Don’t Ignore the Elephant: Audio’s Role in Big Data


First, Some History

Everything changed, and few noticed. Many people ask me, what is “Big Data,” and why do so many periodicals, books and business gurus say it’s changing everything? The answer has a lot of facets and remains partially unknown. What I do know is that big data is impacting everything. Let me explain why.

From the beginning of time until 2003, all of humanity created about five exabytes of data. That’s a lot. This year (2013), the same amount is created in only ten minutes! Let’s put that into perspective. Every ten minutes we create the equivalent of enough “old-school” CDs, each containing 650MB of data, to reach from the north pole to the south pole when stacked flat.

What Does This Mean For Today?

But while the sheer volume of information might not be relevant, what does matter is that interactions between people and companies are now in data form. You blinked, and in the last five years this happened. People tell people what they think, how they feel, and what drove them to those opinions and emotions. These interactions hold an unbelievable amount of data, all with human context. People are emitting interaction data (words) constantly. For the first time in history we can track and understand the patterns and extract the collective meaning of it all.

Where’s The Data Coming From?

In the big data scurry, two things have become abundantly clear: 1) storage is ridiculously cheap and 2) anything and everything can and will become digitized.

In 1995, it cost about $1000 (US) to store one gigabyte of data. Today, you could pay for the same with less than a nickel. The other thing that happened is that some smart people within the Apache project invented a way to distribute data across these now cheap, commodity storage servers. Basically, “unlimited” computing is now available.

Things you do every day are becoming data. You own and love them – smart phones, location based sensors, smart cards, energy consumption monitors, click streams generated on your tablets, etc – a hyper-proliferation of devices, all producing data. Now, when you actually use your device to call, your voice is no longer an amplified signal – its bits and bytes.

Ok, so you can digitize the experience, and create the data and you can even store it. But what seems clear is that most entities get lost in the technology and forget about the impact.

Audio Still Reigns Supreme

The analyst firm, Forrester, studied people’s channel preferences (like: phone, social, web, or chat). They found that when something is wrong like: a medical claim not being settled, an auto payment failing, or a home network not connecting to a cable company’s network, people don’t tweet or update Facebook with the details, they call another human being to fix it. In fact, in the United States, the study showed that all other channels combined (web, self-service, web-chat, text, and social), do not equal the preference for a person-to-person interaction.

What’s been forgotten is that of the almost 200 billion interactions between a company and its customers, 60% of them remain voice interactions. They’re the big elephant in the room. You know you need to look at them, but you aren’t sure how they intermingle with the rest of what you have going on. But to ignore the voice interactions would be to overlook one of the biggest assets you have because those unstructured interactions contain a golden nugget. The voice of the customer. It’s within the voice interactions captured by the contact center that a customer will tell you why they are reaching out to your company, how their experience with your brand has been, whether they’re likely to remain loyal or churn, if they’re considering a competitor, etc.

The key to getting the voice interactions to intermingle is to use an interaction analytics solution that is scalable and flexible enough to put structure to that unstructured data so that it becomes empirical, quantifiable business intelligence and combine it with already structured data, like customer value scores or relational information. Doing this unlocks a secret weapon of sorts. Companies employing this method will be able to understand and predict customer behavior like never before, allowing them to lower operational costs while improving customer experience. This big data combination will be the recipe of the new generation of successful companies.

Putting it Into Practice

Here’s an example. Most companies collect Net Promoter Scores (NPS). This is structured data. Most companies also segment their customers. This data is also structured. The problem is, most companies do not have a reliable way of knowing WHY their NPS are high or low. There may be some survey data, but that’s likely only a range of numbers, and not tied to specific interactions that help pinpoint the exact agent behaviors or business processes that are driving customer dissatisfaction; thus, it is difficult to implement change. This is where unstructured “big data” comes into play. Putting this into a tangible scenario:

A cable company wants to know why their recent promotion to sell networking and monitoring to the “wired senior segment” isn’t working. A look at their structured data shows low NPS scores for that customer segment, which means they’re dissatisfied and aren’t buying. But why? Analyzing the unstructured data – which is calls placed by that customer segment over the past 6 months – and finding the structure, trends, and empirical information in it, reveals a large uptick in tech support calls. Further investigation shows that other products offered by that company were perceived as “hard to connect” based on challenges the company had with previous installations. Now that the company knows the problem, they can market the solution as coming with a free, in-person install for this customer segment to help alleviate these concerns and increase sales.

The Take-Away

We’re in the middle of an age where data – every kind of data – drives more and more decisions. These decisions will be highly prescriptive and will be real time. Data has become a strategic asset, and a company’s success will depend on how well and how often its employees, at every level, use the asset. IBM Research Labs stated that analytically driven organizations, today, right now, have a 36% higher compound annual growth rate than those that are not. MIT Sloan Management Review, partnered with SAS in a multi-year research initiative. One of the published results: 67% of respondents reported a competitive edge from their use of analytics. Don’t blink again, and most importantly, don’t forget to listen – people will tell you through data exactly how to dominate in this new economy.

Want more in-depth information about big data, it’s roots, how it can impact a business, and how you should be thinking about applying it to your own company?  Download this guided presentation for a breakdown.

[Photo: “the elephant in the drawing room” by Thunderchild 7 via Flickr Creative Commons]
Categories: Intro to Interaction Analytics