Nicole Millard, BT: Day 2 13:50

Botman vs. SuperAgent: AI & Customer Service.

Dr Nicola J. Millard, Head of Customer Insight & Futures,BT Global Services Innovation [email protected] | @DocNicola

From East Coast to West: the Story Starts Here in 1990.

From BT Labs to Burnham-on-Sea

A Brief History of Artificial Intelligence in BT....Expert Systems for Customer Facing Environments (ESCFE) – 1990-93

Use

r int

erfa

ce –

mul

timed

ia m

ixed

initi

ativ

e di

alog

ue

Infe

renc

e en

gine

– n

eura

l net

& N

LP

Com

mun

icat

ions

– re

triev

al &

inpu

t

Product/service knowledgeProcess knowledge

Policy/RegulationCustomer careSystem knowledge

More accurate & consistent

call handling

Quicker access to expertise, regardless of skill level of customer or advisor.

Thanks to Dr Tony Houghton, Mike Ward & Dr Michael Gardner.Copyright: BT Global Services, 2016

Accommodating different levels of expertise in customer & staff.Dialogue created dynamically according to information from customer and/or from database.

A comms expert might offer information which is entered directly into the form window.

Has any testing been done?

Automatically eliminates certain questions based on the answers to previous ones.

Form window allows customer to offer information when they want – rather than sticking to a rigid script – and dialogue paths are modified accordingly.

OR: the advisor is guided by the dialogue window as to what questions to ask a less expert user.

Can you locate the modem?

Enters information from dialogue into multiple databases on the back end.

Copyright: BT Global Services, 2016

A Brief History of Artificial Intelligence in BT....Epsilon, 1993-1995, Newcastle & Thurso.

Use

r int

erfa

ce –

pus

h di

alog

ue a

nd

navi

gatio

n lin

k to

pag

e

Infe

renc

e en

gine

– n

eura

l net

& N

LP

Com

mun

icat

ions

– re

triev

al &

inpu

t

Product/service knowledgeProcess knowledge

Policy/RegulationCustomer careSystem knowledge

Push dialogue &

links to back end

information

Advisor gets quicker

access to relevant

information. Nat

ural

Lan

guag

e Vo

ice

anal

ytic

s



AI in Action at the Customer Interface: Making things easy by curating choice.

"I don’t foresee anything replacing customers calling for help, but their contacts will become more complex. We may also have to reach out to them more often”,

Manager, CCMA Ireland.

Primary function of contact centres in 2020.

Copyright: BT Global Services, 2016Source: Millard & Alcock (2014), SuperAgent 2020, BT White Paper

When things go wrong, which channels will we be using in 2020?

Copyright: BT Global Services, 2016Source: Millard & Alcock (2014), SuperAgent 2020, BT White Paper

Challenges:• Works for simple queries – effectively steering the

customer through the “known knowns”, i.e. GIGO.• More difficult for complaints/complexity - if we

get angry, we tend to tell long, rambling stories which are hard to parse.

• Sarcasm tends to throw algorithms: e.g. “Thanks, @TrainCo for my free sauna this morning”; “It was so good to see that your maintenance department hadn’t spoiled things by making unnecessary repairs”; “I would have cheerfully strangled them”.

• Limited ability to parse emotional context - YET!

Chat is where it’s at: are chatbots “IVR for digital”?


Screenshot: Digital Genius

10

Robo and other automated advice only appeals to 1 in 6 consumers.

Automated advice - you would be presented with a set of multiple choice

options (for example your appetite to financial risk - High Med, Low) and as

you enter your answers the system dynamically changes to ultimately

providing a recommendation/advice based on your input

Might be useful

Really add value 14%

43%

Financial service online games - Fun tools that help you with money, for

example Apps that set goals for savings, give you tips on better driving

for cheaper insurance or help you budget weekly and games which

teach children about money and help first time buyers understand home

ownership

Might be useful


36%

Robo-advice - where you get basic advice about financial

services pensions and investments from a government

regulated online or voice recognition service, but the cost

to you is a fraction of an independent financial advisor

Self-help videos about financial services products - impartial and

useful information about banking, insurance, pensions and

investments which you can watch in your own time

Might be useful


40%

Might be useful


53%

Copyright: BT Global Services, 2016Source: Davies/Hickman (2016), Youbiquity Finance, BT White Paper

Managing social media: DebateScape.

Thanks to Dr Simon Thompson .



"Customers are calling us about an issue because of the complexity of the world we are living in. We have termed this a ‘contact centre job’ – with people being paid the same wage as ten years ago. That is

unsustainable – you will not get expert problem solvers who are shifting organisations in six monthly cycles – they are call handlers – we need expert problem solvers”,

CCMA UK, Contact Centre Strategist.

Super agent skills 2020.

Source: Millard & Alcock (2014), SuperAgent 2020, BT White Paper


Intelligent skills based routing regarded as vital to get customers to the person with the right skills. Cloud and big data analytics are regarded as less important but are key (and often hidden) technologies underpinning flexible contact capabilities & intelligent call steering.

Key technologies 2020.

Source: Millard & Alcock (2014), SuperAgent 2020, BT White Paper

Being Proactive using big data and AI: Aperture.

Thanks to Dr Detlef Nauck .

• Interactive models of customer behaviour

• Data from customer service processes & customer feedback.

• Ability to build in other related data, e.g. weather forecasts, Internet of Things.

• Identification of drivers.

• Ability to build “what if” scenarios, e.g. volumes, staffing levels, backlogs etc.

• Used for churn & complaints prevention.


Proactive analytics and Security: Saturn.

Thanks to Dr Ben Azvine.

Enabling automatic notification of:

• Agile fusion of information feeds from multiple sources

• Artificial Intelligence based knowledge management and interactive visualisation

• Interactive Human/Machine data analysis for identifying patterns or links

Failures

Anomalies

PotentialThreats

Attack


44%Would share social media profile to get better service

4 in 5Like orgs using technology to listen to calls if the agent is rude or a problem arises

71%Like it when organisations notice consumers have problems on their website and try to help

55%Might share location with organisation for a better service

The emerging “ego” system:

62%Find dealing with customer service issues exhausting

55%I put a lot of effort into safeguarding my rights

54%Give feedback to companies but they don’t change

The more information I give, the better customer service I expect

48%China

63%UAE

75%India

62%Singapore

66%USA

69%Indonesia

63%UK

56%Belgium

57%Germany

74%Spain

Source: Davies/Hickman (2015), The Autonomous Customer, BT/Avaya

Creating smart customer service: it’s all about me!


Botman + SuperAgent.

“Hollowing out” & “Uberisation” of work

Pattern matching & unstructured problem solving, e.g. Watson,

Digital Genius.

Chatbots, e.g. Siri, M, Cortana, Amelia and many, many more.

Robot customer service – e.g. Pepper, EMIEW3

Robotic Process Automation.


Thank You.Dr Nicola J. MillardHead of Customer Insight & FuturesBT Global Services Innovation [email protected]@DocNicola

Nicole Millard, BT: Day 2 13:50

Technology

Transcript of Nicole Millard, BT: Day 2 13:50