recognition, whether it's what we scribble notes onto a tablet or when we scan handwritten text into a PC. We wish that it was smarter, that it recognized more characters and that the text was searchable and shareable.
To be honest, I shouldn't say "we". It turns out that I have zero confidence that any machine -- even all the power of the CIA and NSA computers combined -- could accurately decipher my own handwriting. In fact, my chicken-scratch cursive is a family joke. So, I'm left using the generic "we" in this case.
What you may not have noticed is the progress that handwriting recognition in our everyday technology has made. It has gained intelligence in terms of accuracy, breadth of languages, searchability, and even logic in work such as flow charts.
Behind that improvement is machine learning, a concept that is evolving in how it supports not only handwriting recognition, but applications such as facial recognition and language translation. That machine learning isn't so much about writing code -- yes, you need code -- but about the actual learning process. Basically, developers forcefeed massive numbers of examples into neural network based systems, and those systems start to detect patterns.
Consider a high school English class where your weekend assignment is to read the complete works of Shakespeare and his late 16th century cronies so you understand their messages by Monday morning.
That's how we approach machine learning today.
On Thursday of this week, we will be looking at the parallel evolution of machine learning and handwriting recognition. Gary Baum, vice president of marketing for handwriting recognition software provider MyScript joins All Analytics Radio at 2 pm EDT.
MyScripts software tools allow users to take, edit, and convert handwritten text in real time, and to more intuitively create, interact with, and share content in digital form.
Maybe it's ironic that we're making progress in handwriting recognition at a time when school systems are considering eliminating cursive education.
However, I suspect handwriting -- whether using paper and pen or screen and stylus -- will be with us for a long time. Register now for Thursday's streaming audio interview, How Machine Learning Takes Handwriting Recognition to New Levels, and join us Thursday, August 25, at 2:00 p.m. EDT.
It became apparent pretty earlier in the discussions about the predicted data science talent gap that no single constituency was going to address the shortage of analytics talent on its own.
Potential employers in myriad industries couldn't train enough data scientists on their own, and the universities that had to groom future generations of analytics talent needed some ramp-up time and an understanding of what skills and experiences graduating students would need to find their way into and through the workforce.
If you can't go it alone, you turn to partnerships, and that is just what universities and industry -- particularly the tech sector -- did in working together to launch and shape new data science programs.
To the traditionalists in the education space, "industry partnership" is a concept that has no place on a college campus. College is viewed as an idyllic place where learning should be unencumbered by business relationships. I'm not sure when that view was reality. Maybe it was true 100 years ago, but I know that business interests were in place on campus 50 years ago. Even that long ago, colleges struck deals that brought potential employers on-campus to advise on a curriculum and to recruit students.
The goal is to add some value to that college degree, value that can include being job-ready on day one of employment. So we are seeing colleges develop data science programs designed to meet the needs of the workplace.
Among the tech companies supporting data science education is SAS (sponsor of this site), which will highlight the work of more than a dozen colleges that it has partnered with at the Analytics Experience 2016 conference, taking place in Las Vegas next week.
For a tech company like SAS the benefits of such a partnership include seeing a growing number of young people and career changers embrace analytics and find meaningful work. While some students who are products of such partnerships might end up working for the sponsoring company, most help to fill the need in myriad industries.
Among the colleges participating at Analytics Experience is Oklahoma State University. Goutam Chakraborty is director of the data science programs at Oklahoma State, and has been working to expand data science studies since long before the gloomy predictions of an analytics talent shortage emerged in 2012.
Chakraborty joins All Analytics Radio on Wednesday, Sept. 7, at 2 pm EDT, to discuss the evolution of data science education and how Oklahoma State and other universities are preparing data scientists for the real world of employment.