Contact Center Survey Summary

By Peter DeHaan

September 2010

ContactBabel recently released the third edition of its The US Contact Center Operational Review.  Last year, in the March 2009 issue of Connections Magazine, I highlighted some data from the "outsourcing" section of their second edition.  This time I want to report on some of the broader findings.  Of course, you can study the 251-page report yourself, but gleaning the highlights here will be much quicker.  

For the 5 percent of Connections Magazine readers who work outside the United States, do not dismiss this report.  If you serve the US market, this information is most applicable to your work, and if you don't serve the US market, know that many US contact center trends and realities still have relevance for your center.  To include as much information as possible in this short space, I will be concise.  Here goes:

Quality: The majority of respondents said that the most important measure of quality is giving correct information to the customer; second is providing a consistent experience.  Interestingly, complying with regulations ranked fifth out of five.  This is not to dismiss compliance, but to indicate that most do not see compliance as a chief quality indicator.  The lack of investment in systems was cited as the biggest challenge to providing quality.

Scripting: One third of contact centers do not use scripting; 20 percent use it only for "new or inexperienced agents"; and almost half use scripting for all agents.

Quality Assurance Methods: The top QA method is call recording, followed by call monitoring.  The use of performance dashboards was third, followed closely by customer surveys (the only external measurement on the list).  Interestingly, call scripting did not rate that high as a QA method.  This might that suggest that the benefit of scripting is primarily for consistency purposes.

Customer Satisfaction: Customer satisfaction (c-sat) is primarily tracked by one of three methods: an IVR survey at the end of the call, a post-call mailed survey, or a post-contact outbound survey call (sometimes by a third party).  Overall, c-sat increased slightly from the prior survey, with 66 percent giving it the highest rating.  However, c-sat varied greatly by vertical market, with insurance leading the way (outsourcing was second, but trailing considerably).  Verticals with the lowest c-sat ratings were transportation and travel; healthcare was the second lowest.

Agent Training: The agent training method most often deemed as "very useful" was live call taking, followed by mentoring (which was the highest esteemed when also considering those who also rated it as "useful.")  The buddy system also received good ratings, but e-learning and classroom lectures did not.

Becoming Fully Productive: It takes almost seven weeks for an agent to become "fully productive."  The time for inbound agents was slightly longer at 7.5, whereas outbound was a relatively short 5.3.  Interestingly, mixed agents (doing both inbound and outbound) become fully productive quicker than those doing just inbound, at 6.7 weeks. 

This also varied considerably by vertical market.  Insurance was by far the longest at sixteen weeks.  (Recall that the insurance vertical also had the highest c-sat ratings.)  The lowest time to become fully productive was healthcare at four weeks.  (Likewise, note that healthcare had the second lowest c-sat t rating).  Interestingly, outsourcing had the second lowest time for agents to become fully productive at 4.2 weeks, but also enjoyed the second highest c-sat ratings.  Clearly, the outsourcing vertical has the ability to train quickly and well.

Ongoing Training: On average, four hours a week is giving to ongoing training, over half of which is coaching, with the remainder split between e-learning and classroom.  This training is focused on three areas, with roughly equal emphases: soft skills/behavioral, internal systems/processes, and product/marketplace changes. 

Agent Activity: In considering how agents spend their time, 59 percent is spent on calls, with 15 percent on wrap-up activity, 10 percent on administrative and training functions, and the remaining 15 percent as being idle.  Presumably, paid breaks and lunches were excluded from these numbers.

Call Stats: Although the variation is significant between verticals, the average call length for customer service calls is 5.6 minutes, with sales calls taking slightly longer at 6.7 minutes. 

Abandonment rates average 6.1 percent, with a median rate of 4.0 percent.  The mean rate is much higher than the median rate because of some centers with "very high call abandonment rates."  It is noteworthy that call abandonment rates are higher in smaller contact centers and lower in larger centers, proving the "economy of scale" that contact centers enjoy.

The average speed to answer was 29.1 seconds, with public service (thankfully) leading the way at a mere 10 seconds; healthcare was second at 13.7.  The technology, media, and telecommunication verticals, as a group, was at 71.7 seconds. 

The first call resolution rate averaged 75.9 percent. 

Cost per call is a problematic metric to ascertain, with great variability within and between sectors; however, for inbound centers, the mean average cost per call was $6.80 and $9.10 for outbound.

Contact Center Costs: Not surprisingly, agent compensation was the highest cost item at 53.3 percent, with non-agent compensation at 15.7 percent.  (Therefore, staff compensation is 69 percent of a contact center's total expenses).  Telephone costs were 8.8 percent; IT, 7.2 percent; rent, 4.9 percent; utilities and taxes, 4 percent; "other" expenses, 6.1 percent.  Surprisingly, there was not a great deal of variation between centers of different size.

Multisite and Virtualization: Forty-three percent operated multisite contact centers.  The degree of virtualization varied by contact center size, with small centers most likely to operate stand-alone (60 percent) and large centers most likely to be virtual (62 percent).  Overall, 29 percent of multisite operations were virtual; 43 percent, stand-alone; and 29 percent used a mix of stand-alone and virtual.  Reasons for virtualization include improved ability to handle call spikes, reduced call queues, more equitable agent utilization, increased agent skill sets available to callers, and cost savings.  The top reason for not going virtual was concerns over managing multisite teams.  This was closely followed by apprehension over cost and complexity. 

Work-at-Home: A related issue to multisite and virtualization is work-at-home, with respondents' views being significantly more positive regarding the option than in the past.  Over one third, at 36 percent, are using some home-based agents.  Large centers, at 53 percent, are more likely to use home-based agents, with small centers' adoption rates at 33 percent.  Interestingly, medium-sized centers trail in this trend at only 20 percent.  Noteworthy is that those not using homeworkers cite effective agent management concerns (57 percent), whereas those actually employing homeworkers are much less anxious over managing them (23 percent).

These are the highlights from the first 100 pages of the report; look for more next month!

To read other articles written by Peter DeHaan, go to From The Publisher or check out his blog, Musings of Peter DeHaan.  In addition to publishing Connections Magazine and AnswerStat magazine (for healthcare call centers), Peter also publishes several websites, including ArticleWeekly.com. He may be reached at 616-284-1305, dehaan@connectionsmagazine.com or the Peter DeHaan Publishing website.