seat58F
outsourcing relationships: international, inter-economic, interpersonal
September 4, 2010 ~ 10:45:29 AM * -07:00ST
get in there
Post # 309 by
rc on April 27th, 2009 ~ 08:12:29 AM
Posted as csat, knowledge |
No Comments »
Level 3 agents who actually share Level 1 duties have a realistic view of the morale of Level 1 and the effectiveness of the knowledge db. Organizations that rely on the creativity of Level 1 engineers have a big problem with respect to recruitment and retention, constant problem solving is frustrating if the organization has an ineffective feedback loop into the services development process.
The support organization should have a strategy for content development and maintenance of the KDB. Level 3 can provide a realistic assessment of KDB effectiveness and is in a position to close gaps.
csat positive correlation
Post # 303 by
rc on April 26th, 2009 ~ 07:59:45 AM
Posted as csat, incident |
No Comments »
csat is positively correlated to active end user support by Level 3. If these guys view themselves as mentors to Level 1, feeling some sense of ownership to Level 1 overall performance, solutions are higher quality and re-work rates drop in a major way.
on the other hand, if Level 3 is viewed as an out by Level 1, agents and supervisors are missing the point.
(http://www.historyofpia.com/tails.htm)
statistics are your friend
Post # 173 by
rc on April 2nd, 2009 ~ 10:53:08 AM
Posted as csat |
No Comments »
using sampled data to evaluate an agent’s performance must be done extremely carefully. the data must be treated with respect and the resulting information must have validity attributes attached to it - this is where statistics come in.
usually we’re talking about the domain of someone playing a quality manager role. that person and the supporting processes needs to ensure that the information is developed in an unbiased way. the approach should generate statistically relevant information. sampled numbers must always be conveyed with a confidence interval. if sampled csat data is published in graphical form, clearly illustrate the confidence interval, please don’t bury it in the legend or footer.
one intro to statistics resource: Statistics in a Nutshell: A Desktop Quick Reference
IVR ergonomics
Post # 159 by
rc on March 31st, 2009 ~ 11:25:56 AM
Posted as csat |
No Comments »
start with the assumption that inbound callers are frustrated by the IVR and try to limit that exposure to negative cu-sat. abandonment rates give a general idea of call routing effectiveness; usually this metric does not include early termination while the caller is interacting with the IVR. get your voice people to gather and report these disconnect numbers.
the quality people can also monitor IVR sessions. callers tend to talk to the IVR even though its a keypad UI. monitoring the call at this point yields interesting insights into the perception of the service provider and/or the supported products.
as always, be aware of privacy concerns when recording these sessions.
