classification of sales opportunities for software company · introduction sales operations portal...

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

Classification of Sales Opportunities for Software

Company

Jeanne M. RussoDaniel Rozas

Pablo MacouzetSriram Gandhi

Vino Babu

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

Introduction

Sales Operations Portal introduced in 2004 to replace SalesForce.com system.Design as a single source for remote access to all sales related activities: Account management, calendar, tasks, opportunitiesUsed by all field employees : Account Executives, Sales Consultants, Sales Managers & DirectorsData related to Sales Opportunities will be the subject of our analysis

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

Introduction- Sales Opportunities

• Life cycle of a Sales OpportunityCustomers

- Partners

- Homepage

- Events

- Industry Seminars

- Customer Referral

- Cold Calls

Lead

Opportunity

Order

Lost DealSales Process

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

Objectives

1. Taking advantage of data gathered by the system to predict the likelihood of open opportunities becoming won or lost deals

2. Improve the process to assign resources to opportunities based on the sales potential and skill of the sales force

3. Better understand the most influential factors that lead to a win

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

Sales Opportunity- Parameters

Opportunity TypeLead SourceRevenue StreamRegionOpportunity Creation DateOpportunity Close DateAge (days since creation)AmountMilestoneCompetitors

New CustomerCustomer - New BusinessCustomer - ExpansionPartners

ReferralsEvents ….

- Lost- Order

22 Regions

LicenseConsultingSupport Maintenance

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

Dataset

Raw data from ROLAP database exported to Excel2080 recordsOpportunity Milestone is response variableNaive rule: 86% of opportunities result in orders

Cleanup & TransformationsConsolidate Regions (from 22 to 8)Consolidate Opportunity streams and create dummy

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

DSI Expansion New Business (existingcustomer)

New Customer

Order Rate

Mean Rate

Single Dummy

Opportunity Type

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

0%

20%

40%

60%

80%

100%

<$100K $100K - $250K $250K>

Order RateMean Order Rate

Opportunity Amount

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1 2 3 4 5 6 7 8 9 10

Age Bin

New CustomerDSI ExpansionAge Distribution: New CustomerAge Distribution: DSI Expansion

Bin To Count1 6 2162 14 1923 27 2124 38 1785 56 2086 82 1897 114 2008 169 1979 269 199

10 1329 199270

FromOpportunity Age (Days)

5783

115170

7152839

0Sales Divergence

Uniform Distribution

Non-uniform Distribution

Opportunity Age

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

0%

5%

10%

15%

20%

25%

1 2 3 4 5 20 21 22

Consolidated Region

Single Dummy Variable

Loss Rate by Region

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

0

50

100

150

200

250

1 2 3 4 5 6 7 8 9 10 11 12

Month Created

Volu

me

0%

5%

10%

15%

20%

25%

Loss

Rat

e

Volume

Loss Rate

Mean Loss Rate

Dummy Variable

Creation Period

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

0%

20%

40%

60%

80%

100%

120%

1 2 3 4 5 6 7 8 9 10 11 12

Close Month

Ord

er R

ate

0

100

200

300

400

500

600

700

Volu

me

Order rate

Mean Rate

Volume

Closing Period

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

Regression ModelsLogistic Regression

ComplexityHit R

atio

New Customer

Business Cycle

More Involvement

Product Placement

Validation Set

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

Applications

Opportunity probability can be used in conjunction with other data to determine a master plan for sales efforts

© 2005 Robert H. Smith School of BusinessUniversity of Maryland

RecommendationsIncorporate the predictive model as part of the Sales Force reports already available.Automate the process so that the model improves as more data is gathered.Quantify costs associated with false positives and negatives to be incorporated into the model.Add granularity to competitors data (currently missing values)Investigate why particular regions have lower hit ratio and losing high amount dealsInclude opportunity probability as part of the quota and commissions system.

© 2005 Robert H. Smith School of BusinessUniversity of Maryland