IN THE MUSICAL THEATER ADAPTATION OF LEO TOLSTOY’S “WAR AND PEACE,” THE LYRICS TO THE SONG “NO ONE ELSE” BEGIN WITH: “FIRST TIME I HEARD YOUR VOICE/ MOONLIGHT BURST INTO THE ROOM.” It’s just a hunch, but given Tolstoy’s lack of references to call centers, this lyric is probably not referring to agent-customer conversations. In fact, when an agent first hears a customer’s voice, the only thing that should be bursting into the room is a sense of urgency.
Industry experts have long estimated that 25 cents of every dollar spent by a call center goes toward fixing issues that weren’t properly addressed during the customer’s initial call. Those experts were overly optimistic. A recent study by the Yankee Group revealed that 66% of all contact center costs can be attributed to callbacks.
It’s no wonder that more organizations than ever are focusing on first-call resolution (FCR) strategies designed to reduce the number of interactions between their agents and customers. Solving a customer’s problems during their first call:
- is the fastest and easiest way to significantly reduce costs;
- is an excellent way to measure your call center’s effectiveness;
- improves customer satisfaction; and
- increases customer retention rates.
According to the SQM Group, FCR is the highest correlated measure to customer satisfaction of all contact center measures. The statistics are stunning: a 1% gain in FCR translates into a 1% gain in customer satisfaction.
What’s the best way to improve FCR? Speech analytics. A speech tool is an effective, scientific means to quantify the number of customers who call more than once to resolve their issues. Contact center management can leverage speech analytics to identify trends and specific behaviors, customer issues, hints of attrition, process challenges and inhibitors to FCR.
A recent study by the Yankee Group revealed that 66% of all contact center costs can be attributed to callbacks.
Many companies that incorporated speech analytics as part of their FCR strategies have reported improvements of 8% to 10% simply by correcting how their agents respond. In fact, just the awareness of speech analytics can change agent behavior. Prior to deployment, one organization’s customer callback rate was nearly 8%. Once agents knew that all their interactions were being monitored and evaluated through speech analytics, the rate dropped closer to 3%.
Elevating FCR rates is a big reason why speech analytics tends to pay for itself many times over. Although no speech tool is 100% accurate, an automated approach will be far more accurate than any individual’s speculation or judgment. “Time and time again, analytics has helped our clients discover more effective ways to address customer issues when they first surface,” says Scott Kendrick, VP of marketing for CallMiner. “Analytics isn’t just a better way to track FCR or repeat calls. It provides actionable information and supports root cause discovery, allowing call centers to minimize repeat calls and the need for callbacks. This should be a priority for every call center.”
MY AGENT SAID WHAT??
Believe it or not, the following statements were actually uttered by agents.
“You don’t need a landline with us. That’s why we are soooo expensive!”
“That’s just our policy. I know it’s bad.”
“We are endorsed by the NFL and just about every other organization.”
A recent benchmarking study by the Ascent Group disclosed that the most common non-SA method used by call centers to measure FCR is massaging data from customer management systems like CRM and CIS (27% of participants reported using calculations from CRM/CIS systems). The other top methods reported in the survey were: call monitoring (20%), recent contact surveys (18%), and post-call surveys (15%).
|PHRASES SAID BY CUSTOMERS INDICATING THEY HAD CALLED BEFORE
Called a few days ago
|PHRASES SAID BY CUSTOMERS WHO WERE CALLING AGAIN BECAUSE THEIR ISSUE HAD GONE UNRESOLVED
Hasn’t resolved it
Unfortunately, these commonly used traditional methods are inherently flawed due to factors such as:
- Subjectivity of the agent in what and how he or she chooses to report.
- Lack of a simple way to correlate key utterances to patterns of caller behavior.
- Limited source of data from customers willing to participate in surveys.
- Inability of QA team to have confidence in the data they’re analyzing.
In contrast, speech analytics offers distinct advantages:
- Customer comments are captured and examined objectively and scientifically based on actual utterances, instead of relying on individual agents’ biased views about whether an issue has been resolved.
- All phone interactions are scrutinized, not just a random sampling.
- Calls are filtered based on the relevance to issue resolution.
- The data collected can reveal patterns and expose issues that can be used to retain customers.
The Three Components of a Successful FCR Program
Optimizing FCR is both a science and an art. The best outcomes occur when all three of the following elements are in place:
- Identify repeat calls
- Understand the root cause of why customers are calling
- Identify process improvements
IDENTIFY REPEAT CALLS
Let’s begin by examining the anatomy of a typical repeat call:
Caller: “Hi, maybe you can help me. I talked to someone yesterday about the service light on my converter box and they told me that if it stopped working again, I should just call back. This is kind of frustrating. The light went out last night and the TV won’t work. Actually, I’ve called four times over the last two weeks. They keep telling me to try something new and if it doesn’t work, I should just call back. At this point, I’d just like a new box.”
Although it’s clear that this customer has called in before, identifying a repeat caller is not always so obvious. In some cases, customers may even deny having called before. Only by auditing hundreds of recordings can you isolate all the potential phrases (e.g., “I just spoke with” or “I contacted you”) that single out a repeat caller. You may be surprised at how many ways repeat callers can be identified.
The overall efficiency of this process will depend on the capabilities and features of your speech tool. To ensure accuracy, you will also need to distinguish repeat calls from agent callbacks, dropped transfers, previous interactions through different channels and other potential reasons for a callback.
UNDERSTAND THE ROOT CAUSE OF WHY CUSTOMERS ARE CALLING
Imagine if your annual review consisted of just one statement, and that your compensation was dependent on it. If your review stated that you were an “average employee” without offering any insight into how you were performing in each of your various responsibilities, you wouldn’t know what areas were in need of improvement or even where to begin. Consequently, little if any improvement would actually be made.
The same dynamic holds true for call center operations. Repeat callers often call about issues that are unlikely to be detected by random samplings. If you don’t know what issues are driving these calls for each segment or workgroup, you cannot hope to resolve those issues and reduce call rates. These issues can only be uncovered and addressed when speech analytics provides solid evidence that a recurring problem exists.
The first step in root-cause analysis is creating categories. Start with big buckets (e.g., “billing issues”) and create smaller buckets (e.g., “billing late payment”) as needed. Second, identify the cause of the problem for each category. Is it an operational issue? A language barrier? Insufficient agent training?
Third, arrange meetings with the departments responsible for the issues at hand. Present actual recordings for each team to listen to and review. Listen to about 10 calls in each category so the issue is clearly understood. If clarity is lacking, you may need to refine your measurements or widen your investigation. This process of reviewing high-value recordings is the most effective way to communicate problem areas and spur action. Even so, be prepared to encounter defensiveness and resistance. You may be required to ask tough questions and challenge the status quo.
Fourth, develop recommendations. Ensure that the appropriate department and personnel take ownership of the problem and design an action plan for change. Last, implement the action plan and follow up by monitoring the overall number of calls and FCR metrics to ensure that the solution was effective.
Pinpointing the root causes of calls and resolving the associated issues allows you to track FCR performance on a per issue basis. For example:
- After spotting a pattern of customers calling to complain about their password not working, one organization was able to reduce call volume by 2% just by identifying and fixing issues with its password reset process.
- A retailer told me he was able to reduce the number of callers who wanted to speak with a receptionist by 5% by identifying why customers were calling, and then adding a new option in their auto attendant menu.
- My company recently helped a healthcare organization figure out why it was receiving so many midday calls. By recording hits on utterances like “keep having to call” and “third time I’ve called,” analytics revealed that a significant percentage of calls were coming from patients who were unable to reach their doctors. Once they understand the problem, they could work on closing the loop with a solution.
IDENTIFY PROCESS IMPROVEMENTS
Regularly review your entire operations, from people to processes to technology, to identify errors, inefficiencies and bottlenecks contributing to repeat calls. One organization’s speech tool revealed customer confusion about its loyalty program. Alerted to the issue, marketing found an error in the program’s Silver package, which was quickly corrected.
Searching through an agent’s calls for targeted key words and phrases almost always uncovers opportunities to improve performance. Average handle times will increase and the number of callbacks will spike if the agent lacks knowledge on a given topic and consequently provides wrong or incomplete information. You may also find that well-intentioned agents are unwittingly inviting customers to call back by saying things like, “If that doesn’t work, you can always call.”
Through valuable insights gained with speech analytics, one call center retrained their agents and decreased the percentage of repeat calls from 40% to 30%, with a 20% improvement in average handling time. By becoming more efficient, management could cancel plans to hire 20 new agents.
Of course, not all repeat calls are the fault of the agent. Agents are tasked to resolve customer issues on the spot, which may mean answering a billing question to the customer’s satisfaction or finding a technical solution. It’s therefore incumbent upon call center management to ensure that the correct processes are in place so agents have access to necessary information and can find this information quickly and easily. If the agent doesn’t know how to answer basic questions like, “How quickly will the product be shipped?” it’s unlikely that your FCR goals will be met.
First-call resolution is the holy grail of call center effectiveness. No other statistic comes close to its impact on performance. Applying speech analytics to conduct structured searches of agent-customer conversations, perform root cause analysis and identify areas in need of improvement throughout the enterprise will make your call center leaner, more efficient and more profitable. When a customer’s first call is their last call, everybody wins.