Where does your customer's emotional trail lead?

I was reading a recent article in Speech Technology Magazine, the subject matter was speech-to-text software for recorded calls, and this passage caught my attention:

Many people live and work under the false assumption that speech analytics is a form of plug-and-play technology; in reality the applications require a significant amount of work and resources to make them function properly. Users constantly have to fine-tune applications, improve definitions, realign parameters, and then work and rework facets of the applications until they get them right for their specific operating environments. Once that is achieved, everyone has to understand that if they’re going to use speech analytics effectively, then they’re going to need to dedicate resources to it.

To me, that sounds expensive and slow with many potential points of failure. But maybe it's worth it. Let's see, after committing your time, talent and treasure to this analytics approach, what is it that you finally get to unveil at the boardroom readout...


Yeah baby, nothing says, "I've got my finger on the pulse of customer experience" like transcribed customer service calls.

I know there is the whole argument that speech analytics is needed because text can be indexed, sorted, aggregated, and mined for the keys to customer bliss and angst (if only our company had more data we'd know the answers....wait a minute... what about putting the text of all those calls in a ginormous database?).

But lets peel back this argument and look at another reason companies adopt this technology, although it may not be explicitly stated: speech-to-text allows us to avoid listening to the customer's actual voice. Let's admit it, hearing a customer treated poorly, or hearing them get upset about an issue is painful. Even as a third party listener, its painful (if you are in the business of improving customer experience, it SHOULD be painful). But at risk of having the cliché-police knock on my door, I tender that customer analytics is a case of "No Pain, No Gain."

When we experience the customer's pain, we gain depth of understanding what to fix. Their emotions are a trailing indicator of what has happened to them and a leading indicator of what they will do with (or to) you. Does a customer leave for a competitor and show no emotion? Does a customer call you and show no emotion? Does a customer's emotion typically stay the same over an entire call? Of course not.

Neutral emotions may characterize a customer's "business as usual" mode, but by definition they are on the phone with you because they are not in "business as usual" mode. And as all the mis-interpreted e-mails that have caused arguments (ok, I'm guilty on both sides of those) will attest, text is and always will get mis-interpreted much more frequently than voice. Voice is the most reliable medium for understanding emotion, and correlating the events of an interaction with customer emotion is the most effective way to fix what may be broken in customer relationships.

I don't have an issue with deployment of speech to text technology, just want to make sure companies have the right expectations about what they have in the end, and more importantly, what they have lost in translation.

Why do Sharks Keep Swimming?

So they don't die, is the short answer (OK, that sentence had a distinctly Yoda-speak structure, didn't it?!). Sharks keep swimming and companies keep selling for the same reason: so they don't die.

I think that is why many of our projects lately keep getting pulled toward sales issues, even when they start out looking at some other aspect of the customer interaction. Times are tough, competition is high, and companies have a real feeling of sell more, sell better or die. If you are looking for motivators, fear of death is usually a pretty strong one.

So here is a scenario: company executives fear for their survival, and make a company-wide sales push to get the top-line numbers up. This trickles down to the call center where agents are instructed, nay scripted, to sell to every caller. Trouble is, there is a difference between a caller and an opportunity. An opportunity is created when the caller is in the right disposition to consider an offer, we say they are in a sales context. Some callers begin the call in a sales context, others will never get there, and repetitive sales scripting will have the opposite effect, making the customer want to buy less, or even leave your annoying company.

We've identified and tracked several factors that contribute to development of a sales context on a customer call. The call driver, customer attitude changes, agent rapport-building, the sequence of preceding events on the call, and the pattern of agent-customer speech all contribute to development of a sales context. Often when a customer reacts negatively to a sales script, it is not due to agent incompetence, but lack of sales context on the call. This is a process and training issue, but without understanding the factors that lead to sales context on a call for your specific business and customers, your agents will continue to be a fish out of water.

Swim teams and quality management

My kids are on a swim team this summer, their first experience with that, and we all find the meets to be lots of fun. One of the best things is the next day practice, when they hand out ribbons. There is one ribbon in particular that stands out in my mind, its the personal best ribbon when a swimmer improves their recorded time for a particular event. Its a colorful ribbon, and the kids always seem particularly proud of them, which shows a wisdom beyond their single-digit years. It occurred to me there is a corporate quality analogy here.

How does this apply to call centers? Well, first I should confess that the analogy breaks down because, despite an industry obsession with Average Handle Time, customer interactions are not a race against time like a swimming heat. They are a race against discontent customers, and since customers don't like spending time on calls any more than call center managers do, lower handling time is a byproduct of winning this race. The pride in a personal best carries with us from childhood. My kids are learning quality management in the pool: define a process (the entry, the stroke, the reach for the wall), measure the process, analyze and improve the process (morning practice!), and the real reward is not the ribbon, but what it represents: results.

The opportunity is to see, through the example of kids on a swim team, how when we are made aware of the elements that are in our control, how we can improve them, and we know the measures used to validate that improvement, the drive to better results and its own internal reward is a natural human inclination.

So what is in an agent's control, that is vital to winning the race against discontented customers, that we can isolate and coach, and that is measurable? Two such measures are Net Attitude Change and Net Experience Gap. The approach is to listen to calls and measure attributes (can be agent or customer oriented or general) that apply to a particular phase of the call, a key event, or the entire call. Out of the typical 60-80 attribute-value pairs that can be captured for a single call, views of what is really happening on calls can be quickly created. Views that are not available via CRM, surveys, ACD data, QA conformance monitoring, or CRM systems.

The most important view is how the customer's attitude changed over the course of the call, and the difference between the customer's measured experience and what the experience could have been based on our mash-up analysis. Agents and contact center management now have a measure that is vital to the organization, that is within their control, that is quantifiable in simple terms, and that provides the underlying detail to foster effective coaching and improvement. There are many organizations that are at a crossroads from chaotic changes in their call centers, from offshoring to onshoring, insourcing to outsourcing, mergers and acquisitions, changing call routing strategies and agent specialization, and shifting priorities and associated metrics. Maybe its time to empower something more basic: ignite that natural human inclination to chase those personal best ribbons and put the tools in place to support it.

The truth about support costs vs. customer satisfaction

As Jack Nicholson said, “You want the truth? You can't handle the truth!”

The truth is, we need to understand customer interactions, at the ground level, in order to improve customer interactions, and improving customer interactions not only retains and grows the customer base, but it has this interesting side effect of actually reducing support costs. I know this because the company I work for has been consistently achieving 10-20% increases in customer satisfaction concurrently with 20-30% decreases in applicable customer support costs. Those results contradict the assumptions of many, who assume there is a conflict between raising customer satisfaction and meeting cost reduction targets. Knowing that the companies we've worked with who are implementing this truth are outperforming their peers in profitability and market value, what will you do with the truth?

Back in the day, Henry Ford remarked, “you can have the car in any color you like, as long as it is black.” We know that we don't live in that world anymore, and we've seen the wreckage from companies who lost market dominance from not listening to customers. A case in point is Motorola, who lost control of a market that they owned when an upstart from Europe started making colorful mobile phones. Do you think any customers told them they wanted more style, more pizazz? I'm sure many did, but no one was listening, at least no one with an orientation toward larger business issues and market drivers. But those are the competitive advantage benefits to listening, we are focusing here on the more pragmatic, near term objective of improving contact center performance.

Traditional methods of analyzing customer support grew because they were readily available and they generated simple metrics. These methods include the metrics generated from call center equipment, the analytics generated from CRM and other IT systems, QA montoring and surveys. They have their uses, but not for telling you what is really happening on calls and how it can be fixed. At best, they provide a trailing indicator of the performance of customer facing operations. You know, the metrics that look in the rear view mirror and announce, “yep, we ran over a customer back there, don't know how that happened.” They also have the pitfall of being one-dimensional: focusing on a measure like handle time just adversely affects other important measures, like customer satisfaction, retention, and first call resolution. And while it is true that QA monitoring is a listening technique, it raison d'etre is process conformance, not process improvement.

So if we want to improve the customer experience, and we acknowledge that this customer experience is formed by the real interactions they have with us (I call these "moments of truth"), the solution must involve getting inside those interactions. We need to be able to understand real interactions as they occur, observing real customer behavior. I'm not "in the moment" when I respond to a survey, I am just using my imperfect recollection.

So you want the truth? You can handle the truth: support costs can be reduced while also increasing retention and customer spend. But it is done not by chasing one-dimensional metrics (e.g. handle time, reported CSAT) but by improving interactions: their flow, sequence, key moments and events, and missed loyalty-building opportunities.

Agents are part of the value equation

Many stakeholders must combine and interact to make a successful company. These includes employees, management, customers, vendors, business partners, and investors. So, when we look at contact center agents, whose actions affect many of these stakeholders, we are not just looking at the cost of the calls they handle, but also the value (or cost) of the interactions they have within the system of stakeholders. This tells us clearly that we need measures for evaluating, training, and incenting agents that are in synch with the higher level goals of the company.

Many measures today are flawed proxies for company goals, and often focus too rigidly on cost containment. Driving business processes by over-weighting a single-dimensional measure, such as average handling time (AHT), causes side effects that negatively impact the equally important goals of customer retention and revenue growth. This blog will contain my observations, news links, and other posts related to improving contact center performance.