Home Tags Posts tagged with "big data analytics"

big data analytics


Has big data had its moment in the sun? Interestingly, a few years ago, (that is, when the concept was making headlines) this query would have elicited a resounding NO. Today, though, the response isn’t that clear-cut.

Here’s why-on one hand, industry data suggests that companies are looking for new ways to leverage analytics to the fullest. The bottom-line? Well, boosting their bottom-lines, actually (pardon the pun). Essentially, the idea is to highlight a shift towards leveraging big data smartly and not just for the sake of it. In other words, don’t just analyze large data reservoirs; unearth those vital bits of information that make the difference to your business. To add to this, research firm IDC has predicted that revenue from the sales of big data and analytics applications, tools, and services will increase by more than 50 per cent, from nearly $122 billion in 2015 to more than $187 billion in 2019. Interestingly so, services will account for more than half of all revenue by 2019. It means most companies will be using big data technology in conjunction with expert knowledge. That’s that, then.

Now, permit me to play devil’s advocate. Why get restricted by terms like big data? Isn’t data, by itself, enough, given the role it plays in our lives today? Sample this, for instance, the Cisco Visual Networking Index: Forecasts and Methodology, 2016-2021 has stated that in 2016, global IP traffic was a staggering 1.2 ZB per year. By 2021, this is expected to touch 3.3 ZB per year. In other words, we live in a data-rich age, and how!

And this is the interesting bit-companies haven’t given up on the concept of mining vast sets of data for insights. That’s still very much prevalent. So much so that one doesn’t need to fix the “big data” label on it. One just calls it data. It is now taken for granted that the huge data at our disposal will glean actionable insights.

But, there is a catch, though. The technology of collecting this vast data repository isn’t enough. What is to be done with it? What applications can be developed from this information? Actually, to rephrase, which applications can best leverage this data?

And here’s where I bring in the customer experience management angle. After all, what’s a blog on big data analytics without it? As mentioned before, the technology for its own sake simply isn’t good enough anymore. If it were, would we actually be reading articles heralding the demise of big data? And there are a fair few, I must say! On that note, even deciding to use the data at one’s disposal smartly isn’t enough. One must ask the following questions:

What is the quality of insights one obtains from the data?

What is one trying to achieve from this data?

Does one have the appropriate data to address the above?

Will the insights obtained be beneficial for one’s business?

How does one leverage these insights for this purpose?

Net, net, the idea of this blog isn’t to say that big data is dead. Nor is it to applaud its achievements. The idea is to say that big data for the sake of big data simply isn’t enough. The idea is to get to the why and how of existing data. It is learning how and which applications can be leveraged to even reach the why’s and how’s. It is about accepting that big data can (and does) have its pitfalls.

It may just be about smarter data. Or not. Watch this space for more.



January 24, 2018 0 comment
0 Facebook Twitter Tumblr Pinterest

It would be a bit of an understatement to say that we live in a data-rich age. To illustrate, sample the following data released by Cisco, in a whitepaper titled, Cisco Visual Networking Index Forecast and Methodology, 2015-2020,

  • Annual global IP traffic will surpass the zettabyte (ZB) threshold in 2016, to reach 2.3 ZB by 2020.
  • Global IP traffic will increase nearly threefold over the next five years, and will have increased nearly 100-fold between 2005 and 2020.
  • Smartphone-based traffic will exceed PC traffic by 2020, to account for 30 per cent of total IP traffic. (PC traffic will account for 29 per cent).
  • Globally, mobile data traffic will increase eight-fold between 2015 and 2020.
  • The number of devices connected to IP networks will be three times as high as the global population in 2020.There will be 3.4 networked devices per capita by 2020, up from 2.2 networked devices per capita in 2015.

Net, net, it would be safe to assume that the flow of data-based traffic isn’t going to abate (anytime soon, in any case). But, what’s an operator to do with this barrage of facts and figures? These numbers will, of course, come in handy, especially given the fact that operators are fire-fighting on several fronts simultaneously. (I allude to the threat of over-the top players, razor-sharp competition and wafer-thin margins). Here’s the catch, though-these numbers will remain mere massive repositories of unstructured data, unless the operator deploys tools that would help them sift through the pile to uncover actionable insights about the customer’s behaviour.

Here’s where big data analytics steps in. Very briefly, today’s customer demands constant engagement with any brand. Things become even more complex, as we live in a multi-channel world. Companies thus need to think on their feet to provide an optimum level of customer experience management. Why? Well, to ensure that the customer is constantly engaged and is able to interact with them across multiple screens. Big data analytics thus helps companies to constantly innovate and make fast and streamlined business decisions, in real-time, of course!

So, isn’t the decision to deploy big data analytics in one’s business a no-brainer? Interestingly, no. Here’s why-most companies are extremely agreeable to deploying analytics to give an edge to their business. What trips them up is the debate on whether to build an in-house team for the same (too complex and time consuming) or to outsource the entire gamut of activities.

In fact, analytics as a service can help in a multitude of areas, which are not only limited to the team of experts who manage a set of tasks for an operator. There is a lot more-such as tools, security, storage, system integration capabilities to integrate new and old nodes and solutions to make this entire methodology more holistic. All these requirements can be bundled under the managed services providers’ scope of work. Apart from this, the managed services provider can also help to identify the appropriate action required (through tools or otherwise), after analysing the huge pile of data flowing through the network.

Of course, the argument isn’t that stark. According to Syntelli Solutions, a few key factors driving the argument in favour of the latter include:

High demand for a precious few data science practitioners– Typically, companies would rather turn their attention to solutions like managed analytics, than comb the markets for this scarce resource.

Exorbitant prices and equally exorbitant risks-Companies usually deployed analytics services through fixed-price or testing and measurement consulting models. As a result, the prices and risks associated with that project rose significantly. The solution? Begin relying on third-party analytics via the subscription model, of course!

Enter the Cloud. Today, cloud-based services are becoming increasingly reliable and secure. And this is just the beginning. It goes without saying that any future improvements in cloud-based processes will drive managers to a less expensive, more reliable managed service model.

It is little wonder, then, that operators are increasingly opting to outsource their managed analytics-centric activities to a managed services provider. Shall we take a closer look at the benefits this shift will accrue?

  • Managed analytics services provide the necessary push for an organization to move from a capital expenditure to a predictable operating expenditure cost model. Why predictable? Well, isn’t a set of easily definable results an excellent business case to put before the higher-ups in the organization?
  • Advanced tools and coherent and tested processes are deployed by the managed service provider. This, naturally, not only makes the service itself more outcome-driven, but price can be tied to quantifiable business value-extracting meaningful insights into the customer’s mind!
  • Organizations can achieve faster time-to-market. This is because managed analytics services cuts the internal process of building an analytics team by half (no hires, no systems to implement).

This, in a nutshell, is why a company ought to gravitate towards managed analytics services. On a parting note, I would like to add a brief caveat-operators would do well to remember not to get carried away and outsource all their analytics-related requirements. This is because the type of value this tool can provide varies from company to company. In all probability, each company already has a certain level of managed services implementation. The trick is to simply integrate and enhance the existing set-up. So, what’s your managed analytics services strategy?

November 24, 2016 0 comment
0 Facebook Twitter Tumblr Pinterest

The rapid proliferation of smartphones has brought with it many changes like the increase in data consumption, migration to 3G/4G, emergence of OTT providers besides increasing customer’s expectation from the networks. Today’s mobile networks are perpetually in a state of flux, where the end user experience is dependent upon several internal (device related) as well as external factors. While earlier, customer engagement management could be managed by pulling a few network levers, it is now more complex, calling for bespoke solutions that leverage big data analytics to provide holistic view of the user’s actual experience on the networks.

Operators should realize that in today’s mobile enabled networks, it is usually the subscriber’s perspective of the mobile operator and not the mobile operator’s perspective of the subscriber that counts. In order to drive customer engagement, telcos must step into the shoes of its customers to understand them better. In this context, the maturing of technologies, like, big data analytics couldn’t have come at a better time  enabling telcos to sort through huge volumes of data to draw inferences as well as trends needed for understanding customer’s actual (not inferred) user experience.

A good example of this is churn management. Big data analytics is used to spot customers who are about to churn on the basis of dormancy scoring models based on several parameters.  Depending upon the dormancy score and the customer’s expected life time value, the system picks the best campaign from the campaign portfolio to drive contextually driven customer retention programs.  Similarly, operators can analyze data usage patterns of their customers to profile multi-SIM users, which can go a long way in winning them back through timely customer engagement programs. For example, usage pattern of customer Y reveals “zero” activity during peak hours during the day. In-order to drive engagement the operator could send a promotion, say: “Get 50% off for all calls during 8 AM to 8 PM.”

Any such tool or system should answer the “who”, “when” and “where” of customer side of engagement: Who is the customer? Where is the customer present? When to roll out the engagement? For example, in order to roll out highly contextual campaigns the operators has to first segment the customer (the who) on the basis of demographic information, transactional patterns, life cycle, device type and social groups. Similarly, the “where” of any customer engagement model provide information on customer’s location and network environment. Finally, we have the “when” which lays down the timeframe and the conditions for rolling out any program of customer engagement.

The biggest challenge before the operator is to derive context from the humongous amount of data available in the cloud and on the server and utilize it for driving highly personalized CEM. Data poses its own share of problems – especially, the three Vs: Velocity, variety, volume. How to get insight from data on a real time basis? How to deal with structured as well as unstructured data? How to store this data? Also, operators must realize the customer relationships, if they are not moving forward, tend to atrophy over a period time, and hence need constant tending and nourishment (customer life cycle management).

With so much risk riding on networks, operators have to locate issues fast or risk losing their customers to competitors. In the highly competitive telecom marketplace, operators need bespoke solutions that leverage big data analytics for driving customer engagement processes that are customer triggered, aligned to customer life cycle, interactive  and near real time.

March 3, 2016 0 comment
0 Facebook Twitter Tumblr Pinterest