Opex Chapter Meeting-Symbios Consulting
While lots of companies race to implement new technologies (Artificial Intelligence, digital self-service, data analytics), they overlook obvious and easy ways to fully utilize what they have.
-Opex chapter meeting video on youtube:
https://www.youtube.com/watch?v=2AUMl4nC0UU
- Symbios Consulting Youtube Channel: https://www.youtube.com/channel/UCx7HKNnSIlEEE5JKI4JsA4A
2. Index
2 20 June 2019
• Intro
• Strategy for prediction.
• How to Predict.
• Results of predictive models.
3. 3
Profile
16 years of overall experience across variety of markets, led deployment of experience strategies in
the gulf region with heavy exposure to the overall EMEA region.
Certified Lean/Six Sigma Master BlackBelt.
Digital strategist certified by Google (squared online), understands how to build cutting edge digital
strategies and bridge them to the physical world to ensure maximum return.
Winner of more than 20 CX awards including, Best CX Brand, best CX Transformation Program and
Best Customer Service Strategy to mention a few.
Ahmed Nour
20 June 2019
5. 20 June 20195
“ Predictive Analytics is the process of learning from historical data in order to
make predictions about the future.”
Insight
Maturity of analysis
FuturePast
“Sense & Respond” “Predict & Act”
Hindsight
Insight
Foresight
Reporting
What Happened?
Analysis
Why it happened?
Predictive
What could happen?
perspective
What should we do?
7. 7
Four steps to formulate your strategy
Service Demand Contact Strategy Customer Preference Customer Management
“Who is going to contact you
and why?”
• Predict which customers are
in demand for service and the
root cause for their issue(s).
“Which customers should we
act on and which should we
“sleep”?”
• Predict which customers
need to be proactively
contacted to deflect/ avoid
issue.
“What are the
right time and right channel to
engage with customers?”
• Predict the right time and
channel(s) to engage with
customers for both inbound
and outbound contacts
“What should we do
to best serve our customers?”
• Predict the right service
action in response to
customer demand
Step 1: Who & Why Step 2: Act or Sleep Step 3: Right Channel / Time Step 4: Right Action
20 June 20197
8. 20 June 20198
Proactive Reactive
Proactive actions aiming to reduce
frequency of contact (FOC)
Maximize efficiency when receiving inbound
contacts and optimize experience
All customer interactions would be more effective if
we could know upfront which customers are affected
by an issue and why
21. 21
Next Best Action
doesn’t mean
offering fries with
every order, or…
Network & Usage
Customer
Interactions
Handset &
Tenure
We take
everything
we know about
and
Only offer
what is relevant
NBA
20 June 2019
21
22. 22
Target TNPS with NBATNPS without NBA
+71%+53%
Made 20 million
USD from
previously
unmonetized
channels
20 June 2019
22
23. 23
Unexpected Extras
Operation
People
1 second improvement in AHT
SLA improved by 2 points
Abandonment Rate improved by 1 point
Employee engagement improved by
8 points
20 June 2019
23
24. 24
92 96 97 105 116 120 151 163 181 197 211
275
364
364
405
560
507
560
616
6.6 6.8 6.6 6.7 7.1
6.9 7.8 8.6
9.5
10.6
10.2
9.9
9.9 11.0 11.4
10.5 14.4
15.4 18.6
0.0
5.0
10.0
15.0
85
185
285
385
485
585
Sept'16
Oct'16
Nov'16
Dec'16
Jan'17
Feb'17
M
ar'17
Apr'17
M
ay'17
June'17
Jul'17
Aug'17
Sep'17
Oct'17
Nov'17
Dec'17
Jan'18
Feb'18
M
ar'18
Thousands
Unique Customer and Freq of Visit
Unique customer Frequency of Visit
Prediction on it’s own is not really useful unless you know what
to do with predicted data!
25. Contact by Design … Complexity & predictability
25 20 June 2019
High Value Medium Value Mass
50%
Chat&AppIVR&USSD
10%
20%
10%
10%
10%
20%
20%
40%
* 100% outbound * 20% outbound
10%
10%
10%
40%
15%
10%
5%
Mass
Market
26. What was also made possible by prediction is drastically
lowering cost to serve by using a RFT commercial model.
26 20 June 2019
27. 27 20 June 2019
Not Scary Multipurpose Useless on its own
Prediction is…