When there’s a problem, your help desk is you: negative, positive or neutral.
Your sales people worked hard to acquire that business, but it’s help desk
performance that will determine whether or not you retain it.
The business rule we all know is that you can only expect . . . what you inspect.
Yet many people choose, and retain, a help desk with far less “due diligence” than
other “mission critical” services.
The work of the Help Desk starts with an incoming call, answered by a person
or ACD – (automated call distributor.) It is critical that you call your help desk
periodically. It’s equally important that you have someone else, less knowledgeable
and coming from a different perspective (a teenage son or daughter works well)
do the same.
The purpose of this exercise is to insure that callers at all levels of expertise have
a satisfactory experience. You, as the industry insider, can pose difficult issues,
and expertly evaluate responses, but your wanting it to work makes you less
than totally objective. The coached teenager is an excellent mimic for a major
segment of the call-in population: young, and often technically savvy (at least
in their own minds) but often not clear on terms, troubleshooting, processes –
So, point number one is make the calls, twice a year at minimum, and fine tune
as needed. If there’s an ACD, make sure the “3X3 rule” is followed. Three choices,
each with 3 options, and always a way to zero-out. Customers are not there to be
manipulated. Don’t deny them an operator when they call for one.
Side note: if the help desk you are using, (or, heaven forbid, considering)
forces callers into a “no-zero” system, dump them. They have no clue what
customer service means. And if you encounter this mistreatment elsewhere,
visit http://www.gethuman.com/. It provides instructions for short-circuiting
the irritating closed-loop menus of hundreds of companies.
Point number 2, of course, is evaluate the responses you receive during those calls.
Answers must be understandable, and executable by a lay person, with reasonable
speed and accuracy. Do they have a pleasant voice, speak English fluently, and use
correct grammar? Do they use the correct words and terms, and speak in terms
the caller understands, rather than industry jargon?
Listen for the tone of voice (enthusiastic, not giddy) what they say, and how they
say it. The best customer service agents have a voice of “gentle authority” that
allows them to manage and control the flow and direction of calls. They also speak
from a flexible position (“Here’s what I’m hearing”) that is never confrontational.
And rather than barking orders (“you need to . . .”) they offer suggestions,
(“here’s what I can do.”)
It’s wise to round-out your research by surveying customers on their help desk
experience. Not only will it yield important insights, but it demonstrates a level
of quality control on your part that is exceptional.
I want to return to the subject of ACDs for a moment. As a provider of help desk
services, I evaluated various automated systems, and their overall effectiveness.
My take-away was that standards within the help desk function run a broad gamut:
terminal calls vs. order calls vs. simple inquiries. To be most effective, calls must
be routed very carefully –difficult to accomplish with an ACD.
After routing, evaluation of talk time, a common matrix, must take into account
that order calls (“I need roll paper and also my account balance?”) can often be
completed in a minute or two, where a technical issue (“printer down!”) can require
a half hour to resolve. (The mid-range question, I suppose, is something on the
order of, “what does that 'nd’ on my display mean?”)
You can’t combine these for analysis purposes, and there’s no relevant industry
standard to help you: the banking industry does not provide technical assistance
with hardware and software issues. So, if you want to evaluate your help desk
performance on technical issues, you need to look elsewhere for an applicable
standard. Trouble-shooters for computer systems or copiers, or other products
with technical content, are good choices.
Consider, too, whether you’ve you empowered them with sufficient authority to
be effective. This is tough to do, because it involves relinquishing control. If you
have, at what point does that authority end? Is your help desk empowered to
spend $15 to resolve a major issue? $100?
Also, who does the help desk contact when the call moves outside the parameters
you’ve established? (An example is a merchant demanding a free terminal when
the policy calls for a swap-out.)
I’m big on empowerment, and believe it should be in writing, clear, and not prone
to assumptions. It can be fluid, and the rules can change, but the core document
must be clear. For most companies outsourcing their help desk function, the
document is a Service Level Agreement.
In addition to levels of authority, it typically covers “time to answer” and other
operational issues which, frankly, and are not based on anything except what
will fly. The point here is that customer expectations, rather than “industry
standards” or “what will fly” are what matters.
One thing CardWare does to measure performance is to survey 20% of all
help desk calls. We ask if the phone was answered in a timely manner, and if
the problem was resolved satisfactorily. We understand that customers –
not our expectations, and certainly not some standard - are what matter.
Another help desk evaluation matrix Is tenure. Is the help desk position a
stepping stone at your provider company? If it is, turnover will be constant
and experience and knowledge levels lower than you’d like.
Which brings me to the topic of knowledge access. Information, particularly
complex information, may not be immediately available from the help desk.
What’s important is that it be quickly accessible, accurate on a call-to-call basis,
and accurate between help desk agents.
OK, so your help desk is on-track: well, trained, motivated, knowledgeable.
The way to keep it that way is regular reporting. Whether your preference is
weekly, monthly or daily, you’ll want to see first, an Itemization by client –
merchant by merchant. If one merchants calls three times in one month,
you want to know why. You also want a database you can query. And you’ll
want a list of customers, dates they called, the time the call was taken
and the time it ended.
I also want to know about all calls over 15 minutes.
I also want a record of who took each call, and the reason for the call. This is
do-able with simple coding. The nomenclature must be consistent so that I can
look at all the printer issues together, for example, and also drill down for a
particular printer model if needed. Finally, I want to know how the call was resolved:
programming of the terminal, replacement of equipment, a call to the processor?
The main benefit of a good report is that it identifies trends quickly. If there’s a
specific problem with a terminal we haven’t seen before, there may be a looming
hardware issue. And that information will help us both resolve the immediate issue,
and plan for future equipment purchases – and recommendations. If there are
calls from new merchants within a few days of start-up, we may have a training
issue. This is important because good training reduces help desk costs. If there’s
little or no investment in training, understand that the help desk will be more critical,
and more involved, all the time.
So, get maximum value from your help desk. Remember it’s human nature to
pay attention to what you inspect – not what you expect. You can’t manage
what you don’t know, so regularly measure performance. Even if everything is
great, they will know you care, and are listening. Help desks are like nursing
homes. Over time, without monitoring, quality declines. With a little attention,
service stays where everyone wants, and needs, it to be.
Biff Matthews is President of Thirteen Inc, the parent company of
CardWare International. He is one of 12 founding members of the ETA,
serving on its board, advisory board and committees. (740) 522-2150