Salesforce Issues

1
Salesforce Issues

• Goals
• Sizing
• Allocation/Alignment
• Organization

2
Salesforce Issues

• Goals
• Sizing
• How many?What type? (Syntex)
• Allocation
• Over products (Syntex/Detailer)
• Customers/prospects (Syntex)
• Territory design (Geoline)
• Call planning (CALLPLAN)(Integrated models ZS)
• Organization/Control
• Compensation

3
Salesforce Management Decisions
4
Salesforce Size

• of sales method (back into number of people)
• Breakdown method (add up sales potential)
• Based on response model
• Syntex

5
Sizing ExampleSalesforce Sizing at Syntex

• Current salesforce size (1982) was 433 reps,
  expanding by 3040 per year.
• Salesforce allocated to
• 7 Drugs Naprosyn, Anaprox, Norinyl, etc.
• 9 Physician Specialities Family practice,
  internal medicine
• OB/GYN, etc.
• by judgmental/historical norms.
• Questions
• How large should the salesforce be?
• How should it be allocated to drugs and to
  physician specialities?
• Approach
• Judgmental estimation of response functions.

6
Syntex Model
Each management team member separately
estimatesa response function for each sales
entry As a group, the management team members
discuss and develop consensus estimates of
response functions Run the model Test profit
consequences of alternate scenarios Do results
make sense? Can they be implemented? Implement
resource allocation decisions Monitor and
evaluate sales performance
No
Yes
7
Opportunity Costs Identified by Syntex Model
15M
Opportunity cost of maintaining current sales
force size
10M
Change in Profit after Change in Salesforce Size
and Reallocation of Effort
5M
0
5M
0 Current level
30 Optimal increase in sales force size
Change in Salesforce Size from Current Level
8
Allocation

• Allocation to existing customers and prospects
  (CALLPLAN)
• Allocation to products
• Rules of thumb.
• Response models.

9
How to Estimate theResponse Function?

• Historical data
• Available?
• Sufficient variability for estimation?
• Experiments
• ABB (Remember?).
• Difficult to execute.
• Managerial Judgment
• Quick.
• Builds consensus.
• Encourages systematic thinking.

10
Allocation Example CALLPLAN Model at United
Airlines

• Step 1
• Determine expected contribution of all possible
  call policies for each product account.
• Step 2
• Allocate time to accounts.

11
Each Salesperson is Asked . . .

• What would sales of this account be if . . .
• . . . you dont visit the account? X0
• . . . you visit the account the same as last
  period? X1
• . . . you visit the account 50 more than last
  period? X2
• . . . you visit the account as frequently as
  possible? X3

12
This Yields
Expected Sales
0
Base
1.5 Base
Number Of Calls
13
Problem

• How much time to spend at each account to
• Maximize
• sum of account revenues account margin
• Subject to
• salesperson time constraint
• min./max. calling restriction
• Output
• Recommended calling frequencies for each account

14
Optimization Procedure in CALLPLAN

• Number of calls
• 1 2 3 4 5 6 7 8 9 10
• 1 100 300 600 400 300 225 160 11 70 50
• Account 2 1400 1100 850 650 490 360 250 165 85 15
• Number 3 3600 1800 800 200 100 75 50 40 30 25
• 4 180 170 160 150 140 135 130 125 120 115
• The numbers in each row of the matrix correspond
  to the response function for that account,
  showing incremental contribution per call at that
  call level.
• Indicates marginal contributions at current
  level of deployment
• Indicates marginal contributions at the optimal
  level of deployment

15
Current Allocation vsModel Recommended Allocation
8
8
7
7
6
6
5
5
Number of Calls
Number of Calls
4
4
3
3
2
2
1
1
0
0
1
2
3
4
1
2
3
4
Account Number Current Allocation
Account Number Model Recommended Allocation
16
Does CALLPLAN work?

• The United Airlines Experiment
• 20 sales people participated
• 10 West Coast
• 10 East Coast
• 10 pairs selected
• 5 East and 5 West matched by managerial judgment,
  personal characteristics, compatibility of
  territory size, and revenue account mix.
• 10 CALLPLAN participants selected
• 1 from each pair.
• Other 10 were control group.
• Control group
• Told they were in an experiment.
• Participated in judgmental call frequency/account
  response development.
• Received computer feedback of their input but no
  CALLPLAN recommendations.

17
Results after 6 Months

• Sales Change vs Previous Year
• Experimental Group Control Difference
• 11.9 3.8 8.1
• (Probability that this result is due to chance
  lt 0.025.)

18
Profit Impact

• Time Needed
• 1/2 day per salesperson
• Change in Sales
• ( 95 profit in airline business) gt 1,000,000
  (for the 10 sales people) in 6 months.

19
Territory Design ExampleGeoline

• ObjectiveTo design territories that are
• 1. Of equal workload.
• 2. Contiguous.
• 3. Compact.
• ModelMaximize compactness
• S.T. Workloads are equal.
• Contiguous.
• Sales people are assigned to one or more
  geographical units.
• (Uses legislative redistricting software.)

Workload is measured using variables such as
customer count, number of calls, sales potential,
or some weighted combination of these variables.
20
Sales Territory Design Output

• Optimal boundaries and resident locations for
  three sales territories.

10
8
6
Miles
4
2
0
2
4
6
8
10
12
14
16
18
20
0
Miles
21
Sales Force Alignment/Allocation Influences

• Environmental Changes
• Competitor activity
• Shifts in buyer behavior
• Demographic shifts
• Government regulations
• Organizational Changes
• New products
• Maturing product lines
• Restructuring
• Mergers and acquisitions
• New data sources

22
What Can Realignment Accomplish?
Workload Balance
23
Territory Design Objectives and Criteria
Sales Workload Potential Minimum Territory
Balance Balance Disruption Compactness Customers R
esponsiveness Relationships Reps Earnin
gs opportunity Reduced uncertainty Contro
l of overnights Firm Sales
results ? Effort control Travel cost
control

• Conflicts?
• Tradeoffs?
• Priorities? (How to set?)

24
Compensation (Indirect Control)

• 1. Level
• 2. Form
• Salary
• Commission (on sales? on profit? short/long
  term?)
• 3. Other
• Contests
• Bonuses