A

project on
service time variation
in Café coffee day
SUSHANT
Structure
Service time

Project methodology

Analysis

Observation and recommendation

What do we sell?

Experience

◦ Physical product

◦ Ambience

◦ Service
Quality of service depends on
following aspects:

1. Quality of customer and staff interaction

2. Quality of complaint handling

3. Optimum service time

Optimum service time
What is it?
 Forthis project, we have taken it as the time
interval between taking the order to the service
of order at the table.
Is it important?
◦ Type of customer
◦ Time of the day
◦ Format of the outlet
◦ Time period between the order is taken and the
order is served
◦ competition
Research methodology
Data collection

Sources of data

Sampling
Sources of data and collection
Primary
 The primary data was collected by putting timing
of order received and service timing on the KOTs
of the respective order.
 Data collection was done for 22 days in the S P
Road outlet, Hyderabad.
Secondary
 Forsecondary data, I referred to the websites of
Café coffee day, Google, Wikipedia.
 Books Statistics for management (Levin &
Rubin),Principles of retailing(J Fernie).
 Papers : Café coffee day and Barista, a
comparative study, ASBM, Bhuvneswar.
 Sampling
Sampling was done randomly and 50
samples (KOTs) were taken daily.

Extraneous effects were removed by

removing or not choosing very large orders
and very small order.
All the orders were also noted with the
time of delay due to factors other than
considered.
Analysis
Statistical process control
◦ X bar chart
◦ R bar chart

Total quality management

◦ Fishbone diagram
◦ Pareto chart
X- bar chart
 X bar charts are the control charts for
process means. The control limits in X bar
charts place bounds on the amount of
variability we are willing to tolerate in our
sample means. In X bar charts we have  
A center line X bar-bar or X grand mean
An upper control limit (UCL) line, with
value = grand mean + 3σ
A lower control limit (LCL) line, with value
= grand mean - 3σ
R- chart
InR charts, we plot the values of the
sample ranges for each of the samples. The
center line for R charts is placed at R bar.
UCL (upper control limit) = R bar (D4)
where D4 = 1+3 d3/d2
LCL (lower control limit) = R bar (D3)
where D3 = 1 – d3/d2
D4, D3, d2, d3 are constants.
Analysis
 X
14
– BAR CHART
12
12
11 10.8 10.8

10
(UCL
7.8
8 )
Grand mean6.5
6.01 6.3
7.52
6 5.28 5.34
(LCL)

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
ANALYSIS
 As we plot the CL, UCL and LCL and the daily
values of the X bar, a quick glance at the chart
shows us that on the dates 15/05/10, 18/05/10,
22/05/10 and 30/05/10 the average serving time
jumps above the UCL(upper control limit).
 To check whether the cause for this variation is
assignable or random, close investigation shows
that there is some relation between sales on that
particular day and the delay in service.
 Now we can see clearly that whenever the sale is
more than Rs.20,000, it is affecting the service
time and the order is served late. Now we will try
this same by excluding the outliers or the values
above the UCL.
X – BAR CHART (excluding outliers)

Series 1
9
8
7
6
5 Series 1
4
3
2
1
0
1 2 3 4 5 7 8 101112 14151617181920 22
X- BAR CHART (without outliers)
From the above chart, we can see
that once we remove the outliers
the service time is well within control
limits.
R – BAR CHART
Series 1
UCL 12.535
14

12

10

4 LCL 3.758

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
R BAR CHART
The values of the daily ranges are in
the control limits. This shows that
there are not great differences
between one day range to another
day range, but these range values
have to be compared with another
competitor to get the advantage of
this chart.
FISHBONE DIAGRAM(Total quality management)
The TQM approach to any business starts with
the realization that all errors, defects, and
problems have causes and there is only finite
numbers of these. The fishbone diagram
takes an unstructured list of factors that
contribute to delayed service and organizes
that list in two major ways. First, it gathers
the factors into logical groups. And then,
within the groups, it indicates how the
various factors feed into one another in cause
and effect relationship.
 FISHBONE DIAGRAM
CUSTOMERS
inside counter equipment
operations
Sudden breakdown D
less skilled team members e
low awareness no time constraints for order prep. l
Low efficiency a
No audit or checking for service time y
Less number of team members e
d

Less skilled TM skills S

e
r
Less no of TM v
Recruitment
process training i
Wrong order c
commitment
e
no serving time check provision

Compensation Bonus

Service area operations team members

 PARETO CHART
From the above fishbone diagram we see that
there are several factors that affect the
service time.
In TQM, we distinguish between vital few and
trivial many. So that, we can solve the vital
problem first, and then the next vital one, it
is a continuous process of improvement.
To identify the vital problems, we use Pareto
chart. A Pareto chart is a bar graph showing
groups of error causes arranged by their
frequencies of occurrences.
 PARETO CHART
FREQUENCY OF FACTORS
80

70

60
F
R 50

E 40
Q
U 30

E 20
N
C 10

Y 0
S S S S S
ON ON ENT MER BER
I I M
A T A T PM ST
O
ME
ER ER Q UI U
OP O P E C AM
R EA TE
NTE AR
U
CO I CE
ERV
S
PARETO CHART
As from the Pareto chart, the data
shows that 60% of delay can be
attributed to inside the counter
preparation, 20% to service area ops
and rest are not significant now. Thus
in the TQM parlance, as we say slay
the dragon first, we can concentrate
on the inside counter part of the
operation first.
Conclusion & recommendation
Through continuous observation, it is
observed that there are no control
measures for service time.
The variation in the serving time is
out of control
The variation is assignable and not
random.
 Conclusion & recommendation
Introduction of service time sheets for
customers: these sheets will be given to every
customer at the moment the order is taken.
These sheets will show the time that a
standard order should take and based on the
statistical calculations on an average the
standard deviation that may happen; it will
also show minimum time and max time an
order should take. We can make this format
interesting by inserting concepts like before
time, well in time, average etc.
 Conclusion & recommendation
Introduction of service time record in the
billing system itself: here in coffee day when
an order comes to the counter it gets fed
into the system, now if we can introduce a
automated barking system in the three
sections (cold, hot, food) by the means of
LED boards to display the orders it will make
the recording of service time possible. Once
the order is prepared and placed at the
counter the button on the board is to be
pressed to receive the new order from the
system.
Conclusion & recommendation
Introduction of better equipment
those can be helpful at the peak time
of business. For example, big ice
blenders which can continuously
blend ice for some while without
melting it, it will decrease the time
for making ice blended drinks.
Conclusion & recommendation
Reports for evaluation :
Having a digital recording system
for service time will provide us with
data that can be easily converted
into useful reports. These reports
when compared with sales report and
other reports over a period of time
will provide useful insights.
Thank you