Big Data: Does it Improve Customer Service?

Posted on by

About Martha Brooke

Program Director and Founder, Interaction Metrics
1

Big Data. Conferences, businesses, bloggers—everyone wants in on the action. In particular, customer service organizations want Big Data because they believe that parsing multiple streams of information will improve service and, therefore, boost customer loyalty.

And, certainly, customer service needs to improve. After all, the best companies in the world barely achieve an NPS (Net Promoter Score) over 60. Meanwhile, the average ACSI (American Customer Satisfaction Index) hovers in the mid-70’s.

But will Big Data work for customer service? Is Big Data a worthy investment?

Yes and No. What we find is that for some aspects of customer experience—such as the ability to customize customer service recommendations—Big Data helps. But, when it comes to customer service conversations, its value is limited.  In fact, when talking, emailing or chatting with an actual human being, Big Data is less than 5% relevant. Here’s why: Conversations are, by nature, emotionally complex and unpredictable. Big Data can’t show associates how to listen, engage and connect in the present moment. And, based on listening to tens of thousands of calls, we find issues arising in the course of the conversation make up approx. 95% of an interaction.

For mastery of the present moment, you need Interaction Data—because Interaction Data uniquely captures what happens, and how customers respond, throughout conversations. Plus, Interaction Data doesn’t require a large investment in information processing. Even better, Interaction Data funnels into Interaction Analysis which pinpoints gaps, opportunities and how to improve.

Speech analytics and voice recognition catch pitch changes and word choice, which provide basic interaction information. But since they don’t deal with emotional complexity, these tools only offer a primitive form of data-gathering and analysis. Basically, they trade meaning and nuance for the ability to process volumes of content quickly.  Remember, even the costliest, smartest computers like IBM’s Watson merely produce possible answers and their associated confidence intervals—they can’t have real conversations or understand the meaning of what’s going on.

When it comes to improving customer service, how much should you invest in Big Data? It depends.

  • Are you in a simple service environment where it’s possible and important to predict customers’ preferences and answers?
  • Or are your interactions complex, such that it would be better to show your customers that you’re in sync and care about their questions?

Also, consider: Big Data could be a distraction. As Nate Silver says in The Signal and the Noise, “The signal is the truth. The noise is what distracts us from the truth.” If noisy data impinges on your ability to hear customers’ signals, Big Data will do more harm than good.

Customer Feedback: Social vs. Traditional Tools

Posted on by

About Martha Brooke

Program Director and Founder, Interaction Metrics
1

Finding out how your customers feel about you, i.e. collecting customer feedback drives most customer experience improvement; after all, “you can’t manage what you don’t measure.”

So what about social media? Certainly, by its nature, social media is where customers go to communicate their feelings. Therefore, monitoring review sites, along with Twitter and Facebook for customer feedback just makes sense. But, unlike traditional customer feedback tools, social media never provides a complete, accurate and objective perspective – so don’t abandon your traditional tools yet.

The upside, Social Media can be Actionable and Does 2 Things Well

Social media:
1) Captures complex thoughts and feelings about the customer experience.
2) Lets customers define which aspects of the customer experience matter most to them.

Compare this to surveys which prioritize multiple choice questions and, by necessity, work from a pre-existing hypothesis about what’s most and least important to customers.

On the downside, Social Media is Inherently Misrepresentative

A major downfall of social media customer feedback is that it is subject to extreme response biases.

Company solicited social media always skews toward loyal customers or “friends.” User-generated customer feedback (like that on Yelp, Amazon reviews, and Epinions) captures only the extremes of gushing or venting. And, both of these sources can only provide customer feedback from one group: members of social media sites, which may not be the majority of your customers.

Use social media for gathering actionable insights; use surveys for accurate data. But for actionability and accuracy use customer interviews, they’re a vastly underused feedback tool.

The Value of Customer Interviews

Customer interviews enable you to ask your customers questions that get at the heart of the customer experience, but that are not what customers address in their social media rants. For example, an interviewer can ask:  “who is the competition and what do they do particularly well?” Plus, interviews can be designed so that the sample group is large enough and balanced in a way that accurately represents your complete customer base.

Why are customer interviews underused? Sometimes, because they are more costly than generic email surveys; although they are rarely as expensive as focus groups and usability studies. But mostly, because companies don’t know customer interviews are an option and the lure of social media has obscured some of the best, existing methods.

Branded Customer Service: Don’t lose the customers you already have

Posted on by

About Piper Smith

Lead Customer Experience Analyst at Interaction Metrics
1

When you think about branding you may think about logos, billboards and ads—mostly directed at enticing new customers. But customer acquisition is expensive, and while you are busy attracting new consumers, your current customers could be slipping out the back door.

So how do you shift your focus to customer retention? You could try the usual approach, which is to probe customers’ thoughts by sending yet another customer feedback survey. But the problem is that most customer feedback surveys don’t show you how to create a rewarding experience that engages customers through their hearts, minds and senses.

For this, consider branding customer service. Branding customer service is a way to support your advertising and marketing messages. Perhaps more importantly, branding customer service is a proactive way to demonstrate your core value and key differentiating points to your current customers.

In Branded Customer Service: The New Competitive Edge, Jenelle Barlow and Paul Stewart make a great case for branded customer service. Their book outlines how customer service can be more than just good or bad; it can embody the values of your company by being “on-brand” or “off-brand.” They argue (and we agree) that when brand promises and the actual customer experience align, customer service multiplies the effect of advertising. Conversely, when the two are mismatched, your advertising is completely undermined. Basically, customers are more likely to get your brand and remember your company when your messaging is consistent.

For companies trying to maximize value by branding customer service, Barlow and Stewart’s book gets you started. However, their approach is lacking because they don’t explain how to accurately and objectively measure whether you are succeeding in your endeavor to brand customer service. As Peter Drucker so aptly quipped, you can’t manage what you don’t measure.

In the weeks to come, I’ll further explore branding customer service with an emphasis on how to measure success in this area.

Improve Your Customer Service with Our 5 Step Plan

Posted on by

About Kelsey Zorn

Research Lead at Interaction Metrics
Reply

Improve Your Customer Service with Our Step-by-Step Plan

Everyone’s trying to improve their customer service, some with more success than others. Last month, the New York Times column The Haggler, which usually posts customer service horror stories, gave voice to tales of exceptional service. The column was a refreshing change and likely sent plenty of business towards the companies mentioned.

So how can a company improve their customer service enough to be column-worthy? Google turns up a hodgepodge of papers, articles and books, some of which may bump your customer service up a notch, but you’d be hard pressed to find a clear, step-by-step plan.

That’s where we come in. We’ve improved customer service for dozens of companies: for example, GE CareCredit attributes a 31% increase in revenue to our customer service improvement program. So, we thought it was time to share the plan we use to improve customer service.

Our 5 Step Plan to Improve Customer Service:

  1. Pick your goals. Ask yourself what, in addition to answering customers’ immediate questions and concerns, you’d like to accomplish through customer service. There are seven main goals that customer service can achieve. These include optimizing self-service, expanding your brand and increasing customer loyalty.
  2. Catalog your interactions. Identify each customer touchpoint and decide how you will realize your goals from step 1 in these front-line situations.
  3. Create a model communication for each customer touchpoint. Your model communications may take the form of call scripts, email templates, checklists or talking points. Regardless of the format, you must set the bar and show associates exactly what you expect from each touchpoint that you identified in step 2.
  4. Measure meaningful details. Averages often mislead. So be sure to measure the weird stuff—the stuff that gets at the heart of interaction quality—like percent of time associates demonstrate they care about the customer, or end their interactions in an authentic and positive way.
  5. Keep your employees in the loop. Monitor your performance with mystery shopping, customer service monitoring and other forms of customer service quality assurance. Make sure your whole team knows how they’re doing, and which specific steps they need to take to improve their performance.

Read more about our 5 step plan in our full article here.

Here’s Why U.P.S. Failed the Top Haggler

Posted on by

About Martha Brooke

Program Director and Founder, Interaction Metrics
6

David Segal’s “The Haggler” column in the 2/12/12 edition of The New York Times covered a few horror stories about U.P.S., all of which had one thing in common: a customer’s package was lost, and U.P.S.’s customer service was useless. The haggler conservatively estimated that if U.P.S. is truly exceptional at keeping track of packages (only losing 0.05%), then every day U.P.S. bungles about 8,000 deliveries.

8,000 mistakes a day means 8,000 upset, confused customers and 8,000 calls in to U.P.S. customer service. Segal spoke with one U.P.S. rep who was “extremely helpful” and several who were “pointlessly brusque”. Other customers spoke with Segal about their U.P.S. travails and had a similar ratio of helpful to useless service.

Now, I’m going to extrapolate: If only one out of several calls (let’s say 1/3) was helpful, that means U.P.S.’s error rate when it comes to these critical conversations is 67%. How could a company with an enviably low error rate for deliveries have such a high error rate when it comes to a key, memorable customer conversation that happens 8,000 times a day?

Here’s why: Apparently, U.P.S. Managers don’t consider the “hazy conjecture” and inconsiderate care their team gave Segal to be a problem. If they did, these frustrating conversations wouldn’t have happened—or, at least not 67% of the time.

Most companies simply don’t recognize the importance of customer interactions. They think that with the right software and generalized metrics, they’ve got things covered. In fact, according to a much quoted Bain & Company study, 80% of companies believe they deliver a “superior experience,” but only 8% of customers agree. In our customer service audits, we find that poorly designed surveys that result in inaccurate information are often the source of staggering gaps between executives and their customers.

What’s the solution? Because getting covered by The Haggler can’t possibly be good for business, U.P.S. needs to change its perspective. U.P.S. managers need to turn their gaze away from averages and take a closer look at their actual interactions. After all, by definition, averages obscure the details.

If we had a U.P.S. exec’s ear, we’d tell him/her that they need to study their customer interactions with fresh eyes and different metrics—metrics that more realistically capture the lived experience. If they did that, they wouldn’t have to learn the weak points of their customer experience by reading The New York Times. As always, snaps to The Haggler for giving a voice to the actual customer experience!

You Responded, But Did You Answer the Question?

Posted on by

About Martha Brooke

Program Director and Founder, Interaction Metrics
5

Customer service boils down to a simple give and take: customers have questions or problems and companies provide answers. It’s pretty straightforward. But all too often, customers are left befuddled, scratching their heads and wondering: “What did they mean? Were they listening? What do I do now?”

Transparency and authenticity are the buzzwords of business. So why can’t customer service be more clear?

Let’s look at the word “answer”. Merriam Webster gives it two very different meanings:

  1. Something spoken or written in reply to a question. 
  2. A correct response.

Companies and their call centers often conflate the two meanings and assume that because they replied, they provided an answer.  But while all answers act as replies, not all replies provide an answer.  A real answer is accurate, complete, understandable and addresses the customer’s unique situation.  When customers ask for solutions, but get vague replies and obfuscations, brand loyalty and advocacy plummet. While yammering happens in stores, it is particularly prevalent in call centers.

There are a number of contributing factors:

First, follow the money: sadly, blather is profitable. Typically, revenues rise as call centers chat, talk and email more, regardless of the information conveyed and quality of the answer. And unstrategic call centers view web self-service as a revenue-eliminator, instead of seeing self service for what it is: a way to increase customer satisfaction.

Furthermore, good answers require an investment in resources and a level of spending that is beyond the scope of a typical call center.  Between putting out fires and filling seats (no easy task) call centers are stretched thin—at least relative to the budgets clients require.

Plus, sometimes companies are in a hurry to provide the bells and whistles of excellent customer service, the kind of service that personalizes, customizes and brands. While we always advocate doing more with customer service, this should never be at the expense of the basics. As Dixon, Freeman and Toman in the Harvard Business Review note, “what customers really want (but rarely get) is just a satisfactory solution to their service issue.” In other words, make sure your answers pass muster before aspiring to excellence.

How do you guarantee correct answers? For this, you may very well need an answer engine technology and you’ll also need what we call a Total Answer Management strategy. As Tom Wentworth and others have pointed out, technology is great but it won’t cure your customer experience by itself.  Here’s what a strategy of Total Answer Management means:

Total: Take inventory and prepare a report that details every question (including small variations) that customers have.  Make sure you study enough interactions—it should be at least one week’s worth of conversations.

Answer: Nail the responses to each and every customer question and thoroughly vet those answers for substance and style. Chefs show their cooks everything: how they want the egg poached, how the dish should be garnished and plated, etc. Likewise, if you can’t show your associates a perfect answer, don’t expect them to come up with it on their own. Without clear examples, customers will get different answers depending time of day and who they reach. That’s what Gallup found in their study of the quality of customer service in the Harvard Business Review: “the customer experience still depends almost entirely on the particular rep who takes the call.”

Management: Determine which questions are best answered by which customer service channel.  While some questions can be answered by an FAQ, others require the kind of interaction that only a phone call provides.  Management is about directing customers and measuring answer gaps. It’s also about continually monitoring to determine when answers need to be upgraded or added to—all while gauging how well call centers deliver on their promises.

How well do you answer customer questions?  We’d love to hear about your answer solutions and call center strategies.

Links and Resources:

 

CEx 2012: Nail Email!

Posted on by

About Martha Brooke

Program Director and Founder, Interaction Metrics
3

The biggest challenge for the customer experience in 2012 will be to make customer service emails work.  I am not talking about simply sending responses in a timely manner, although that’s a good start. Rather, it’s time for companies to step up and provide thoughtful answers to customers’ questions, answers that show the company and its representatives are listening and care what customers have to say.

Consider the advantages of email:

First, according to numerous studies, emails cost much less than phone calls. In some cases, email customer service costs half as much as the phone experience. Savings of this magnitude deserve to be front and center on any company’s radar.

Second, it’s the customer-centric thing to do.  The customer gets an answer on their time. They don’t have to wait in your call queue and they don’t need to plan around your company’s hours. Instead, customers state complaints or pose questions when it’s convenient for them.

Third, companies can provide information that’s difficult to provide over the phone. For instance, contextual information such as how to avoid problems in the future and extra resources that the customer can refer to at their leisure are better served through text than conversation. In the customer service monitoring business, we’ve heard far too many calls suffer when CSRs tell customers to go to a long-link-web-address and the customer simply gives up trying to write the whole thing down. Links in customer service email solve this problem.

But in spite of these advantages, the state of customer service email is bleak. Often, customers don’t really get answers and when they do, those replies can be a robotic string of FAQs that do little to forge a bond between company and customer.  And customers have been burned by email. They have learned that the only way to get a meaningful answer is to call a company and when that doesn’t’ work to ask for the supervisor. If that’s a dead end, customers often air their complaints on social media sites or in letters to the legal team or the company president.

Why is email customer service such a sorry state of affairs? Here are two reasons:

  • Call centers are exactly that: they specialize in calls. While nearly every call center offers email customer service, the fact is email and chat are just add-on services, not their specialty.
  • And then, there is the Peter Drucker quote that we return to again and again because it offers a reasonable explanation for all kinds of execution failures: “you manage what you measure” or, as other quality gurus have said “you can expect what you inspect”. The majority of customer service email is an improvised mess because companies lack the right guidelines, procedures and metrics.

All companies that have nailed customer service email have done so because they track critical factors like customer effort, answer completeness and overall empathy.

If you are a company with great customer service email, tout it! Tell your customers that you are different and that while their calls are welcome, that’s not the only way to get a straight answer. Promise replies within the day written by associates who take the time to read—not just gloss over—questions.  Even better: if you can verify it, state up front that your customers report that 95% of your email responses are smart, easy-to-understand and friendly.  By communicating these facts from the start, you’ll get the transition from call to email customer service underway.

2012? Let it be the year of excellent customer service email. Mastery on this front means you’ll have more satisfied customers, better interactions and a more profitable customer experience. Email on!

Experience Counts: The “Downgrade” Debacle

Posted on by

About Martha Brooke

Program Director and Founder, Interaction Metrics
3

Shep Hyken’s reflections on his hotel experience illustrate a common customer service problem. Here’s what happened: Shep reserved a room, but when he got to the hotel it wasn’t ready, so the front desk worker offered to “downgrade” him to a room that was already prepared. No one is ever happy about a “downgrade,” especially when it comes to an unknown hotel room. Will the room be a closet? The bed a cot? But what rightfully surprised Shep was that the “downgraded” room was nearly identical to the one he originally reserved. So why didn’t the associate describe the room as a virtual lookalike? Why promote the negative when the positive is right at hand?

In subtle ways, frontline employees can undermine an otherwise excellent customer experience. And this slow erosion of the company’s image isn’t captured by big-picture metrics like Net Promoter Scores or overall satisfaction. If Shep had filled out a satisfaction survey at the end of his stay, he would probably have said he was “somewhat” or “very” satisfied, assuming nothing else went wrong. If he were just the average customer, he probably wouldn’t even have thought to take note of the rep’s use of “downgrade.” But it still impacted his perception of the company.

Some might say that small details like word choices don’t matter, that as long as the customer didn’t complain about them it’s not a problem. But Zaltman’s studies, amongst others, find that while customers don’t take note of very much about their buying experiences, small cues have a huge impact on customers’ unarticulated feelings and repurchase rates. Harvard Business School: The Subconscious Mind of the Consumer (And How To Reach It): “95 percent of our purchase decision making takes place in the subconscious mind.“

Moment-by-moment details are critical to a company’s sales success, so it’s hard to understand why managers tend to rely on overly-simplified averages that don’t represent the lived experience.

Action Item: To maximize customer loyalty, ask fine-grain questions like “how many positive words do our reps use in each customer interaction?” Or, “how often to associates use language that supports or tarnishes our brand?” Set expectations for how associates should act to align with your most critical objectives. That way, instead of some amorphous average score, you get a clear, actionable picture of what’s going on on your frontline and what your reps need to do to make it better.

Branding Customer Touchpoints: Why Most Companies Aren’t Cutting it

Posted on by

About Martha Brooke

Program Director and Founder, Interaction Metrics
11

With the dramatic success of startups like Dropbox, Groupon and Twitter, many people are wondering, what’s their secret? Forrester’s Kerry Bodine took note of one innovative approach shared by all three companies: these newer companies seek to actively brand and engage customer feelings at even minor customer touchpoints such as the confirmation email, bill, etc. This attention to managing brand detail could be driving heightened loyalty and advocacy levels.

Certainly, this idea of branding customer touchpoints has made its way into the zeitgeist with Zappos CEO Tony Hsieh describing the “telephone as one of best branding devices out there” and other marketers quick to point to the power of customer experience branding.

But the biggest companies mostly leave their customer touchpoints untapped for their brand value and when it comes to customer service (the most memorable of all touchpoints), branding is often completely ignored.

Bodine argues that these opportunities are squandered because large corporations are crippled by too many silos, over-concerned with the demands of ROI and lacking customer focus.

As Customer Experience Analysts, we find these factors are contributive, but hardly the root cause. The real reason why large corporations tend to lag when it comes to touchpoint branding is that corporations largely subscribe to an antiquated notion that engaging customers’ feelings belongs in the back room, not the board room.

Until CEOs and COOs start asking essential customer experience branding questions, like “what percent of our customer interactions are branded?” “how engaging are our frontline reps?” or “have we trained our customer service reps to deliver our core marketing messages?”, these large companies won’t have the institutional inertia to carry their big marketing ideas through to the front line.  And that means they won’t be able to keep up with the hip Zappos-types and Dropboxes of the world.

Unstructured Data Needs an Unstructured Approach

Posted on by

About Martha Brooke

Program Director and Founder, Interaction Metrics
5

We heard a great webinar from Bruce Tempkin at CXPA yesterday about how to gain customer insights from unstructured data.  Bruce addressed the cavernous disconnect between limited, multiple-choice responses and the reality of complex customer thoughts and feelings.

Here’s what we would add: when dealing with unstructured data (call recordings, free-form response cards, social media posts, etc.), it’s essential to start with an equally unstructured approach. All too often, we see speech analytic programs failing because structure is imposed too quickly. In other words, sentiment and specific words are defined as key indicators of the customer experience before the entirely of the customer experience has been dissected and understood.

The problem with picking indicators too early in the unstructured data mining process is that analysts will only see what they thought was important, not necessarily what IS most important. Also, customer experience branding opportunities will most likely be overlooked. While the idea of ‘when to impose’ structure may seem academic or an ‘overly fine’ point, qualitative researchers will tell you, this is far from the case.

Interaction Thinking approaches unstructured data with few predefined expectations, allowing the data (i.e. customers or experiences) to show their deepest meanings and implications.  It takes time to dwell in a state of unknowing before moving to everyone’s favorite part of customer experience analysis: next steps, statistics and conclusions. But this approach pays off with richer, more relevant insights.