MARKET REPORT AI and Machine Learning
Portals to Success
© sdecoret-stock.adobe.com
In the channel, we have
become accustomed over the
years to having user portals;
web based interfaces where
resellers and users can order and
congure products and services.
ese ‘portals’ have developed
into ever more sophisticated tools
that can provision and bill these
services down to a micro level,
for example the turning on of a
voicemail box for one extension
for a week.
At the same time we have
seen the rise of the bot, a piece
of software designed and created
to automate the kind of tasks we
would usually do on our own,
like making a reservation, adding
an event to your calendar or
displaying information.
And then there are chatbots.
ese are used most commonly
in the customer service space,
assuming roles traditionally
performed by living, breathing
human beings such as sales
Judging from the conversations we are having, machine learning and especially
the use of bots, is surging ahead in the field of user interactions with suppliers of
services but what is it that customers want when interacting with suppliers?
support sta and customer
satisfaction reps. ey work by
processing the text presented
to them by the user (a process
known as parsing), before
responding according to a
complex series of algorithms that
interprets and identies what the
user said, infers what they mean
and/or want, and determine a
series of appropriate responses
based on this information.
Chatbots are useful for
frequently asked questions but
can become very frustrating to
use when straying o-piste with
unusual requests. Inevitably, in
these circumstances, they resort
to transferring the call to a real
person.
Here the smart money has
to be upon using machine
learning to reduce these human
interventions to a lower, more
acceptable and commercially
attractive level to eke out every
piece of productivity that can be
garnered for these tools. ey are
there not just to improve service
and customer levels but to also
improve business protability.
As Emily Nerland of Masergy
says, “A contact centre that needs
optimisation or a well-dened
social marketing strategy - bots
are great for contact centres that
have a high volume of calls, but
the cost of a powerful bot means
that they are usually limited to
larger organisations, with contact
teams in the double digits.
For marketers, bots can be
a great way to customise the
customer journey and make
it interactive for the prospect
without using physical resources.
Again, Bots aren’t cheap.
erefore, having a realistic plan
of how your bot will provide a
return on investment is vital.”
Nerland says that Bot users
need to have an extensive data set.
“How do you want your bot to
sound? Bots need to learn how to
use your business’ language and
how to refer to your products,
so having as many transcripts/
recordings as possible, (as well as
a clear idea for tone of voice) for
the bot to learn from is advisable.
And have clear expectations.
Bots aren’t going to be able to
wax lyrical about how amazing
30 | Comms Business Magazine | December 2019 www.commsbusiness.co.uk
/sdecoret-stock.adobe.com
/www.commsbusiness.co.uk