CALLER SATISFACTION MEASUREMENT

According to you, can we measure customer satisfaction for debt recovery departments ?

This has been tried, but to no avail. Customer satisfaction in a collection call center is an elusive item and probably not worth measuring...
--Dr. Jon

Are there any best-practices for determining caller satisfaction/completion rates from an IVR? For ex, if someone hangs up after we have given them a URL but before we explain an application process, should that be considered an abandon or a caller satisfied metric? I have to hit a 20% caller satisfied within the IVR for a client, but there are disagreements as to determining an abandon versus a satisfied. Any information would be great.

This is a tough, but common question. We are all trying to “make our numbers.” Specifically, if a customer hangs up after you give them a URL for self-service, you should consider this a caller satisfied metric, after all, he/she got what they need to know. Callers usually do not abandon unless they wait much too long, or they get some semblance of an answer.

At our call center, we do monthly caller satisfaction surveys using a scale of 1 to 5, where 5 is a perfect score. No matter how hard we try, we are not able to get our percentage of perfect calls to go over 50%. Are we doing something wrong? How do others do in this performance category?

Although your scores are slightly below average, your survey results mirror our call center benchmark research. Callers are not easy to please. In fact, many respondents to post-call surveys simply do not give perfect scores. I suggest you add an open-ended question to your post-call survey that asks, "What is one thing we could have done better to improve your calling experience?" In working with call centers, I have found that this question results in many new ideas for improvement in call quality and delivery.

Dear Dr. Jon, Is there an industry standard for email and fax response times? And if so, what would it be? Sincerely, Peter Demaitre

Peter, the answer depends on the industry. If you consider all industries, the answers are as follows:

e-mail response time = 18 hours
FAX response time = 6 hours

-Dr. Jon
Answered: 01/10/05

From recent member discussions around customer satisfaction, first contact resolution, voice of the customer and more, it has been clear that members employ a wide variety of approaches/philosophies when it comes to the timing of customer/external measurement. Some advocate within 3 days as being sufficient to retrieve an accurate customer response, some longer and a few even suggest that real-time is the way to go.

In my book entitled “Listening to the Voice of the Customer” your members will find a rich array of methods to obtain timely and accurate customer feedback regarding their contacts with the company’s inbound customer service contact center. As is so true of almost everything in life, “timing is everything”.. and, with no exception, timing is critical in obtaining accurate customer feedback.

We have developed the Every Customer Has Opinions (ECHO for short) survey system for caller, e-mail, and chat customer feedback. Our research shows that the less time that elapses between the customer “contact” and the time you ask for feedback, the better. Regardless of how you get the feedback, i.e., IVR, e-mail, chat-window, we recommend that it either be immediately after the contact, or never more than 24 hours later...the simple fact is...customers forget the details and their answers change with time...immediacy is critical for statistically, and emotionally accurate customer feedback.

How did you come up with the number of 30 minimum completed surveys per rep to ensure statistical validity?

To be statistically valid at a 95% level of confidence with an accuracy of plus/minus 5%, you need 360 completed surveys from a random sample of participants (in this case callers). If you're looking for statistically
accurate feedback from callers at the agent level, 30 per month, equals 360 per year, and therefore you can statistically "quote" the voice of the call
in your annual reviews of agents.

How many call centers use ECHO?

Currently we have between 100 and 200 in the customer pipeline. Since we established our exclusive partnership relationship with Witness, the ECHO product/service has mushroomed. ECHO allows the customer to monitor and coach the agent…a unique concept that saves the call center tons of money.

Dr. Jon

If we were gathering customer feedback for agents on a team level only, how many surveys would we need to gather for each and what would be the associated cost?

At the team level, you would again need 360 completed surveys over the period of planned evaluation. Typically the ratio of agents to supervisors (teams) is fifteen to one.
Logically then, the previously quoted agent-level cost would be divided by fifteen to get the cost per team or supervisor.

In a BenchmarkPortal report it indicates that: "92% of calls were resolved on first contact versus 79% for other financial services companies." Does this mean that if there was a problem it was completely resolved during the first contact?

Percent “first contact resolution” is a key performance indicator for all call centers. It means that the caller’s issue and/or questions is resolved/answered on the first call, with no transfers and no call backs.

In regards to post call surveys, is there a "rule of thumb" as to how long (time) the surveys should be or how many questions we should ask in each survey? Should you limit the number of questions to avoid callers from hanging up early?

Length of time to complete the post-call survey and/or number of questions is definitely an issue in maximizing completed surveys...tell the caller up front how the survey will take and stick to that time or less.

From our experience, absolutely no more than 10 questions and/or less that three minutes.

-Dr. Jon Anton
Answered 02/10/05

In the BenchmarkPortal report on service recognition, it indicates that: "92% of calls were resolved on first contact versus 79% for other financial services companies." Does this mean that if there was a problem it was completely resolved during the first contact? (i.e. If a shareholder calls re clarification on language within a prospectus the service representative provides an answer during the call and does not have to call the shareholder back with an answer.)

Percent “first contact resolution” is a key performance indicator for all call centers. It means that the caller’s issue and/or questions is resolved/answered on the first call, with no transfers and no call backs.

Is abandon rate still a common performance standard for call centers? How does the following compare to industry standards for call center performance areas: Average Speed of Answer - 80% within 30 seconds Abandon Rate - 5% Response to inquiries - 90% within 3 business days

Abandon rate is not a common performance standard for call centers. It depends too heavily on the patience of the caller. ASA 80/30 is not particular great. ASA in 30 seconds if an average performance. Abandon rate of 5% is definitely poor across industries, and response of 90% in 3 business days is almost at unacceptable email levels. I suggest you get one of our industry reports for your industry and compare all performance metrics.

Is there a basic ratio of number of satisfied customers that send in positive feedback to dissatisfied customers that send in complaints? (For example 1 in 100 satisfied send in a positive feedback and 10 in 100 dissatisfied send in a complaint. Which would be a 1:10 ratio.)

Rick, this is an excellent question. In my research on customer behavior regarding survey response, I have found the following:

1. A customer’s willingness to respond to a survey following a customer-event (phone call, email, chat, and the like) is highly dependent on:

a. The loyalty of the customer to that company (for instance, eBay members are more likely to answer surveys simply because eBay has created almost a “cult” company whose members really believe in, and want to make the company better.

b. The ease with which the customer can switch vendors (for instance, a utility is more likely to get a higher response rate because customers can not change to another utility company. Therefore, the customers try to change the utility by responding to surveys.

C. And finally, if the customer is really happy, or really unhappy they are more likely to respond.

2. If you research a customer’s emotional disposition following a customer-event (phone call, e-mail, and the like) their most likely emotional disposition will fall into one of the following five categories: 1 = very dissatisfied, 2 = dissatisfied, 3 = neutral, 4 = satisfied, 5 = very satisfied.

Now, the probability of a customer responding to a survey following a customer-event is as follows (assuming they are in one of the emotional states described above):

1 = 90% probability they will complete the survey with a 1
2 = 30% probability they will complete the survey with a 2
3 = 05% probability they will complete the survey with a 3
4 = 30% probability they will complete the survey with a 4
5 = 50% probability they will complete the survey with a 5

--Dr. Jon
Answered: 01/10/05

Jon Is it possible to get a list of attributes to be included for a caller satisfaction study ?

Dharma,
A caller satisfaction study should include the following:

• Details by customer type (e.g., age, education, demographics, etc.)

• Direct, focused questions related to the customer’s most recent contact experience with your center, to include attributes such as:

oPrimary reason for contact

oOverall quality of service experience using a 5-point scale where 5 equals very satisfied, 4 equals satisfied, 3 equals neutral, 2 equals dissatisfied, and 1 equals very dissatisfied

oLength of time before reaching an agent

oLevel of satisfaction with agent’s quality of service (using the 5-point scale)

oCustomer’s satisfaction level with the agent’s politeness, courtesy, concern with your issue, willingness to help, listening skills, knowledge of the product or issue, and the agent’s overall ability to answer the customer’s question and/or resolve her issue (using the 5-point scale)

oLevel of satisfaction with agent’s quality of service (using the 5-point scale)

oNumber of times the customer has contacted your center about this specific question or issue

oDetermination as to whether or not the customer’s question and/or issue was resolved with this contact

oDetails on reason for dissatisfaction, if customer was dissatisfied with any aspect of the calling experience

•Collected by e-mail or after-call IVR survey.

A study we conducted showed that over 93% of companies surveyed collected customer opinions, but only 33% of those companies used the opinions to influence internal change. To make the utilization and leveraging of customer opinions easier, BenchmarkPortal designed Echo, a web-based application that channels customer feedback directly to front-line CSRs and to contact center leadership in the form of secure, meaningful, actionable reports. For more information, visit our Web site at and click on ECHO (Every Customer Has Opinions) under Customer Satisfaction Measurement.

-Dr. Jon
Answered 12/09/05

What is the benchmark for e-mail response that we should target for our center? What do customers expect?

These are actually two different questions. In our e-mail benchmarking studies, we find that companies with e-mail management software consistently respond to e-mail within three hours. By contrast, when we asked a large sample of customers what they expect in terms of an e-mail response, we found that 24 hours is more than satisfactory. This expectation will, of course, change over time.

What is the major difference that BenchmarkPortal offers with ECHO from the other Customer Sat. Surveys offered?

ECHO has the following features that differentiate it from others:
1. It was developed by and with real customers.
2. It is scientific and statistically based.
3. It drives results to the front-line agent in real time to produce change quickly and seamlessly.
4. It is the only system that uses the caller to “monitor” the call and drive this valuable feedback directly.

--Dr. Jon

What is the ratio of quality audits conducted per CSR on a daily, monthly basis to call volumes in order be confident of the quality being delivered to customers? Jo Carol Hutchins

Jo Carol, here is my answer:
1. For new agents, you should be doing a quality audit ratio of 10 per 100 calls

2. For experienced agents, you should be doing a quality audit ratio of 5 per 1,000 calls

--Dr. Jon

Would you happen to have any articles or research on the benefits of doing IVR vs. paper surveys?

I did a great deal of research in writing my book entitled, "Listening to the Voice of the Customer." I am sure a copy of this would help explain a lot of things.
Basically here are some of the differences to be aware of in comparing IVR surveys with paper (or mail) surveys:


1. An IVR survey takes place immediately after the call where as a paper survey typically takes place a week after the call.

2. IVR surveys, in volume, cost about $3/survey; paper surveys can cost as much as $5/survey.

3. IVR surveys can have a 40% response from callers; paper surveys seldom have more than a 5% response from callers.

4. The IVR survey is completely automated; the paper survey is completely manual.

5. IVR survey results are available immediately, whereas the paper survey results may take three to four weeks.

6. IVR surveys allow for easy branching of questions, but the paper surveys do not allow for any branching.

7. IVR surveys can include the "actual" voice of the caller; paper survey responses only contain written comments.

8. IVR surveys take advantage of modern technology; paper surveys depend on snail mail service.