Evolve by the Data: Collect, Analyze, Visualize, Actualize
There’s data, and then there’s metadata — not to be confused with master data or big data. All of this information is at your fingertips, but what do you do with it? In this episode, we talk with BlueMetal data expert Raheel Retiwalla to discuss data analytics tools, data strategy and making the most of all of your data points.
Note: Complete audio transcript found after author info.
Episode 5 – Evolve by the Data: Collect, Analyze, Visualize, Actualize
Published July 20, 2016
Announcer: You're listening Technomics. Connecting you to insights on digital transformation and the marketplace, with your hosts: Robyn Itule and Jeremy Nelson. The hosts' opinions are their own. Enjoy the show!
Robyn Itule: Jeremy there are few things bigger than big data.
Jeremy Nelson: I feel like this is a set up for a joke.
Robyn Itule: No, I have no punch line to this. Its extremely huge and its a big opportunity, its a big business problem to tackle, and it requires huge amounts of infrastructure in order to make sense of it all.
Jeremy Nelson: Oh absolutely, and even beyond that, just a lot of really smart people to know how to sift though the massive amounts of information collected from the endless amount of endpoints, users, and sensors that we're putting out there these days.
Robyn Itule: And the whole point of it is to run smarter, right? So all of that information really has to be aimed at a singe goal, a strategic objective. And then the flip side of that coin is you can use all of that information to help you make those strategic decisions. But in the view of the customer, every single piece of that data had better improve their experience.
Jeremy Nelson: Absolutely, its the sacrifice. If you're willing to give up that information, it better benefit you somehow.
Robyn Itule: And there's a million different ways you can slice and dice it. And we have a great conversation coming up about the possibilities that exist here.
Jeremy Nelson: Absolutely, who better to talk to about big data, than someone who specializes in big data. We're going to take a short break and when we come back, we will continue this conversation on big data with a very special guest.
Robyn Itule: And now a word from our sponsor. Engaging your customers with modern technology requires a modern platform and infrastructure. Enter Microsoft Azure.
Jeremy Nelson: All right. I have to stop you right there, Robyn. Is it AZ-ure or Az-URE.
Robyn Itule: You know we had this conversation in Season 1. And I believe what we landed on is, it is about the color, which is in fact AZ-ure.
Jeremy Nelson: Awesome, I’m glad that’s finally put to rest.
Robyn Itule: Although now I’ve said it so many times I feel like I’m saying it weird.
Jeremy Nelson: It’s like spoon.
Robyn Itule: We digress. The important message that you need to take away from this sponsorship is that with Azure, you can build faster, move quicker, and empower your organization to reach new heights. If you contact an Insight specialist, they can help you find out how implementing Azure will open up a new realm of possibility.
Robyn Itule: Thank you for staying with Technomics. I'm Robyn Ituley and today I have a special guest that comes from the BlueMetal wing of Insight. And his name is Raheel Retiwalla. Raheel is a really deeply experienced data scientist, Internet of Things scientist, information technologies and systems guru. And so we've brought him in to continue this discussion around evolving by the data. Using that information to collect, analyze, visualize, and actualize for the business. Welcome to the show Raheel.
Raheel Retiwalla: Thank you, it's great to be here.
Robyn Itule: Did I do you justice on that welcome? I like to position myself as the peanut gallery because it is safe. But you are really in this fascinating area of technology right now. Can you expand on what you've been up to the last few years and how big data has maybe changed your career?
Raheel Retiwalla:Sure, its a significant amount of innovation happening in the data space over the last five years across many different capabilities. If you think about how you acquire data, what you do with data, how you use data to understand and predict your business outcomes. At every single element of how data has been used or collected has been impacted over the last five years. And more importantly, as somebody who has spent about 10 years at Microsoft working on business intelligence and data warehousing in the early and mid 2000s, its been fascinating to see the rapid rise of open source innovation specifically in this space. And personally, that has impacted my career going from traditional enterprise software, commercial software, to more of an open-source software model as well.
Robyn Itule: We hear a lot about big data. Its been a major buzzword in the industry for a while now. To what degree do you think that it is just a buzzword and what is it really, how is it really helping businesses?
Raheel Retiwalla:To really answer that question, let's evaluate one of the driving factors where the need for big data or anything around data today is required. And if you think about what those driving factors are, it really starts with changing customer expectations. Customers are used to now receiving real time insights, real time information, expect instant gratification, look for personalized engagement. And when you think about how larger enterprises have been around for 20, 30, 40, 50 years. How do you pivot and respond to those changing dynamics. That's the driving factor. How do you convert that into net new revenue opportunities? That's the driving factor. How do you make your operating model and how do you actually get work done to serve the new products and services to your customers? How does that get impacted? That's where the use of big data is really applicable. Its the force and function to ensure that you as an organization have the ability to respond to these changing dynamics. So its very real, and its very, very critical in my opinion.
Robyn Itule: Critical but also huge right? The estimates that are out there are something to the effect of 2.5 quintillion bytes of data every single day that we're creating as a very broad group of consumers. Which by the way is the equivalent of stacking books from the sun to Pluto and back again. But how do you operationalize out of so much data? And then optimize it right? How do you get to a granular action item for the business to improve a customer experience out of that kind of vastness?
Raheel Retiwalla:It really goes back to thinking about what are the factors that are influencing your business right now. And aligning the use of data to those particular initiatives. For example, you may be an enterprise that is in the middle of a business transformation. It may mean that you're thinking about going from a hardware company to a more software and services company. You're thinking about going from a brick and mortar to a digital business. You're thinking about moving away from Cappix [assumed] based operating business model to a more providing a subscription based services and software. Now as those transformations are being thought of, it is critical that you align the use of that data and how you operationalize that data to those specific initiatives. Another example is around process innovation. How do you improve the things you build, the services you deliver, the way you support your customers? All of those have an impact on when and what data needs to be operationalized. So it really has to be driven from top down to understand, "What are the strategies for my organization this year. And how I'm going to actually use and operationalize data to meet those."
Robyn Itule: So based on that answer, in your opinion, which of these questions matters more? What do we want to do because of the data? Or, what technology do I need to solve this?
Raheel Retiwalla:Well your first question I think is more about assuming that data is going to tell you what you should do. And that's not necessarily the case. The data is not necessarily going to tell you what you should do. Its going to be more that your people have gut reactions. They've been in the industry for a long time. They understand the customers, they understand the channel. The understand the processes that it takes to actually get a product or a service out to the market. So we want to make sure that that knowledge is incorporated in the method in how you're using data. And that's what really what you want from the data is to take that organizational knowledge, embed it inside the data itself and then use that to actually influence outcomes that you're looking for. As far as what technology do I need to solve this, I think that is not as dire of a situation in the sense that you don't necessarily need to worry that technology is of a bigger concern than the actual outcomes from the data. I feel that innovation and technology is so rapid and the vendors, the traditional vendors and the open source vendors are doing actually a pretty incredible job of making it easier and easier and more integrated to actually build and bring new capabilities, or data specific capabilities into existing products and services.
Robyn Itule: So basically, the technology becomes the how. It's how you are able to bring that information to you. The data is the what. But we still leave businesses in the position of ascertaining the why out of all that data. Are there strategies and tools out there that businesses are looking to in order to kind of get to the center of the golden circle on that?
Raheel Retiwalla:You know that's one of those holy grail things right? Where you expect the data to tell you why things are happening and that's, even in a statistical field for a very long time, its been quite a challenge to balance the causal versus identifying factors that influence. So basically saying that I can tell you what factors influence a certain outcome to happen, with some predictive and some higher level probability. Versus telling you why exactly some things happen. So where we are today is that technology has the advancements in advanced analytics, and technologies around machine learning, are making it easier and easier for us to get a better sense of all the different factors that influence certain outcomes. So if you're interested in knowing what influences customers to turn, well there are models that can help you determine what those factors are. But can they tell you exactly why? Probably not. But you're better off at least getting some indication of what those factors are. And what's interesting about factors is factors can be internal and external to your business. You know I can tell you maybe a couple years ago I was working with a very large telecom company and we were trying to understand what factors influenced customer experience as far as streaming video was concerned on a 4G or LTE network. And there are many factors as you can imagine why if you're driving down the highway, and you're in the passenger seat hopefully and you're kind of going through your Facebook feed, and as you go through it somebody shared a video and the video starts playing immediately, and its buffering and its not giving you the fluidity in the streaming so you're getting upset. But you know there are many factors that influence that. The question is what are those factors? Are they in control of that telecom provider or are they out of control? And as a telecom provider, their first inclination is to assume that its internal factors that influence that because they're controllable. But rather, they are actually external factors. It could be that because it was 5:30 and everyone happened to be stuck in a traffic jam. And it happened to be that most of the people were customers of this particular telecom. It could also be because there was a big event going on close by. So there are different factors. The problem is that enterprises have never been able to combine the internal factors and the external factors to understand truly what causes and influences things to happen. And that's where the technology has gained a lot of momentum over the last few years making that more and more easy to do.
Robyn Itule:When we come back, we'll have a chance to speak much more in-depth about big data. You will not want to miss this.
Robyn Itule:Jeremy do you ever feel like you have a hard time figuring out what the most important technology news is on a weekly basis?
Jeremy Nelson: Always, I never seem to know where to go.
Robyn Itule:There's so much of it.
Jeremy Nelson: So much.
Robyn Itule:Fortunately for people like you, there is The Script which is one of our newsletters that is about the news, best practices, and current trends in technology. We have scoured the web and we've looked for only the most important things. So if like Jeremy, you need a concise, valuable, way to get the most out of the technology headlines, visit www.insight.com to subscribe to The Script. They're IT headlines worth repeating.
Robyn Itule:So let me expand on a piece of the example that you just gave. Because I deeply appreciate your sentiment about making sure you're checking your Facebook feed from the passenger seat. Its an important point, please don't social media and drive. But let's talk about how complex that potentially gets when we're talking about the expansion of the Internet of Things and the ambient bandwidth around us that is required. Take for example, a day in the not too distant future when we are all the passengers in the cars that are being driven by themselves. How does the Internet of Things complicate the internal and external controllable factors with something like this?
Raheel Retiwalla:That's a great question. If you look at the Internet of Things, the way I look at it and the way we look at it as working with many customers around Internet of Things based products and services that they're looking to bring to market. Is really that the data coming out of smart devices is just another source of information that can tell you and influence the factors that I was talking about. Its just another source of input. And that's the magnificent thing about the Internet of Things. Its just the source of that data is much more real time, its coming in very fast, it is very low latency meaning its based on events, something happens and data is here. I mean I can give you many, many examples around that. As you said, let's take that driving example. I can already imagine autonomous driving is a basis of a lot of conversation today. But if you can imagine what the decisions the car has to take in order to drive itself, there's tons of them that are happening. How close the car in front of you is. What's going on to the left, the right of the car. What's happening inside the car. How comfortable you are or not. What you're doing in the car. Even the car seats have sensors that can inform the car about your comfort level and your stress level and all of those things. So these are all inputs that influence outcomes. And in this case, the outcome is the car driving itself and doing so safely. Whereas in a business sense, a device that tells me or gives me a measure of my smart building for example, allows me to save money in real, hard dollars around my operation efficiencies. Or if I am a business oriented and providing a product or service in healthcare for example or even in public safety for example. I've spent some time at Motorola Solutions and we did a lot of work around public safety and smart policing. And you think about just capturing all kinds of data from these different devices. Whether its the network itself, whether its the radio that is connected to the police officer for example, or it could be the video stream that is being captured, or it could be the car. Whether its the police car in this case or its the fire arm. I mean all of these things have sensors now built in to understand when they are used, how they are used, what was the condition when it was being used, all of that stuff is available and can be used to actually make outcomes better. In this case, a better policing experience. In the healthcare space, a healthier, more informed person.
Robyn Itule:With that in mind, we're really talking about having a lot of actionable insights from the data. And what percentage of businesses do you think are truly ready to make those conclusions right? Once you identify a trend in your data or something that is really, really meaningful, how do you take that data and leverage it to run smarter?
Raheel Retiwalla:Yeah, the one way of doing that is to think about, it comes back to the applications. When you talk about actionable, who's doing the action? At the end of the day, its most probably a human being that's doing the action right? I mean either your customer, your sales reps are talking to a customer. It is a customer support person speaking with a customer that has a problem. It is your field person going out there and servicing something for a customer. At the end of the day, it is the interactions and the touch points that you have with your customer that are where the actionable piece comes out in the actionable insight. So the question is, how are those people that are interfacing your customer, how are they doing that today. Most probably, they're doing that by some type of an application right? So whether it may be, I am servicing some asset in the field, I'm going to have some kind of mobile device, I'm going to go there and I'll record what I did. Or I'm a customer support person and I probably have a web-based application. I'm looking up what the customer history is, etc. The idea of actionable insight is to actually embed intelligence in those applications so that the outcomes of that application itself is telling that person what is the next best action to do. And that's the difference. Its embedding these static applications that are just reading data and presenting information, to now reading data, understanding and interpreting that data, understanding how best to react to that data, and presenting those options to that person. And that's where the advanced analytics comes into play and that's how you should think about operationalizing these insights.
Robyn Itule: So really, the operationalization that you're talking about is moving to those intelligent applications that eliminate the human analysis from some of that process so that you go from just big data to real time insights that are delivering value to customers.
Raheel Retiwalla:You are absolutely correct. And that's exactly what we want to do, we want to have intelligent applications. But let's take it one step further. It's great that you brought up intelligent application. You mentioned the term human interpretation. So let's look at the process. First of all I need to have some data as a user, as a person. I view that data, I interpret what that data is saying, then I analyze that and then I act upon it. These are all things that are connected and how kind of things work. As far as intelligent applications go, yes you want to embed the ability to reduce the need for the humans to actually interpret what the data is saying because you can take advantage of algorithms to look at all the different internal, external factors that may have influenced that thing you're doing and tell you and interpret it for you. So that reduces the time. But the next thing is, taking the action part. So is there a way for me to reduce taking actions from a human as well? In many cases not. But in some cases as far as let's say, Internet of Things goes, there may be. For example, if its a farming example, can I turn off the water if its raining? I can. Does a human have to go and turn the water off? No, the device has the ability to actually turn it on or off. Smart hose. Same thing. You know, we actually worked on a really interesting scenario where you have people that can't move around easily or have limitations, or have dementia or things like that. But they're in home, they live lives and the home or the kitchen can be smart about how it can satisfy the needs for those folks. Or it could be, as I mentioned in farming, think about it in healthcare as well. How can a device stop performing a certain function or start performing a certain function based on the embedded intelligence? That's where we get into the power of embedded advanced analytics.
Robyn Itule: There's a lot of predictions out there that you're going to see advances around delivering technology stacks that are integrating the key components. Public safety and healthcare, I think that those just seem to be such topical areas where it seems like there's some amazing solutions. What are you seeing in that space?
Raheel Retiwalla:In healthcare, there's many use cases and scenarios we can go into. But, one area that I feel is very exciting is how devices that improve people's lives are enabling the Internet of Things scenario. So for example, we all are used to just our basic phones giving us better information about how we move, when we move, when we sleep, how much we sleep, how much we sit, all of those things. But now we take that one step further where you can improve people's lives who can't move because of certain specific reasons. We have devices that can give you that and understand when you need to move and do that for you. Or when you need to breathe. Or if you're dementia and forget where you are. Being able to actually provide in your home, the ability to know whether you are where you're supposed to be. And when people fall, another area is when people get older, they may fall. And so the question is well, how do we reduce that. And then finally there's areas of your health itself. How do you do proactive health management, how do you involve your primary care and predicting and providing insights into what your treatment plan should be. Those are areas that are just scratching the surface. And I think Internet of Things and how we use and operationalize data are really connected to the heart of all that.
Robyn Itule: Yeah that's such a great example with healthcare because it really becomes the intersection of the healthcare environment. Really important, impactful work that's being done by a lot of people in those industries right now.
Raheel Retiwalla: That's right, absolutely.
Robyn Itule: So let's close on a question here that maybe takes us right back to square one. For a lot of people who probably don't feel like they're really ready to go and tackle this big data thing, if you are in a position where you wanna make some big decisions that are informed by the kind of data that your company is currently capturing, what is the first question that you ask?
Raheel Retiwalla: That's a very good question and I know for a fact that a lot of customers that we work with are in very similar zones. They're trying to figure out what and how. And the thing that we advise them and work with them is first and foremost on the strategy. What is your strategic goals? Is your strategic goal to produce cost, improve your processes, create operation efficiencies? Is it to generate netnew revenue? And really the use of that data has to be aligned to that particular strategy. When you think about, same thing as IOT as well which is really connected to the big data movement as well, its all about strategy. What is the ROI of me doing that? And I think a lot of times, people stumble because they start off with a proof of concept to see if this would work or that will work. And sometimes it does and sometimes it doesn't but there is no real business case around that. There is no specific ROIs around that. There is no understanding of once we actually get something from the data, how will that person on the front line actually use it? What is the application that we'll use. How will we actually enhance customer experience? What elements of the customer journey does this improve? So it really isn't necessarily about technology. It really is more about ensuring that there is a business justification and a reason behind investing in this. And that will give you the focus and not only what data should you use, but what technology platform should be used, whether its the cloud, or on the premise, or a hybrid. All of that stuff will just work out itself because we know that by doing this, we will actually deliver value to the end person who is going to benefit from dealing with your customer rep. And I think that's where I would encourage people to start.
Robyn Itule: Raheel I want to thank you so much for all of your wonderful thoughts and anecdotes today. I think our listeners have a better understanding of how big data is really a business function and where they can maybe start having conversations about their strategy to ultimately improve that customer experience. I certainly look forward to it as a relatively heavy app user myself. I hope you had fun having a conversation with us.
Raheel Retiwalla: This was great, well thank you so much.
Robyn Itule: Well we really look forward to having you back again soon.
Raheel Retiwalla: Same here, thank you so much. Thanks everyone.
Robyn Itule: I feel like this was one of the more profound conversations that I've had with someone about technology in a while. Because the big data topic just gets to be so transcendent.
Jeremy Nelson: If you don't live it every day, the one thing I took from the conversation is that, I don't want to use big again, but big data is a big decision. It’s something that requires a whole level of planning and understanding and support that most organizations don't think about when they get into that area.
Robyn Itule: Well and there's this ever expanding universe of the way analytics can be digested in an easier and easier fashion. And yet the number of factors that are proving inputs to all of that is growing at the same rate.
Jeremy Nelson: Or faster.
Robyn Itule: Absolutely and it comes back to something that we've really seen as a thread throughout the last few conversations on Technomics which is, you have to be aimed at a specific goal. And you have to adopt a specific strategy in order to really benefit from the opportunities that exist here.
Jeremy Nelson: Absolutely and that strategy and that plan has to be focused around your client and what they want.
Robyn Itule: And what's the ROI, right? It’s not just about, dare I even say this on the customer engagement season of Technomics, but it’s not just about the customer. It’s about what the business is doing to serve the customer. And trying to figure out where their ROI is. The data may show you that there's a demand for something. The business may be able to look a layer deeper and see that really that is something else entirely. It can just unveil a whole other area of work that you don't even know you should be doing.
Jeremy Nelson: Now that's a great point. I think it all comes down to maybe who the end customer is of the information. But like you alluded to, at the end of the day, it always has to come back to making sure that your business is running smarter and supporting your client in a way that they want that support.
Robyn Itule: Thanks for listening to Technomics. If you want to find more episodes, you can download the podcasts from iTunes, Google, or your favorite podcast provider. And for more stories on intelligent technology, visit www.insight.com.