Cloud-Based Predictive Analytics for Communications: Chris Checco Interview
We recently caught up with Chris Checco, President and Chief Analytics Officer, Razorsight. We were keen to learn more about his background and perspectives on the evolution of cloud analytics, how data science can impact mobile/communication companies and what he is working on now at Razorsight...
Hi Chris, firstly thank you for the interview. Let's start with your background and how you became interested in working with data...
Q - What is your 30 second bio?
A - At the heart of it all, I’m just a data geek with some business skills who thrives to help others make informed decisions. I’ve spent my entire career delivering insights to businesses to help them run their business more effectively, whether it be for selling more products, identifying fraud, or finding hidden cost leakages. I went from database programmer to project manager to management consultant to executive. That path has helped me understand the world of analytics from the ground up – I feel very lucky in that sense.
Q - How did you get interested in working with data?
A - It started when I was a kid eight or nine years old, collecting comic books. While I never read a page of the comic books, I was interested in the economics of them; I would track the price changes of my comics from month-to-month on graph paper to see where the value was, and then try to invest the bits of money I made from cleaning pools, gardening, and delivering newspapers in the comic books with the best future value. Sad, isn’t it?
Q - Very smart! :) ... What was the first data set you remember working with? What did you do with it?
A - From a professional perspective, the first dataset I worked with was a mobile phone customer database back in the late 1980s. I learned how to program on that dataset, tracked and reconciled commissions and made a marketing database for myself to call customers for upgrades and promotions. It was a crude green-screen system, but it ultimately helped drive sales and reduce commissioning leakage.
Q - Was there a specific "aha" moment when you realized the power of data ?
A - While I spent many years futzing with data in many ways, I really learned the power of data when I was working as a programmer at a wireless provider as a contractor. I was working with a statistician for the first time, trying to create some funky dataset they could use for their analysis on a repeatable basis – it was new ground for me. That statistician was thoughtful enough to show me how she was going to use the data, and shared some of her results with me. I was absolutely blown away. While the reality of advanced analytics is complex, the concept was so simple and natural that I immediately started thinking how else we could apply this magic to other parts of the operation. That moment was a turning point in my professional life.
Very interesting background - thanks for sharing! Given your current role, let's talk in more detail about the world of Cloud Analytics…
Q - In research published recently, Gartner analyst Dan Sommer says, "we’re at the cusp of a series of tipping points which will facilitate unprecedented interest and adoption of analytics." Do you agree, and if so, what's driving this uptick?
A - I think Dan is absolutely right, however, I would say that statement has a long tail. Here’s what I mean. While advanced analytics has been around forever, and really hasn't fundamentally changed since the early days of agricultural use, each industry has it’s own adoption curve. If you went to any major credit card provider, insurance provider or oil company, they would have been using these advanced techniques for decades, at least in certain areas of their businesses. Other industries, such as telecommunications, are in the midst of this adoption phase. Others have yet to adopt those technologies. Dan's statement is profound in another way - there is a proliferation of advanced analytics across industries, with less mature industries borrowing the lessons learned from the more mature industries.
Q - What have been some of the main advances that have fueled the move toward cloud based analytics? How do you see this playing out going forward?
A - I think there are three main drivers in the march towards cloud analytics, and believe the pro-cloud trend will continue. Cost – the investment in the infrastructure, maintenance, and migration has plagued companies since we started using computers. The cloud removes much of that pain.
Second is convenience – the ability to stand up a system in minutes, hours, or days, versus months or years, means business can move at the speed of business rather than at the speed of technology integration.
And finally, core competency. Companies are realizing that cloud applications level the playing field in terms of the technology advantage. Companies need to put their energy into delivering the best products within their industry their core competency, not into the latest and greatest super computer, tool, or programming language.
Makes sense! Let's switch gears and talk in more detail about your the Communications and Mobile industry, given that's your current focus...
Q - For years, the communications and media industries focused on Business Intelligence platforms, but with only marginal performance improvements. Now, with the move to Big Data analytics, data science is all the rage. What precipitated this shift?
A - The realization that BI, while still a necessary and beneficial tool, has limitations if used alone. Imagine the first explorers to the New World trekking across the ocean with a rudimentary map, compass, and a vague idea of where to go. That is BI alone – the people asking the questions aren’t armed with solid facts. The explorers, similarly, were limited by their vision, assumptions, and knowledge. Now imagine if those same explorers had todays GPS technologies and knew precisely the route they were taking and the route to get there. It drives intelligence far beyond because it can see the big picture (which couldn’t be seen by the explorers) and the details all at once. It provides an answer to a question – what precise direction do I need to go to get to India, and what is the best way to get there. Big Data Analytics provides the latter ... and that shift, as has happened with many technologies, is increasing in utility and benefits as it gets used more broadly.
Q - What are the biggest areas of opportunity/questions you want to tackle in the $2 trillion global communications market? Do you see one sub-area such as cable or mobile as the most ripe for disruption?
A - While advanced analytics has great potential in almost every aspect of the global communications industry, there are 2 specific areas that are most exciting.
First is the addressable advertising space – this is an area that has had slow traction since its inception a couple of years ago, but it promises to deliver up to 1000% improvement on a significant portion of ad revenues for cable and satellite TV operators. The historical gap was being able to deliver these targeted ads, but that gap has been filled by the latest ad delivery platforms and IP-based set top box technologies. The gap today is a sales challenge – accurately estimating the impressions at a granular level and optimizing sales to that volume.
The second area that provides significant growth potential is data monetization. Many communications providers are taking big bets on this space, but are in uncharted territory in relation to their traditional businesses. The first gap is the business model – how will they actually drive significant revenue streams. Second is the differentiation, since all of the major players are working toward the same goals with similar data availability. Analytics in my mind will both provide vast differentiation and enable them to have more options in terms of their business models. For instance, selling data versus enriching data versus selling insights warrant three different levels of compensation. Analytics enables a higher multiple.
Q - You've said that customer retention continues to be a major issue for cable companies as well as mobile operators, where recent research shows that every quarter some 16% of postpaid customers say they'd like to switch service providers. What's behind these high levels of "churn" and how can Data Science help?
A - The desire to switch carriers is driven by two key factors. First are the reduced barriers to exit and entry. I can now cancel my service and have my termination fee paid by the competition. Second are the alternative options for consumers. While cable TV is a must in most households, the younger generation is using IPTV as their television medium.
Data science can aid organizations in managing retention in several ways. Historically, these organizations wanted to know who was going to churn, and then provide them an offer. Today, data science can help build a better picture of the customer, telling the operators why they are dissatisfied, how much they will be worth, how to best address the customers’ needs, and the best communication options. It’s the difference between throwing a steak at someone who’s hungry or giving them a seven-course, well-balanced meal that meets their palate. The latter provides a richer experience and a more satisfied customer.
Q - Which companies or industries do you think communications companies should emulate with respect to their use of data science – and why?
A - Casinos. The best-in-class casinos have the customer experience down to a science – literally. They have created systems to gather information on their clients and turn those into actionable insights for each and every one. In the pre-data-science days of casinos, only the wealthy players got a personal experience – the 1% of the 1%. Today, at the analytically driven casinos, customers are treated based on their individual profiles. This drives more repeat business, more time in the casinos, more money changing hands, and a better overall customer experience.
Now, 18 months or so ago you joined Razorsight, let's talk more about that...
Q - What guided your decision to join Razorsight in late 2012?
A - The decision was one of the easiest of my careers. After spending many years consulting for companies and government agencies, and designing/building one-off analytic platforms, I knew there had to be a better way. With these customized solutions, I watched customers either struggle to support it themselves or pay huge consulting fees to keep them alive. I also witnessed them struggle to hire and keep the scarce resources required to fully leverage the platforms.
Razorsight provided the ideal opportunity to create a leading edge, cloud-based solution that avoided the one-off approach, avoided the need for these scarce resources, but still provided the power and results of a customized solution. And that is precisely what we’ve built and delivered on top of the mature cloud-platform that Razorsight has perfected over the last dozen years. The ideas that have been culminating and evolving for years have come together in a repeatable, business-user application that doesn’t skimp on anything ... and now you can see why the decision was easy.
Q - Razorsight has attracted some big name clients -- AT&T, Verizon, Comcast, T-Mobile, Dish, Facebook and some 80 other leading brands. What does Razorsight do in the areas of data science and advanced analytics, and why does your work matter to these companies?
A - It's really simple – we help these organizations make better fact-based decisions by providing them insights that they can't create at scale or through brute force. We do it on a repeatable basis at scale delivered through a business-user-friendly interface. They ask the tough questions, and our RazorInsights system provides the answers and insights. While there's no silver bullet to addressing our clients' business needs, our goal is to help them continuously and to cost-effectively "turn the dial" in areas like customer acquisition & retention, cross-selling, operations, and advertising.
At the core, our client base struggles with one thing in relation to decision-making – Time Management. The first part of the Time Management is making timely decisions – business is moving faster and faster, and traditional analysis can't keep pace. The second part of Time Management has to do with where resources are spending their energy – some customers have estimated that they spend roughly 80% of the time munging data, 15% analyzing the data and 5% making decisions. Hence the analysis and decision-making suffer because of the long data preparation time. If they can eliminate all or most of the data munging time, and have a stronger starting point for their analysis, they can put more effort into the advanced analyses and decision-making processes. It only stands to reason that better decisions will be made and at the speed their business requires those decisions.
Q - What makes Razorsight's approach new or unique, and how does it differ from prior attempts to deliver insights on the customers' mindsets "in the moment"?
A - The uniqueness of our approach is that we don't just provide more data or answer a one-dimensional question – rather we surround the core answer with additional critical decision-making facts. If I'm lost and dehydrated, I don't just need to know where water is located, I also need to know if the water is drinkable, the best way to get to the water, the risks in getting there, and how much water there is available. Our solution provides these types of critical answers, but for the communications industry. If we just tell them a customer is likely to churn, there are still many unknowns. So we also tell them why the customer is likely to churn, what they should spend on that customer to retain them, how they should entice the customer to stay, and where they should interact with that customer. We can even tell them the likelihood that the customer will accept a specific offer, and how long they will likely stay after they accept.
Q - What projects have you been working on this year, and why are they interesting to you? What has been the most surprising insight you have found?
A - While I can't give provide specificity on discrete customer insights, as they are highly confidential, I can tell you that there are definitely surprises. Typically, 60% of the insights were known or perceived – while one might state we are simply replicating existing insights (which is partially true) we are also invalidating some of the previous insights and assumptions. This is the truth serum for perceptions and assumptions, as myths and broad statements are dispelled. One of my favorite examples, which I’ve seen across multiple clients, is the "usage trap", that is, the perception that heavy users are highly satisfied and lighter users, less satisfied. Hence they spend enormous marketing dollars trying to drive higher usage rates. What they don't realize is that the pure usage, in many cases, has little correlation to satisfaction level.
Roughly 20% of the insights are things they had an idea about, but couldn’t quite put their finger on it with confidence or accuracy. The remaining 20% of the insights are the true "aha" moments – these are the ones that, many times, go against conventional wisdom and drive positive changes in the way the business thinks, acts, and communicates with customers.
Q - Very interesting!... It's also been said that IT departments are the place data science and analytics go to die because traditional IT is just too slow on the draw. How does Razorsight deal with this challenge -- do you have greater success approaching different departments within a company? What, if anything, would make this easier?
A - We do interact with IT, but our buyers are generally business buyers: Marketing, Sales, Finance, Operations, and Media teams. It's because we are not selling a technology or tool, per se, but rather a solution that directly addresses business issues. IT typically has a set of enormous challenges, namely to keep everything running smoothly while there are a thousand changes in place and external forces driving new behaviors. In this sense, I see IT as the factory – they have to be great at operationalizing massive processes, putting in fail-safe measures, and dealing (calmly) with the daily fires that arise from the unforeseen. he core of advanced analytics simply doesn't fit this mold – it's not rigid system, which once defined, runs forever. It adapts as the business changes.
Think for a moment about a triage unit. General medics handle the easiest cases. This is equivalent in the business world to the group-specific analysts – they need to be smart and very efficient at what they do, but they can do it with the tools they have on their desks on a small scale. The moderate cases require surgeons to perform more complex but routine operations to address the moderately wounded. These take longer, require more training and experience, and have larger impacts if not done correctly. This is your IT group – harder issues, large scale, and repeatable.
Severe cases require ER doctors who can handle a myriad of issues, get them to a point of stabilization, and then pass them to the surgeons. These are your data scientists. Give them a complex question with a little direction and let them go to work. Once they find an answer, the outputs can be operationalized in the IT systems. level.
That makes a lot of sense! Thanks for all the insights. Finally, let's talk a little about the future and share some advice...
Q - What excites you most about rrecent developments in Data Science?
A - The best development in recent Data Science history, in my humble opinion, is the realization that Data Science is a requirement. Companies can no longer afford to live without this staple and realize continued success. Things are getting better all around due to the collateral impacts – companies become more efficient at their core business, they find new ways to expand into new businesses, and the customers get new and better products at more competitive prices.
Q - What does the future of Data Science look like?
A - Well, I don’t have a predictive model for that question (!). However, I hope to see two things: First, I would like to see some revolutionary advances in the scientific aspects – new algorithms and techniques which I believe will emerge in the next decade. Second, I'd hope to see the creative side prevail as much as the scientific aspects – that is, creating new innovative uses for the science, just as folks like Pandora and Match.com have over the last decade.
Q - Any words of wisdom for Data Science students or practitioners starting out?
A - These students and early-life practitioners have such a leg up – I’m a bit jealous. This is because they are getting into Data Science at an extremely exciting time, and will be adding value in ways that some of us more aged folks didn’t have the opportunity to contribute. With that preamble, here’s my advice.
Experience is King – get as much experience as possible in the shortest amount of time, across a broad spectrum of applications.
There is no silver bullet – analytics is a vast science, and it must be realized that an approach that provides an answer today may not work on the same question tomorrow. Use all of the tools at your disposal and please do not rely on one approach. And finally,
Innovate – the uses of data science are still emerging, so take the non-traditional route sometimes and experiment. Taking a chance may pay off in spades.
Chris - Thank you ever so much for your time! Really enjoyed learning more about your background, your perspectives on cloud analytics and what you're working on at Razorsight. Razorsight can be found online at http://www.razorsight.com.
Readers, thanks for joining us!
P.S.If you enjoyed this interview and want to learn more about
- what it takes to become a data scientist
- what skills do I need
- what type of work is currently being done in the field
then check out Data Scientists at Work - a collection of 16 interviews with some the world's most influential and innovative data scientists, who each address all the above and more! :)