Shouldn’t we
charge banks for keeping us too long in the queue?
By MUNGAI KIHANYA
The Sunday Nation
Nairobi,
17 May 2009
Ooonnne, Twooo,
Threee, Fooour…this is not a typo! It is how a bank teller was counting
the coins that I deposited to my daughter’s account recently. There were
about 300 pieces in various denominations, all sorted out and
pre-counted, yet this cashier took 10 minutes to confirm the count.
Now 10 minutes are
600 seconds, thus the cashier was taking an average of two seconds to
count each coin… Ooonnne, Twooo, Threee, Fooour…!
She was not the only
slow server; her colleagues were not doing much better, judging from the
time I spent on the queue. I was 15th in line and there were four
counters open.
It took 18 minutes to
get to my teller. In other words, the average service time was slightly
over one minute (72 seconds). But since there were four tellers, this
works out to almost five minutes per customer.
While spending five
minutes being served might not sound like a lot of time, we must not
forget that I had already waited for almost 20 minutes on the queue and
that was too long.
Perhaps the biggest
headache for a bank manager is to establish how long is “too long” for a
customer to wait on the queue: Is it 20 minutes or 5 or what? I pondered
on this matter a while ago and then decided to do a simple experiment.
I’d go into a bank,
join the queue and start the stop-watch in my mobile phone. I’d then
stay in line waiting until some one within my vicinity complained about
the long queue…something like, “these people are too slow, don’t they
think we have other work to do?”
At that moment, I’d
stop the timer and leave the queue. Then go to another bank and repeat.
I visited five different banks and the results were amazingly similar:
the average time was 11 minutes, give or take a few seconds. For a bank
manager, this is simply 10 minutes.
Now, banks operate on
the assumption that they can manage the queue by monitoring how fast
individual tellers work. Unfortunately, this is only one half of the
queuing problem.
The amount of time a
customer is likely to spend waiting in line does not depend on speed of
service alone: It is also affected by the rate at which customers are
arriving in the banking hall.
The few managers who
are aware of the two factors also make another equally invalid
assumption: that the queue can be eliminated by equating customer
arrival rates to service speeds.
That sound
reasonable, but statistical simulations reveal that to control a queue,
the service speed must be higher than the customer arrival rate. How
much higher? Well, that is a story for another day; for now, think about
this true incident…
In 1999, a magazine
in
 |
 |