Creating a Winning Data and Analytics Strategy

CIO Talk – Hexaware Blogcast: Welcome listeners, this is Sanjog, your host and the topic for conversation is creating a winning data and analytics strategy. So, we are now living in a VUCA world-a world that is volatile, uncertain, complex, and ambiguous. And to sustain and succeed as a business in this environment, a winning data and analytics strategy is key. With the enterprises becoming data companies globally, we see a clear shift in the analytics paradigms today. Yes, you want to infuse AI across the value chain but can you truly ignore the fundamentals of data management? Garbage in is garbage out after all. So how can you create a data and analytics strategy to future proof your organization’s data management landscape? To discuss this I have with me Vaidya J.R. Vaidya is the Senior Vice President and Global Head of Business Intelligence and Analytics practice at Hexaware- a consulting firm focused on transforming IT solutions and solving complex business problems using a combination of human creativity and intellect. Their Three-Pronged Strategy of Automate Everything^®, Cloudify Everything^®, and Transform Customer Experiences^® enables enterprises to fast track into the digital era.

Sanjog: Hello, Vaidya thank you for joining us.

Vaidya: Hi, Sanjog, thank you. Thank you for having me over and it’s indeed my pleasure to be talking to you today.

If you wish to directly listen to the Podcast: Creating a winning data and analytics strategy, Click here.

Sanjog: Great, so Vaidya, what is your view on the biggest imperatives for an enterprise for getting their data and analytics strategy right and who are the key stakeholders involved in this?

Vaidya: Yeah, So if you look at the biggest imperatives Sanjog, as you rightly said, most enterprises are data-driven enterprises. So to be a data-driven enterprise you need to be able to harness data, manage risks, and create revenue-generating opportunities continuously. What you need is a solid business-driven data architecture that can enable customer-centric news and give you insights into customer behavior like their buying patterns and their spending patterns, so that is a business-driven architecture perspective.

The other part is the compliance-driven data architecture, right, and the compliance human data architecture part needs high quality of data and governance around it to avoid costly missteps. Your data is what enables this and pulls together as to how data is sourced, integrated and consumed across the enterprise. When there is explosive data growth, compliance that needs privacy and security challenges on one side, and let’s not forget that we do have on the other side- time to market pressures for the business. Business wanting to capitalize on a fleeting opportunity. So it is imperative to get a winning data strategy in place to be able to accommodate both these conflicting interests, isn’t it?

So if I have to answer the second part of your question which is around who the key stakeholders are, let me start by saying that CXOS or the business sponsors would be the number one stakeholders. Their expectations are very simple- they expect the data and analytics ecosystem to provide them with a Lamborghini kind of experience, which means a great user experience, state-of-the-art features and all of these features operating at great speeds, immaculately to green insights.

The second set of important stakeholders would be the technical folks themselves, as in the data architects, the data analysts and data scientists. But they are working on building a finished Lamborghini for the end-users, they need the right skills, tools and technologies to take various kinds of data through the analytics value chain by, I would say, bringing in extreme automation, and make the process seamless, like the way a Lamborghini would flow through the assembly line.

Third and the most important stakeholders are the customers. They have in fact very high expectations. They want a hassle-free frictionless experience and expect that the requests are not delayed by the limitations of the technical systems in place and also their personal and transactional data remain safe and secure.

So it was imperative that all of these stakeholders come together like in a formula one circuit group, right, operating to win the race in the respective industry and the respective businesses. Hope that clarifies the second part of your question too.

Sanjog: Sure, whether so, yes I agree that enterprises are indeed racing to gather insights but the question is how should an enterprise go about winning this race so that’s the most important question in the minds of all business owners and IT leaders.

Vaidya: Right, so from my experience of engaging with valuable clients across industries globally, I would pick three key dimensions that play a very important part in winning the race.

The first dimension is getting the data fabric ready and right. So let me just go back a little here and explain what the data fabric means. That’s the foundational piece in the whole game. So if you look at what it entails, it starts with data discovery, data acquisition and data consolidation. So why do I say data consolidation- it is because it’s not just about the data within the enterprise happily sitting inside the firewall anymore. It’s more about the external data in all fonts and shape like your IoT data, text image, and video data, so to be able to make the list, you got to be addressing some very key questions like how do you ingest such disparate data sets from multiple sources into your data lake.

So how do you deal with the proliferation of point-to-point data feeds? Obviously, everyone knows they can be complicated and expensive to deal with. How do you deal with disjointed data sets? When I say disjointed data sets I mean the siloed view of business data coming from various applications because that’s the way applications and application databases have been designed and are in existence and that’s the way we’ve seen it evolve, right. So we need to look at data from all angles and see how to address the challenges that I just mentioned.

If you ask me how do you do it I would say that metadata management, you know, is of paramount importance here. Creating that semantic layer right, enabling the data sharing mechanism to publishing and subscribing to data catalogs getting those data APIs for easy consumption of relevant data across lines of business- you know these are some things which are very key in kind of getting your data fabric right.

The second dimension after getting your data fabric- right let’s remember here one thing, it’s one thing to be able to get all the data under the sun but quite another to be able to put it to effective views. So 80% of the effort typically goes into preparing data for any analytics workload and analytics cycle and enablement. Gleaning insights at a speed and to be able to enable the enterprise to be able to glean them at that speed is the second key dimension.

Bringing in high level of automation in the data preparation phase is a very key element because 80 percent of the effort typically goes into preparing data in any analytic cycle, right, to be able to profile the data coming in, identify outliers, identify the anomalies and define the data patterns intelligently and automatically to infer, you know, what data sets are similar enough to be blended together and last, but not the least, how do you handle missing data. But interestingly today all of this can be achieved in time by automating and leveraging ML. So how effectively you infuse ML is another key dimension and apart from automation, the other important factor is how do you, kind of, enable data democratization.

In fact, what I mean by data democratization is how do you enable self-service capabilities within the enterprise? If you want to bring speed into your data, to inside the value chain, enabling self-service leveraging technologies like, you know, NLQ (natural language querying), voice-based querying, auto insights gathering, you know, auto ML for automatic model selection making ML more accessible and usable for business analysts, these are you know, the key aspects of how you enable an enterprise to glean insights at the speed.

The third and very important dimension is data monetization. Once you enable the enterprise to get the data fabric right as I said before, and also enable the organization to glean insights at the speed, next level of maturity is enabling the enterprise for data monetization. Now data monetization involves bringing up data as an asset by estimating the economic value of data for various stakeholders, as in you know the suppliers, the partners, and other consumers. So what it really involves is creating new business models around data as an asset. Hope that answers your question.

Sanjog: Yes, it did thank you. So can you elaborate a bit on what you mean by newer business models in the context of data monetization?

Vaidya: For sure. So let me give some examples. Out here we recently worked with one of our clients in the pharmaceutical world that collects de-identified health information, I would say from a vast array of sources, and then punching the data, transforming that data and gaining insights. What they do is they sell that information in the secondary market. Buyers could be, you know, former firms that may be intent on gleaning insights to refine their marketing strategies, or attempting to figure out where they invoice next, so it’s like this right.  Our client has agreements with more than 120000 sources of data around the world to get anonymous patient data it collects from providers, payers, and even pharmacies. So we worked with them to enable them to monetize the data through the three dimensions I just described. Another example that I can think of is from the shipping industry. You know this client of ours is involved in providing port call services to ships across the globe. When I say port call services I mean the entire range of those services from underpinning storage even into the wall and free agent and also maintenance of the ships. This client over the years has accumulated a rich repository of some very unique data sets with an interesting, you know, problem to solve in terms of being able to monetize the data. So we are working with them to enable them to build a data platform and hence you know take them to the maturity level of monetizing the data that they have.

I can also think of another very recent example here and this is from the insurance industry. We are working with these clients again, you know, to be able to help them, you know, monetize their data around claims and loss ratios, you know, in the context of water flooding. This involves crunching a lot of data and ultimately the outcome would be to give out location safety scores for properties and you know the real estate industry is a big-time buyer of the data and business.

So to be able to win the race that is based on data I would think these are the three dimensions that an enterprise needs to kind of master and get to the level of being able to monetize their data for all the stakeholders. I hope that answers your question on you know monetizing data.

Sanjog: Yes it did, thank you. So Vaidya, you did mention that the devil lies in the details, so what are some of the execution considerations for setting a winning data and analytics strategy.

Vaidya: Well, another great question Sanjog. It starts at the very top. I would say executive sponsorship that understands and supports the strategic vision and more importantly understands that there may be pain at first and that you have to continue to drive through it. Since we are talking of execution here. We all understand any strategy is only as good as execution. Let’s see some of the key you know factors, success strategies, right.

The first thing I would say is preparing the organization for the people aspects of this change. I would rate it as highly crucial as one embarks on the transformation journey, the existing roles will change, some roles that may go away and then some new roles that will have to be needed, and of course, with a new set of skills. So let me give you a recent example where we were working with a CIO and the team in the e-commerce space, and they wanted to move the complete data warehouse ecosystem to cloud. Surprisingly, when we kind of got started with the engagement lot of internal resistance came right from the team and we figured out that the team was only used to handle a set of technologies required to support the legacy database. So the team had not been exposed to the emerging technologies but this move would warrant a compelling new set of skills. So we were early to recognize that and we worked with them to restore their entire team. These are some of the people aspects that we need to be really, you know, addressing as we kind of embark on the journey and on the process trend it’s how well you adapt to a giant methodology, DevOps, and continuous delivery methodology. How we engage with the business right from the start and show them value continuously, every sprint and that’s going to instill the world of confidence in them and that will also do a lot of good for the other stakeholders involved in it.

So we saw the people front, the process front, last but not the least is the technology front. There’s this problem of plenty for our customers. Every single customer that I work with, they say, every day some or the other technology vendor meets them to showcase their technology as a new coolest thing in the world and that creates a lot of confusion in the minds of our customers. Sometimes we find our customers do not have the wherewithal and they require an understanding of the emerging technology skills to be able to evaluate those media, opportunities, and technologies out there in the market and that’s where we come in, and say leave it to us as that’s what we do for a living. So these are some of the most important execution configurations that, you know, I would suggest that any enterprise embarking on this data should take care of.

Sanjog: So Vaidya, how would you recommend an enterprise go about selecting the consulting partner for such an effort because this is a monumental effort and you may need partners and if they choose your firm how your playbook would read.

Vaidya: Oh, this is a very important question. I’m glad you brought it up. The right partner can put enterprises in the fast lane of the transformation journey right. The first thing that you know you should look for in your partner is the business and data experts. The partners should bring in strong expertise in the complete data while being generous they should be able to drive the data strategy in architecture and I say data architecture I mean from data discovery to all the way through data monetization efforts identifying the right use cases for the stakeholders to help them monetize their data. So automation-enabled frameworks and actual aiders which would help our clients jump-start the whole process of getting the current level of data maturity and complexity and recommending the way forward to be a very key enabler and they should look for this capability in this part. Partners should also bring in subject matter expertise in customer industry segment and area of business.

The second thing that I can think of is speed and reliability from the partners. They should be able to select partners that will have automation throughout the data life cycle as their core value proposition. Recently, one of our large clients in the mortgage space wanted to move out of an appliance-based legacy data warehouse with our secret system and the appliance came with a drop-dead, you know, support to stop the day that deadline was, you know, about a few days. So, when the appliance support reached the end and were also very expensive to, kind of you know, continue operating that they were looking for a trusted partner who could quickly help them move to cloud at speed and in a highly reliable manner without, you know that’s the key here, without disrupting the business and we were there for them. So speed and reliability is another very key aspect in selecting the right partner and I would say a partner that can engage with the enterprise and kind of co-create value to drive business outcomes would be of paramount importance.

So what I’m trying to say here is it’s just not about the technology we have. With all the ability to put the right technologies in place to drive the desired business outcomes as part of the digital transformation- that will be a key differentiator and that will be the essence of our playbook. You asked about what would be our playbook- we would say write technologies to drive the desired business outcome, how do we engage with the various stakeholder’s rhythm and without the enterprise to co-create value working very closely and transparently with our customers. That is what I would recommend.

Sanjog: Finally, how about this question which all enterprises are having today, that is, with COVID 19 the Coronavirus pandemic, how does my data and analytics strategy change?

Vaidya: Oh yes, the very relevant question for the now and the way forward. So this question of you reminds me of a recent meme I saw in the social media. So this question was put to the CEO as to who decided your digital strategy and it had multiple, you know, options with you the CEO, CFO, the CIO or the CTO and the last answer was COVID and invariably everyone picked COVID. So that’s how important the digital transformation is going to be. The post Covid data and digital transformation, in my mind, will be job number one in everybody’s agenda across industries like how we started our conversation discussing on how we succeed in a VUCA world.

You know the business plans are always going to be in flux. Enterprises will move from multi-year technology planning and executing large programs to really depending and relying on the test and learn as you go kind of models meaning very very continuous test and learn approach. You’ll create MVPs (minimum viable products), work closely with the businesses to see the value and that’s how the world is going to move forward. As for digital initiatives, I would think business and IT strategies are becoming synonymous and given the VUCA factor, organizations will have to change directions as demand changes, and hence everybody is going to look to build an adaptive enterprise, more agile, you know, and more nimble like never before.

A few quick things that I can think of as we converse is we’ll see more and more touchless and immersive customer experiences becoming the norm and that will enable, you know, enterprises to become adaptive and those immersive customer experiences in in turn will be enabled by technologies that help in digital leapfrogging, and also helps in ensuring resilience on a cloud.

So we will see enterprises engaging the anywhere employees and I’m seeing the enterprise is already doing that during COVID and on the other side of COVID. It’s going to be, you know, they’re not engaging with anywhere employees but all of these initiatives, right from the touchless immersive customer experience to engaging with anywhere employees are going to be funded by automation-led sustainable cost takeouts. Those initiatives are going to be, you know, coming out, you know, ahead of the rest of the initiatives for any enterprise. So data strategies, data architectures will be very critical as they are the backbone that will enable anybody to build an adaptive enterprise as we come out the other side of the COVID and an adaptive enterprise architecture, this is what is going to differentiate between the leaders and the legacy, that is my view, you know, going forward.

Sanjog: Once again thank you Vaidya so much for sharing your thoughts and insights about creating a winning data and analytics strategy