Products vs.

Services:
Which is the Better Business Model,
in Software and Other Industries?
Michael A. Cusumano
MIT Sloan School of Management
cusumano@mit.edu
2006

The Global Software Industry

180
160
140

Value in $B

x10

120
3rd Party Services

100

Products
80

Software Servcies

60
40
20
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Year
4

State of the Business

Overall Recovery (esp. in 2005)
But Collapse of Traditional Products
Rapid Growth of Services/Maintenance
as % of Revenues for Product Firms
Future: Intense Battle between Products
Firms & IT Services Firms?
5

The Big Questions

Rise in services & maintenance revenues
temporary or permanent?
Will most software firms become services firms?
Temporary Argument: We are in a transition phase
between platform innovations (client-server to internet to
web services & wireless)
Permanent Argument: Software, like hardware, has
become commoditized and prices will fall close to zero for
standardized products. Future is software as a service.
6

Role of Services in Life-Cycle Dynamics

& Platform Transitions?
Maturity
Performance

..Services.? Disruption
Takeoff

Ferment
Time
7

Typical View of Services in HighTech Companies?

Services will be the graveyard for old tech
companies that can't compete."
Scott McNealy
CEO, Sun Microsystems
Referenced in N.Y. Times, 9-16-04

The Problem with Many Services

Much more labor intensive than products
Hard to scale without adding people
SAP example (1:1)

Costs not as easy to control compared to

standardized product dev & production
Hard to attract VC funding or do an IPO
Low-cost competition from India and
elsewhere pushing margins down further
9

Business Objects: Gross Margins

12 0
10 0
80
Pr od u c ts

60

S er v ic es

40
20

20
02

20
01

20
00

19
99

19
98

19
97

19
96

19
95

19
94

19
93

10

The Problem with Many Products

Hard to write best-sellers (killer apps)?
Products can become commodities?
Example: Price for same (actually better) software
product $1.5 million in 2000 but $250,000 in 2004

Products more subject to discretionary spending.

In bad economic times, product sales more likely to
fall off a cliff ? E.g. Siebel, i2, Oracle

Only guaranteed revenues services/maintenance?

These under downward pressure as well

For products business: 99% of 0 = 0

11

Siebel

Siebel

1200
350
300

800

250

600

Products

200

Services

150

400

Products
Services
Total

100
50

200

1999

1995 1996 1997 1998 1999 2000 2001 2002

2000

2001

2002

Siebel

%of Total Revenues

$ million

1000

100
90
80
70
60
50
40
30
20
10
0

Products
Services

1995

1996

1997

1998

1999

2000

2001

2002

12

PeopleSoft

PeopleSoft
1600
1400
1200
1000
800
600
400
200
0

200
150

Products

Products

Services

Services

100

Total
50
0

19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02

1999

2000

2001

2002

PeopleSoft

Products

20
02

20
01

20
00

19
99

19
98

19
97

19
96

19
95

19
94

Services

19
93

19
92

80
70
60
50
40
30
20
10
0

13

7000
6000
5000
4000
3000
2000
1000
0

Oracle

Products
Services

1999 = 100

200
150

Products

100

Services
Total

50

19
92
19
94
19
96
19
98
20
00
20
02

0
1999

2000

2001

2002

Oracle
70
%of Total Revenues

$ million

Oracle

60
50
40

Products

30

Services

20
10
0
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

14

SAP

6000
5000
4000
3000
2000
1000
0

200

Products

150

Services

100

Products
Services
Total

02

0
1999

20

00

20

98

19

96

19

19

19

94

50

92

m illio n e u ro s

SAP

2000

2001

2002

SAP

70
60
50
40

Products

30

Services

20
10

20
02

20
01

20
00

19
99

19
98

19
97

19
96

19
95

19
94

19
93

19
92

%of Total Revenues

80

15

IBM

IBM
120
115

30

1999 = 100

$ billion

40
Software

20

Services

10
0

110

Products

105

Services

100

Total

95

19
92
19
94
19
96
19
98
20
00
20
02

90
1999

2000

2001

2002

IBM

70
60
50

Software

40
30

Services

20
10
20
02

20
01

20
00

19
99

19
98

19
97

19
96

19
95

19
94

19
93

0
19
92

%of TotaR
evenues

80

16

Business Objects

Business Objects

300

300

Products

150

Services

100

1999 = 100

250

200

200

Products

150

Services

100

Total

50

50

0
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

1999

2000

2001

2002

Business Objects
90
%of Total Revenues

$ million

250

80
70
60
50
40

Products
Services

30
20
10
0
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

17

Business Objects Products vs. Services

100
80
60

Products

40

Services

20

19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
20
03
20
04
20
05

18

Services & Maintenenance as Percentage

of Total Revenues
Business Objects
i2
Seibel
Peoplesoft
Oracle
IBM S&S
SAP
Compuware

19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02

100
90
80
70
60
50
40
30
20
10
0

19

90
80
70
60
50
40
30
20
10
0

Products

ut
io
So
l

yb
rid

Se
rv
ice

ns

Services

Pr
od
uc
ts

% of Total Revenues

Three Business/Life Cycle Models

Typical Breakdown Per $1.00 of Enterprise

Product Revenue

Year 1

Software Software MainteProduct Services nance

$1.00
$1.00
$0.00

Total
$2.00

Year 2

$0.30

$0.15

$0.45

Year 3

$0.25

$0.15

$0.40

Year 4

$0.15

$0.15

Year 5

$0.15

$0.15

TOTAL

$1.00

$1.55

$0.60

$3.15

30 %

50 %

20 %

100 %

21

Initial Research Questions

Products not as good as previously thought?
99% of 0 = 0, especially in down economies?
Platforms products and some niche products do better?
But new products drive services & maintenance?

Services not as bad as previously thought?

Can double or triple a firms sales and profits?
But maintenance much better than other services?
Services help create stickier product solutions?

Hybrid the best business model?

Most stable performance & strategy re commoditization?
But requires most complex combination of skills?

22

New Database Study

MC with Steve Kahl and Fernando Suarez, and
undergrad students; earlier Vikram Mansharamani
Identified 463 public software products firms under
SIC code 7372 PrePackaged Software (NAICS #51121)
Financial information from Mergent Database & 10K
reports. Avg. 9, maximum 15 years of detailed financial
information, from firms listed in 1995 or later.
3788 total yearly observations (4449 including nobreakout firms).
Now doing exploratory analysis
Also starting database of non-software firms
23

Public Software Product Companies Listed in US

400

(% Annual Product Sales, 378 firms & 3314 yearly observations)

100% Product

200

Hybrid Balanced, 34-85%

100% Service

Hybrid
Service

Hybrid
Product

100

Frequency

300

.2

.4

.6

.8

prod p2

Notes: -- Excludes 86 packaged software firms with no sales breakout and unclear status.
-- 1 (100%) includes some product firms that did not break out revenue mix (MSFT, Adobe,
SPSS, Visio, Symantec, and Fair Isaac, and game software firms).

24

Data Analysis
Broke out hybrid using standard deviation. Distribution
approximated normal. Used 1 standard deviation to
calculate the middle group. The mean is .6 and standard
deviation is .257
Total observations for the 5 groups:
Services:
76
Product:
330
HybridS: 464
HybridB: 2206
HybridP: 302
Total:
3378
25

Software Product Firms

by Business Model
Year

100%
Product

Hybrid
Product

Hybrid
Balance

Hybrid
Service

100%
Service

Total

1995

25

26

162

19

240

1997

31

34

207

34

313

1999

23

17

208

38

10

296

2001

19

10

166

57

255

2003

15

12

113

52

196

2004

13

101

45

170

Percentage Breakout By Year

100%
90%
80%

% of Total

70%

Service

60%

Hybrid Ser

50%

Hybrid Bal

40%

Hybrid Prod
Product

30%
20%
10%
0%
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Year

Average Revenue Breakout by Age

80%

Average Percentage

70%
60%
50%
Products

40%

Services

30%
20%
10%
0%
1

10 11

12 13 14

15 16

17 18 19

20 21 22

23 24 25

26 27

28 29 30

Age

Average Revenue Breakout by Year

80%

Average Percentage

70%
60%
50%
Product

40%

Services

30%
20%
10%
0%
1990

1991

1992

1993

1994

1995

1996

1997
Year

1998

1999

2000

2001

2002

2003

2004

28

Service-Maintenance % by Product Type and Firm Age

70%

Average Service Contribution

60%
50%
Total

40%

Infrastructure
30%

Pre-packaged Apps

20%
10%
0%
1

10

11

12

13

14

15 16

17

18

19

20

21 22

23

24

25

26

Age

Service-Maintenance % by Product Type and Year

50%
40%
Total
Infrastructure

30%

Pre-packaged Apps
20%
10%

Year

04
20

03
20

02
20

01
20

00
20

99
19

98
19

97
19

96
19

95
19

94
19

93
19

92
19

91
19

90

0%

19

Average Service Contribution

60%

29

Revenue Mix & Performance (1)

Shift to services-maintenance driven by (1)
declining product revenues/prices, (2) aging ERP
firms/fewer new customers, (3) platform shifts
(e.g. client-server to Internet)

Hybrid solutions firms generally have (1)

higher and more stable profits and (2) higher
market valuations than software product firms if
we exclude Microsoft.
30

Revenue Mix & Performance (2)

Service-maintenance revenues generate
higher and more stable profits than product
revenues for all software product firms if we
include the costs of R&D against product
revenues

Optimal mix for profitability: 73% product

revenues, but hard to achieve. Very high
maintenance revenues another strategy?
31

Services

Growth index

Products

A: Case of a firm
where products and
services revenues
reinforce each other

Time

Services

Growth index
Time

Products

B: Case of a firm
where services
as % of revenues
rise because
products business
is falling
32

Mean Profit Contribution by Revenue Type & Year

60%

Mean Percent Contribution

50%
40%
30%

Product Profit Mean

Service Profit Mean

20%
10%

04
20

03
20

02
20

01
20

00
20

99
19

98
19

97
19

96
19

95
19

94
19

93
19

92
19

91
19

-10%

19

90

0%

Year

Mean Profit Contribution by Revenue Type & Firm Age

80%

40%
20%

-40%

25

23

21

19

17

15

13

11

-20%

0%
1

Mean Percent Contribution

60%

Product Profit Mean

Service Profit Mean

-60%
-80%
-100%
-120%
-140%
Age

Product profitability = (product sales (product cost + R&D)) / product sales

Service profitability = (service & maintenance revenue service & maintenance cost) / service & maintenance revenue
Notes: Product profit mean corrected for outliers, eliminating values outside 1 standard deviation

33

Median Operating Income by BM and Year

20%

0%

19
90
19
91
19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
20
03
20
04

Median Operating Income

10%

-10%

Products
Hybrid Product
Hybrid Balanced

-20%

Hybrid Service
Total Sample

-30%
-40%
-50%
Year

Note: Service group not included because small sample size median was always negative and below the group

34

Market Value by Business Model

Mean Market Cap by BM and Year
4000

Mean Market Cap ($M)

3500
3000
Product

2500

Hybrid Product
Hybrid Balanced

2000

Hybrid Services
1500

Total

1000
500

19
90
19
91
19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
20
03
20
04

Year

NOTE: Excludes Microsoft and Services group

35

Survivor/Exit Comparison
T-test of Difference of Means for Product Firms
Survivors
Total Sales

Exited Firms

$371M

$104M

Total Sales Growth

0.69

0.77

Product Percentage

0.60

0.59

Product Growth

0.60

0.72

Service Growth

0.73

0.88

Gross Margins

0.66

0.64

R&D %

0.27

0.35

SGA %

0.84

1.03

Operating Margins

-0.68

-0.85

* Means significantly different at 5% level

For all % except product growth, eliminated outliers by taking all values within 1 S.D.

Product firms
who exit are
smaller and
spend more to
generate similar
growth

36

Maintenance Contribution?
Sample: 598 data points of firms per year that broke out
maintenance from other service revenues (i.e. probably a
biased sample favoring firms with large maintenance %)
Mean of 61% maintenance as % of total service revenues
Adj. mean of 55% if eliminate 75 data points of firms per year
reporting 100% maintenance
Ran random effects regression using all observations in which maintenance was
broken out.
Dependent variable: service margins
Explanatory Variable: maintenance as % of services
Control Variables: age, size, market, year
Maintenance comes out significant with a coefficient of .53.

10% increase in maintenance as a % of

service = 5.3% increase in service margins!!

37

Strategic & Operational Challenges

How Manage the Crisscross?
Best balance of products vs. services & maintenance?
What new products to generate services & maintenance?

How Servitize Products?

How add special value and revenue opportunities?
How make products stickier, less commodity-like?

How Productize Services?

Create two organizations within one?
How develop standardized products, platforms, or customized
solutions more efficiently reusing components, leveraging
knowledge, tools, & best practices across customers/projects
38

Public IT Services Firms in US

600
500
400
300
Projection

200
100

19
77
19
79
19
81
19
83
19
85
19
87
19
89
19
91
19
93
19
95
19
97
19
99
20
01
20
03

#of companies
39

Revenue Breakdown (estimate)

120%

Projection
100%
80%
60%
40%
20%

19
77
19
79
19
81
19
83
19
85
19
87
19
89
19
91
19
93
19
95
19
97
19
99
20
01
20
03

0%

% service

% product

40

Comments on Preliminary Data

IT Services business also undergoing similar
shakeout and commoditization?
Number of publicly listed firms in our database
dropped from ca. 550 in 2000 to under 400 in
2005 (preliminary estimate, multiple SIC codes)
Product sales used to be 20% of service firms
revenues; sharp decline to 3% today (estimate)
41

Challenges for IT Firms

Product revenues shrinking for Western software
product companies, so they will challenge IT
services companies for service & maintenance
revenues, both from onsite and off-shore centers
IT service companies in India, Japan, US, and
Europe must figure out how to compete beyond
process & quality competence or low wages
(India) or growth rates & margins will decline
42

One Key Option

Develop more Japanese-made products or
semi-products, for license fees and to drive
service & maintenance revenues (full products +
tools, or reusable frameworks, components)
US & European enterprise products not so well suited
to developing markets in Asia

But beware of collapsing product prices; so the

business model should be HYBRID
43

Second Key Option

Find ways to differentiate services
Indian IT services companies differentiated from US,
Europe & Japan competitors by process maturity,
scale, technical skills, English language, low (but
rising) labor costs.
But to outsiders, the Indian IT firms all look alike,
except for different sizes & slightly different prices

How can Japanese services firms differentiate

themselves and prevent copying of their best
practices or prevent Indian competition?
44

Path Towards Differentiation

Complex management of multiple variables
Executive leadership, marketing & sales capabilities,
recruitment strategies, training techniques, process & quality
management, technology depth, knowledge management,
one-to-one customer relationships, semi-products business
through tools or reusable platforms
Need to have a primary base but also be global & local

Need to think harder about Services R and D

Need to impact the brand (= higher prices),
scope economies (= lower costs), and ability to
lead the future (= thrive, not just survive)
45