1. CALL CENTRE MANAGEMENT BASICS
The Planning and Management Process of Call Centers
Ice-breaker:
Introduction to the course – call center is a unique environment; The session is about pulling together a lot
of different concepts and putting them into perspective so we can all understand the part we play in its
success.

2. Introduction 6-hour course - 3 hrs – 2 sessions PLEASE TURN OFF YOUR
MOBILE PHONE NOW

3. Course Outline Key Topics: Call Center Management
The Planning and Management Process of Call Centers
How incoming call centers behave
Service Level
Service Level and Quality
Forecasting
Scheduling
Immutable Laws of Call Center Management
Real time Management
Course Outline
Course primarily focuses on Inbound call centers.

4. Call Center management
What is Call Center Management?
Is defined as the art of having the right
number of skilled people and supporting resources
in place at the right times to handle an accurately
forecasted workload, at service level and with quality.
Call Center management
Have the right number of people in the right place at the right time
Do the right things
Changing of names: Call center vs Contact center
What are the types of callcenters? Inbound, outbound, blended
Could also be classified according to FUNCTION: customer care, telesales, tech support, dispatch, operator services, collections, etc
Ownership: Inhouse, outsourced, BOT
Financial Structure: Profit center, cost center

5. QUIZ
Quiz Questions:
Call arrival can be predicted exactly within the half hour for every day with 12 months of historical data (false)
Telephone customers behave just the same as customers in a bank branch or supermarket checkout (false)
80%:20secs is not the best choice for a service level (true)
It is always cheaper for the client to have an 80:60 SL vs and 80:20 SL (False)
Focusing on quality contradicts achievement of SL (False)
One person in a large high volume callcenter will make a significant difference to service level (true)
Erlang c tables provide you with the right number of people to employ (false)
Only Data directly from CMS(ACD system) should be used for forecasting (False)
The more all agents stay in “available”, the less calls they will get (true).
Larger call skill groups are inherently more efficient than small groups (true)
High occupancy is an indication that service level will also be high (false)
Agents control occupancy (false)
AHT usually remains the same throughout the day for an established center (false)
It takes as many staff to increase service level from 75% to 80% as it does from 90%-95% (false)
IVR reduces AHT (false)
Queue messages will reduce abandonment (False)
The call workload will be at its highest on the day of the week with the highest call volume (false)
If agent productivity is low, this means that occupancy will also be low (false)

6. Call Center management
What makes call centers unique ?
Externally Generated Work
Random Call Arrival
Invisible World
Service Benchmarking
Main Customer Touchpoint
Call Center management
No actual control over the amount of work to be done in an hour or day. A call center is at the mercy of workload generated by others.

7. What makes call centers unique?
Random Call Arrival
Call Arrival Exercise
What makes call centers unique?
Prepare balls and bucket
Linear
Is linear call arrival common in our callcenter? No
Did any call have to wait in queue ? No, all calls were answered immediately
What did you notice about the agent’s occupancy? How hard were they working? No avail time so almost - 100%
2.Random
Is this how calls arrive at Telus? Yes
Did any call have to wait in queue? Yes, some.
What did you notice about the agents occupancy? How hard were they working? Vary according to The balls rolled.
Quiz question: Can agents control occupancy? False. It is determined by factors outside their own sphere of control
Occupancy defined as = time on calls/ (time on avail + time on calls) x 100
Affected by: staffing; volume; aht variations
Ideal levels: 80-85%
agent productivity = Time on avail + time on calls/staffed time x 100 (or Staffed-AUX/Staffed)
Affected by over-breaks; other auxes (coaching; meeting etc)

8. GRAPH What makes call centers unique?
The Effect of Random Call Arrival: Calls Queue!
GRAPH
What makes call centers unique?
Peaks and valleys throughout the 30 minute interval
Pattern a: Monday
Pattern b: Tuesday less calls – same general pattern
Quiz Question: Call arrival can be predicted exactly within the half hour for every day with 12months of historical data
Whilst it is possible to have fairly accurate forecasts with historical data, it is almost impossible to consistently be exact
due to the random arrival of calls.
This causes the ratio of staff hours to workload hours to always be higher. There is often an imbalance of work and staff
to handle that work. – Critical things: forecasting, scheduling, real time management

9. What makes call centers unique?
Random Call Arrival
Occupancy VS Agent Productivity
Occupancy = time on calls
x 100
time on calls + time on avail
Agent Productivity = time on calls + time on avail
x 100
staffed time
or staffed time – AUXes
x 100
What makes call centers unique?
Quiz question: Can agents control occupancy? False. It is determined by factors outside their own sphere of control
Occupancy defined as = time on calls/ (time on avail + time on calls) x 100
Affected by: staffing; volume; aht variations
Ideal levels: 80-85%
agent productivity = Time on avail + time on calls/staffed time x 100 (or Staffed-AUX/Staffed)
Affected by over-breaks; other auxes (coaching; meeting etc)
compute AP:
Coaching: 30 mins/week
Meeting: 45mins/week
breaks: 30mins/day
huddles: 15mins/day
Others: 12mins/day
If occupancy was at 85% and AP was at 85%, how many hours did the agent spend time on the phones talking to a customer?

10. Call Center management
What makes call centers unique ?
Externally Generated Work
Random Call Arrival
Invisible World
Call Center management
You cannot compare how good a callcenter is being run just by looking at the call center itself.

11. - Bank, Ticket Office, etc.
Invisible World
Visible Queue
- Bank, Ticket Office, etc.
What makes call centers unique?
Invisible Queue
- Majority of Call Centers (not all)
Note: just click on slide once first so as not to show faces
Use flipchart – Quiz question: Do customers waiting in a telephone queue behave the same
manner as customers waiting in a bank or supermarket checkout? No
Brainstorming:
Tel queue cannot see how long the queue is the progress they are making it it
Tel customer may hang up just before their call is answered because they don’t know they have
Reached the top of the queue while it would very rare for a person in a visible queue to
Walk away just before being served
3. A person going to the bank may not even join the queue if they see it too long
4. A person in the visible queue becomes happier as they progress in the queue.
A person in an invisible queue expects to be assisted soon and as they wait become
more frustrated.
IVR technology – Knowledge of IVR (?)
Can advise caller where they are in the queue/ estimated wait time. Is this good or bad?
Advantage:
Caller is advised up front how many people are in the queue or how long they are to wait
and can therefore make an informed decision about whether to wait or abandon
Disadvantage:
The customer is not told how quickly the queue is progressing. They may be 100th in
Queue which may free up in a short time due to agents logging in
2. The customer may expect an estimated wait time as told but the next message states
a longer wait time (how is this possible?) Based on historical data and is not always reflective
Of what is happening at present time
Some call centers have the technology to make
the invisible queue visible to the caller

12. What makes call centers unique?
Invisible World: 7 Factors of Caller Tolerance
Degree of Motivation
Availability of Substitutes
Competition’s Service Level
Level of Expectations
Time Available
Who’s Paying for the Call
Human Behavior
What makes call centers unique?
Question: Why will some customer’s hang up after only a few seconds in a queue while
others will wait forever?
1. Is there always a correlation between SL and abandonment? Not always the case.
Special promo (Cebu pacific low rates promo) – higher call volume than normal, low
service levels but callers are prepared to wait longer.
2. Online or go visit branch etc
3. Pizza Hut may answer sooner (Domino’s 30 minute or free)
4. Platinum or Gold card holder for Citibank
5. Lunch break, morning, work break vs after office hours and weekends; Mom’s vs singles
6. Toll free or with payment
7……just cant explain
15 minute break!!!!

13. Call Center management
What makes call centers unique ?
Externally Generated Work
Random Call Arrival
Invisible World
Service Benchmarking
Main Customer Touchpoint
Call Center management
4. Service Benchmarking: the call center’s main performance measure is customer satisfaction. So they have to benchmark their performance
against the best of class call centers and not necessarily just those within the same type of industry in the account being serviced
5. Important role it plays in any organization. Callcenter is the most common way for customers to interact with business and for some it is the
only way. It can cause lost or retained customers. Customers make the assessment of the company overall by how they are served by the
Callcenter.

14. The Planning and Management Process of Call Centers1.
Choose a Service Level Objective
The Planning and Management
Process of Call Centers 4.
Calculate Base Staff 9.
Repeat for a Higher and Lower Level of Service 2.
Collect Data 5.
Calculate Trunk and related system resources 7.
Organize Schedules (Schedule Inflex) 3.
Forecast Call Load 6.
Calculate for Shrinkage
The Planning and Management Process of Call Centers 8.
Calculate Costs
Look at the cycle of planning and management process of an inbound callcenter…
There are a number of factors that need to be considered when Convergys and our clients
are negotiating a contract, price being one of the most important. To determine the
Price, we first need to calculate the cost for Convergys to operate the programme.
1st – Choose a service level objective
2. Collect data
3. Forecast call load
4. Calculate base staff
5. Calculate Trunk load
6. Calculate budge factor (shrinkage)
7. Organise schedules
8. Calculate costs
9…*** Repeat the entire process for Higher or Lower Service levels
Let us look at each process in turn…

15. SERVICE LEVEL – THE CORE VALUE
What is Service Level ?
What is the difference between SL and Response Time
What is Service Level? (anyone?)
** the percentage of calls answered within an agreed time frame
(Ex. 80% in 20 secs)
SL applies to work that needs to be answered immediately
Response Time is applied to work that does not need to be answered
as it arrives ( , case work) Ex. 100% of in 24 hours

16. SERVICE LEVEL – 2 Formulas
1. SL = NCH in X secs + NCA in X secs
X 100
NCO
2. SL = NCH in X secs
X 100
There are only 2 things that could happen to a call – it is either answered or abandoned
Compute for SL (80:20)
Calls answered within 20 secs = 6
Calls answered after 20 secs = 2
Calls abandoned within 20 secs = 1
Calls abandoned after 20 secs =1
Convergys computation: Formula 2
6+1/10= 70%
6/10 = 60%

17. Considerations for the Optimum SL:
The Impact on Abandonment (Caller Tolerance)
The Impact on Agent Burnout
The Impact on Costs
The Value of the Call
5. The Desire to Differentiate
SERVICE LEVEL
Generally, the lower the service level, the higher the abandonment rate (although not
always the case).
2. We will learn later (tom), that at a given call load, a higher service level results
in a lower occupancy. If service level is low, the agents will usually be working at
high occupancy rates and therefore taking call after call.
3. Costs:
Labor costs: More agents are required to achieve a higher SL than a lower SL.
Therefore, labor costs would be more expensive for a high SL programme
Trunk costs: Phone costs (VOIP – 70 calls per E1 line). Trunking costs are
incurred when a customer is in queue or talking to an agent. The longer a
customer waits in queue, the more expensive the phone costs wil be.
4. Value of the Call – Generally a sales call will be more valuable to a client
than a technical support call as it represents the potential revenue to the customer.
Emergency calls have a higher SL.
5. Looks at the factors of caller tolerance that their business is likely to support
to determine the likelihood of customers waiting on line
6. A company may want to blow their competition out of the water by offering a
Very high SL compared to competitors.

18. Service Level And Quality SERVICE LEVEL and QUALITY
Quality is the consistent delivery of service that meets or exceeds customer’s expectations.
Does Quality and the achievement of Service Level contradict one another? Do we really sacrifice quality for AHT?
SERVICE LEVEL and QUALITY
Does Quality and the achievement of SL contradict one another? Do we
really sacrifice quality for AHT?
No, in fact they should work hand in hand. If a call is handled with quality,
the proper call handling techniques should have been utilized thereby allowing
the call to be completed within the required time frame.

19. When Quality is Lacking: SERVICE LEVEL and QUALITY
Escalation of calls and complaints to higher management
Repeated calls from customers
Callbacks to customers for missing or unclear information
Unnecessary service calls
Diversion of agents to activities that should be unnecessary
Agents taking the blame for errors made by others
SERVICE LEVEL and QUALITY
Effects to service level:
Increase handle time – escalations/irate
Repeated calls from customers – volume increases
Callback to customers
Unsupported or when irate already, you have to do more to appease
Lack of knowledge, long holds
5. Start the call irate, AHT lengthened to appease client
In program AHT reduction (benchmarking or timestudies) – There is only
so much an agent can do to reduce overall aht. You have check the process
Or steps a CSR needs to accomplish to carry out each request… call speeds
plateau at a certain point. The process is where there is most leverage (i.e
Scripts, software, tools).
Quality is built around customers’ expectations.
Quality and Service Level work together by reducing rework
and waste and facilitating and positive culture.
The process is where the leverage is.
Use Quality tools to find and fix root causes.
Our operators are the most important source of our success.

20. The Planning and Management Process of Call Centers1.
Choose a Service Level Objective
The Planning and Management
Process of Call Centers 4.
Calculate Base Staff 9.
Repeat for a Higher and Lower Level of Service 2.
Collect Data 5.
Calculate Trunk and related system resources 7.
Organize Schedules (Schedule Inflex) 3.
Forecast Call Load 6.
Calculate for Shrinkage
The Planning and Management Process of Call Centers 8.
Calculate Costs
COLLECTING DATA AND FORECASTING CALL LOAD

21. Data Gathering Sources
Forecasting
Data Gathering Sources
FORECASTING
Historical ACD data (including call volume, handle times, arrival patterns)
Business Drivers
- Internal:
* Marketing activities
* Mergers and Acquisitions
* Network capacity/deployment
* Product launch/ product defects
- External:
* Weather disasters/ flooding/fire
* Market fluctuations
* World and local news events
* Industry rumors
ACD data needs to be checked on for volume, handle time and arrival patterns.
Volume: Need to see if trunks are enough to handle all traffic. Not enough trunks, inaccurate forecasting (in case trunks are increased or improved)
Data Aberrations

22. FORECASTING Data Gathering: CALL WORK LOAD
Definition of Call Work Load
= Call volume x AHT VS
Call volume only
Difference with just looking at call volume – it looks at the total amount
of work that needs to be completed rather than how many call need to
be answered (i.e 40 hours of work)
Importance of getting it right: “Everything to do with resourcing the call center
starts with determining what the calling load will be. In the ultimate sense, the
call load determines how many agents are hired. The number of agents on the
busiest shift determines how much space is required; how many lines are necessary,
how many chairs and desks are puchased;etc. If you get this part wrong, no matter
how good your management and no matter how efficient your agents are, you will
be lost in terms of meeting your SL objective.”
It is important to use Call Load rather than call volume when calculating how many
staff are required for a programme.

23. Call Volume v’s Call Load
Call Volume v’s Call Work Load
Call Volume v’s Call Load
Would you need more staff working on Monday or
Tuesday based on the example above?
AHT =
200 seconds
300 seconds
CALL LOAD
Compute using definition for Call load..
1000 X 200 = 200,000 seconds (or man hours)
700 x 30 = 210,000 seconds (or man hours) – Tuesday
Because AHT is higher on a Tuesday than a Monday

24. ACD Data: Check for volume, AHT, patterns
Volume: Trunk capacity
Data Aberrations
Handle Time: Time of day
Day of week
Events
Arrival Patterns: Holiday factors
Monday effect
FORECASTING
FORECASTING
ACD data needs to be checked on for volume, handle time and arrival patterns.
Volume: Need to see if trunks are enough to handle all traffic. Not enough trunks, inaccurate forecasting (in case trunks are increased or improved)
Data Aberrations

25. Interval Call Arrival Patterns
Why does the AHT or call volume vary according to the time of day?
Note: Sudden dip on Tuesday, 10:30 am. Why?
For arrival patterns and for forecasting, it is important to single out the unusual
for better analysis.

26. Day of Week Factors Why does aht vary from day to day?
Note: Monday – Wednesday of the week of May 6 shows an unusual pattern: Why do you think so?

27. Average Handle Time (AHT):
AHT in seconds
Why does the AHT vary according to the time of week?
Time of Day

28. Monthly Volumes
Why does the call volume or AHT according to the month of the year?

29. FORECASTING How to Break Down a Forecast:
720,000 Current year’s calls (1)
X To add 12% (2)
806,400 Estimated calls in forecast year (3)
X January proportion (4)
57,254 January calls (5)
/ Operations days in January
1,847 Average calls per day (6)
X Monday Index Factor (7)
2,713 Mondays calls (8)
X :00 to 10:30 proportion (9)
149 Forecasted calls 10:00 to 10:30
To determine operations days, count the days the call center will be open
Example – January
S M T W T F S
FORECASTING
To calculate the day of the week index factor, divide day of week proportion by average day of week proportion.
Avg Index
Example: Prop Prop Factor
Monday /
Tuesday /
Wednesday /
Thursday /
Friday /
Saturday /
Sunday /
Based on historical data from January to December
7% for january…..
Based on what??? Increase in customers (growth rate); marketing activities; environment etc
Estimated number of calls to be received for following year
Monthly proportion – based on historical data (see slide 26) – several years can be averaged
Get the % of total year’s calls received in a particular month
5. January’s forecasted calls
6. Get average calls per day by diving the number with Operational days in January (some
programs are not open 7 days a week;take out also holidays that the center is closed)
7. To calculate the day of week Index factor, divide the day of week proportion by the
average day of week proportion ---- Or divide one day over number of operating days in the wk
8. Mondays calls ( see that it is higher than the average calls per day)
Other index factors used are: holiday factors; promotional factors; etc
9. Proportion for each half hour interval
Quiz Question: Call arrival can be predicted exactly within the half hour for every day with 12months of historical data
Whilst it is possible to have fairly accurate forecasts with historical data, it is almost impossible to consistently be exact
due to the random arrival of calls.

30. FORECASTING Forecasting Meetings Assign responsibilities
Assemble the right players
Assign responsibilities
Establish recurring meetings
Assess progress
Adjust the process
FORECASTING
** To ensure that we are as fully informed as possible, the scheduling and forecasting team
of TELUS facilitates a monthly forecasting meeting with each of our clients.
Ten Common Forecasting Problems (that should be avoided):
No systematic process
Events that should be exceptions become part of the forecast
Good ties with other departments or clients do not exist
Planning is done around goals and not reality
Bad judgement of temporary vs permanent change
No one is accountable for the forecast
END OF FIRST DAY!!!!

31. The Planning and Management Process of Call Centers1.
Choose a Service Level Objective
The Planning and Management
Process of Call Centers 4.
Calculate Base Staff 9.
Repeat for a Higher and Lower Level of Service 2.
Collect Data 5.
Calculate Trunk and related system resources 7.
Organize Schedules (Schedule Inflex) 3.
Forecast Call Load 6.
Calculate for Shrinkage
The Planning and Management Process of Call Centers 8.
Calculate Costs

32. Staff Required Calculate Base Staff
Base Staff – the number of agents required to
be on the phones for a given period to handle a particular
call load.
Staff Required
How is it done? Erlang C
Erlang C is a formula developed to take into consideration the random call arrival. It calculates
the minimum number of agents required to achieve service level for a given period of time. It
assumes that all agents will be “in their seats” for the entire period.
It assumes that no caller abandons or receives a busy signal, so it will sometimes have the
tendency to overestimate the number of staff requirede.
How many staff able to take calls in a given interval
and NOT how many people to employ

33. Erlang C Erlang C Module
ERLANG C FOR INCOMING CALL CENTERS BY ICMI, INC.
TALK TIME IN SECONDS = 180
AFTER CALL WORK IN SECONDS = 30
CALLS PER HALF HOUR = 250
SERVICE LEVEL OBJECTIVE IN SECONDS = 20
OPERATORS
ASA
SL%
OCC%
TKLD 30
208.7
23.5%
97%
54.0 31
74.7
45.2%
94%
35.4 32
37.6
61.3%
91%
30.2 33
21.3
73.0%
88%
28.0 34
12.7
81.5%
86%
26.8 35
7.8
87.5%
83%
26.1 36
4.9
91.7%
81%
25.7 37
3.1
94.6%
79%
25.4 38
1.9
96.5%
77%
25.3 39
1.2
97.8%
75%
25.2 40
0.7
98.6%
73%
25.1 41
0.5
99.2%
71%
Operators:
- Number of operators required to be on the phones
ASA:
Average speed of Answer in Seconds
SL%:
Service Level
Percentage of calls that need to be answered in the number of seconds you specify.
OCC%:
Percent agent occupancy. The percentage of time agents will spend handling calls,
including talk time and after call work.
TKLD (Trunk Load):
The hours (erlangs) of trunk traffic, which is the product of:
(talk time + ASA) X number of calls in an hour.

34. Erlang C Delay Module ERLANG C FOR INCOMING CALL CENTERS BY ICMI, INC.
TALK TIME IN SECONDS = 180
AFTER CALL WORK IN SECONDS = 30
CALLS PER HALF HOUR = 250
SERVICE LEVEL OBJECTIVE IN SECONDS = 20
|==================== Number of callers waiting longer than X seconds ======================>|
OP:
SL% 5 10 15 20 30 40 50 60 90
120
180
240
===== 24
203
119
195
191
184
177
170
163
145
129
101 80 31 45
156
149
143
137
126
115
105 97 74 57 34 32 61
118
111
104 85 65 56 38 25 11 33 73 89 81 67 47 39 19 4 1 82 58 52 46 37 29 23 18 9 35 88 41 36 14 2 92 21 8 6 95 16 3 13

35. The Planning and Management Process of Call Centers 1.
Choose a Service Level Objective
The Planning and Management
Process of Call Centers 4.
Calculate Base Staff 9.
Repeat for a Higher and Lower Level of Service 2.
Collect Data 5.
Calculate Trunk and related system resources 7.
Organize Schedules (Schedule Inflex) 3.
Forecast Call Load 6.
Calculate for Shrinkage 1.
Choose a Service Level Objective 9.
Repeat for a Higher and Lower Level of Service 2.
Calculate
Base Staff
The Planning and Management Process of Call Centers 8.
Calculate Costs 8.
Calculate Costs 3.
Forecast
Call Load 4.
Calculate
Base Staff 5.
Calculate Trunk
(and Related
System
Resources)

36. Shrinkage Calculating Shrinkage VACATION LEAVES 15 days 5.77%
SICK LEAVES
6 days
2.30%
BREAKS
Two 15-minute breaks per day
6.25%
COACHING
0.5 hour per week
1.25%
TRAININGS/MEETING
1 hour per week
2.50%
NON-SCHEDULED BREAKS
3 minutes per available hour
0.08%
TOTAL
18.15%
How much total shrinkage if the breaks are 45mins per day, non-scheduled breaks for 10 mins/day.
Will shrinkage vary from day to day, time of day?
So how will call center schedulers cope with this?
Shrinkage VS Agent Productivity
Assumptions: 52 weeks a year; 260 days; 2080 hours; 124,800 minutes. Based on 8-hour days.

37. Shrinkage Calculating Schedule Requirement
Schedule Requirement = Base staff Requirement
1 – Shrinkage Factor
Common math mistake – use shrinkage factor as a multiplier factor to inflate the bodies in chairs number

38. Schedule Inflex Organize Schedules
Schedules are fundamentally inefficient.
The more flexible your workforce is,
the more efficient your schedule can be.
Factors in managing schedules.
Shrinkage
Coverage
Length of shifts
Days off
Break Definitions
Start and stop times
Other work rules
Schedule Inflex
Full coverage approach – every single half-hour of the day is covered (up to peaks).. Most callcenters
do not use the full coverage approach.
Balanced approach – key is to keep the amount of the understaffing and overstaffing minimized
each half hour. Over/ Under Scheduling

39. Schedule Inflex Match Shifts with the Demand Curve –
Balancing Sufficiency and Efficiency
Schedule Inflex
Expanded schedules – CWW
Full time and part-time mix

40. Schedule Inflex Envelope Strategy
Outbound; Quality Improvement; ; Special projects

41. Immutable Laws of Call Center Mgt
When Service Level goes up,
Occupancy goes down (at a given call load)
2. With more staff, ASA goes down
With more staff, trunkload goes down
Law of diminishing returns
Powerful Pooling Principle
Larger answering groups have higher occupancy
at a given Service Level
Immutable Laws

42. Labor v’s Trunk Costs ERLANG C FOR INCOMING CALL CENTERS BY ICMI, INC.
TALK TIME IN SECONDS = 180
AFTER CALL WORK IN SECONDS = 30
CALLS PER HALF HOUR = 250
SERVICE LEVEL OBJECTIVE IN SECONDS = 20
A: Cost per hour for OP: $15.00
B: Cost per hour “800” service $ 6.00

43. Immutable Laws The Dynamics of Size Calls per hour Handle Time
Call Load (in hours)
Number of Agents
Service Level in 20 secs
Agent Occupancy
200
180secs
10 hours 11
39%
90% 12
64%
83% 13
80%
77%
800
40 hours 45
89% 46
86%
87% 47
85%
1600
80 hours 87
92% 88
88%
91% 89
Immutable Laws
Economies of scale Powerful Pooling Principle:
Small group requires 13 agents to meet the service goal.
Medium sized group receives four times the amount of work (200 x4) YET 4 x 13 =52 is not needed.
The service goal can be met with 45 staff instead of 52 staff.
Large group receives 2x calls of medium (8 x calls of small) YET 2 x 45 = 90 (only 87 is needed)
OR 13 x 8 = 104 (104-87=17 less staff) - This phenomenon is referred to as the Powerful pooling principle.
Related to 6th Immutable Law (Large answering groups have higher occupancy at a given
Service level)
Why? Larger groups simply have opportunity to handle more calls during the period of time they’re
working. With a larger volume of calls, there is greater likelihood that when an agent is finished
with one call, there is another one arriving for him to handle. There is less “down time” waiting
for a call to come in and the larger volume of calls represents a smoother calling pattern.
ASA also becomes less for large groups because small groups have larger net effect of waiting
and fast ones… more erratic visually.
Consolidation: multi-skill groups (consider value of cross training); or (intelligent) virtual call routing arrangement

44. Immutable Laws The Dynamics of Size Service Level % 90% Answered in:
Answered in:
30 secs
Staffing Req
Volume
AHT (secs)
ASA (secs)
Occupancy
119
575
340
7.8
91% 65
300
8.2
87% 45
200
8.6
84% 18 70
11.1
73% 8 25
13.4
59%
Immutable Laws
ASA also becomes less for large groups because small groups have larger net effect of waiting
and fast ones… more erratic visually.
As the volume offered and staffing required for each group gets smaller, the resulting average speed of answer (ASA) gets larger even though the service level remains constant. In each group, only 10 percent of callers wait longer than 30 seconds for their call to be answered. The ASA variance between the groups is the result of the amount of time the calls not answered within 30 seconds actually have to wait for an agent to become available. This variance in wait time is generated by the law of large numbers — larger answer groups will have agents available more frequently, thus reducing the hold time when the 30-second objective has been surpassed. Put another way, when comparing groups with the same service level, calls that don’t get answered within the service level objective will generally wait longer in smaller agent groups than in larger ones.

45. Real Time Real Time Management
is about ensuring that all agents are doing “what they
are supposed to be doing” in order to:
- give us the best chance to meet the service level
objectives (Power of One principle)
- maintain phone occupancy at acceptable levels to
prevent agent burn-out
- lost hours on a $ perspective
REAL TIME MANAGEMENT : is about ensuring that all agents are doing “what they are supposed
to be doing” thereby giving us the best chance of meeting the service level objectives
Exercise: Table or Static Chair Lift (Emphasize the benefits of sharing the load amongst the
whole team). One by one, lifters become less. Less to share the load with.
Quiz Question: The more all agents stay on avail, the less calls they will get. Domino
effect of remaining on ACW or AUX.
Power of One principle: looking at the Erlang C table, we can see that one person in
a large high volume call center will make a significant difference to service level (quiz question)
Billed to Paid = Billed hours (to client)
Paid hours (to agent)
Type of programs: hourly billable (includes or exclude auxes? – if exclude, time on avail is paid)
per minute billable (AHT only – opportunity lost even if meeting SL because
at some intervals, you can make up for lost calls on low SL intervals – sched inflex

46. - QUIZ -
Quiz
Larger call skills groups are inherently more efficient than small groups – T
High occupancy is an indication that SL will also be high – F
Agents control occupancy – F
AHT usually remains the same throughout the day for an established desk – F
It takes as many staff to increase SL from 75% - 80% as it does from 90-95% - F
IVR reduces AHT – F
Queue msgs will reduce abandonment – F
The call load will be at its highest on the day of the week with the highest call volume – F
Call arrival can be predicted exactly within the half hour for everyday with 12 months
of historical data – F
10. Telephone customers behanve just the same as customers in a bank branch or supermarket
Checkout – F
11. 80:20 is the best choice for a SL - F
12. It is always cheaper for the client to have an 80:60 SL vs an 80:20 SL – F
13. Quality call handling contradicts the achievement of SL – F
14. Individual agent effort makes the biggest difference to AHT improvement – F
15. One person in a large high volume call center will make a significant difference
to SL – T
16. Erlang C tables provide you with the right number of people to employ