Products vs. Services::Which is the Better Business Model, Which is the Better Business Model, in Software and Other Industries? in Software and Other Industries?

Michael A. CusumanoMIT Sloan School of Management

cusumano@mit.edu © 2006

2

3

4

The Global Software Industry

0

20

40

60

80

100

120

140

160

180

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Year

Valu

e in

$B 3rd Party Services

Products Software Servcies

x10

5

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?

6

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.

7

Role of Services in Life-Cycle Dynamics & Platform Transitions?

Performance

Time

Ferment

Takeoff

Maturity

Disruption…..Services….?

8

Typical View of Services in High-Tech Companies?

“Services will be the graveyard for old tech companies that can't compete."

Scott McNealyCEO, Sun Microsystems

Referenced in N.Y. Times, 9-16-04

9

The Problem with The Problem with ManyMany ServicesServices

• 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

10

Business Objects: Gross Margins

0

20

40

60

80

100

120

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

Produc ts

Serv ic es

11

The Problem with The Problem with ManyMany ProductsProducts• 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

12

Siebel

0

200

400

600

800

1000

1200

1995 1996 1997 1998 1999 2000 2001 2002

$ m

illio

n

Products

Services

Siebel

0

50

100

150

200

250

300

350

1999 2000 2001 2002

Products

Services

Total

Siebel

0102030405060708090

100

1995 1996 1997 1998 1999 2000 2001 2002

% o

f Tot

al R

even

ues

ProductsServices

13

PeopleSoft

0

50

100

150

200

1999 2000 2001 2002

Products

Services

Total

PeopleSoft

01020304050607080

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

ProductsServices

PeopleSoft

0200400600800

1000120014001600

19921993

19941995

19961997

19981999

20002001

2002

Products

Services

14

Oracle

01000200030004000500060007000

1992

1994

1996

1998

2000

2002

$ m

illio

n

Products

Services

Oracle

0

50

100

150

200

1999 2000 2001 2002

1999

= 1

00 Products

Services

Total

Oracle

0

10

20

30

40

50

60

70

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

% o

f Tot

al R

even

ues

Products

Services

15

SAP

0100020003000400050006000

19921994

19961998

20002002

mill

ion

euro

s

Products

Services

SAP

0

50

100

150

200

1999 2000 2001 2002

Products

Services

Total

SAP

01020304050607080

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

% o

f Tot

al R

even

ues

ProductsServices

16

IBM

0

10

20

30

40

1992

1994

1996

1998

2000

2002

$ bi

llion Software

Services

IBM

90

95

100

105

110

115

120

1999 2000 2001 2002

1999

= 1

00 Products

Services

Total

IBM

01020304050607080

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

% of T

ota Re

venu

es

SoftwareServices

17

Business Objects

0

50

100

150

200

250

300

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

$ m

illio

n ProductsServices

Business Objects

0

50

100

150

200

250

300

1999 2000 2001 2002

1999

= 1

00 Products

Services

Total

Business Objects

0102030405060708090

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

% o

f Tot

al R

even

ues

ProductsServices

18

Business Objects Products vs. Services

0

20

40

60

80

100

19931994

19951996

19971998

19992000

20012002

20032004

2005

ProductsServices

19

Services & Maintenenance as Percentage of Total Revenues

0102030405060708090

100

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

Business Objectsi2SeibelPeoplesoftOracleIBM S&SSAPCompuware

7

Three Business/Life Cycle Models

0102030405060708090

Produc

ts

Hybrid

Solutio

ns

Service

s

% o

f Tot

al R

even

ues

ProductsServices

21

Typical Breakdown Per $1.00 of Enterprise Product Revenue

100 %20 %50 %30 %

$3.15$0.60$1.55$1.00TOTAL

$0.15$0.15Year 5

$0.15$0.15Year 4

$0.40$0.15$0.25Year 3

$0.45$0.15$0.30Year 2

$2.00$0.00$1.00$1.00Year 1

TotalMainte-nance

Software Services

SoftwareProduct

22

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 firm’s 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?

23

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 no-breakout firms).

• Now doing exploratory analysis• Also starting database of non-software firms

24

Public Software Product Companies Listed in US (% Annual Product Sales, 378 firms & 3314 yearly observations)

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).

010

020

030

040

0F

requ

ency

0 . 2 . 4 . 6 . 8 1p r o d p 2

100% Product

100% Service

Hybrid Balanced, 34-85%

Hybrid Service

HybridProduct

25

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: 76Product: 330HybridS: 464HybridB: 2206HybridP: 302Total: 3378

1702451019132004

19645211312152003

25535716610192001

296103820817231999

31373420734311997

24081916226251995

Total100%

ServiceHybrid Service

Hybrid Balance

Hybrid Product

100% ProductYear

Software Product Firms by Business Model

Percentage Breakout By Year

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Year

% o

f Tot

al

ServiceHybrid SerHybrid BalHybrid ProdProduct

28

Average Revenue Breakout by Age

0%

10%

20%

30%

40%

50%

60%

70%

80%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Age

Ave

rage

Per

cent

age

Products

Services

Average Revenue Breakout by Year

0%

10%

20%

30%

40%

50%

60%

70%

80%

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Year

Ave

rage

Per

cent

age

ProductServices

29

Service-Maintenance % by Product Type and Firm Age

0%

10%

20%

30%

40%

50%

60%

70%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Age

Ave

rage

Ser

vice

Con

trib

utio

n

TotalInfrastructure

Pre-packaged Apps

Service-Maintenance % by Product Type and Year

0%

10%

20%

30%

40%

50%

60%

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

Year

Ave

rage

Ser

vice

Con

trib

utio

n

TotalInfrastructurePre-packaged Apps

30

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.

31

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?

32

Grow

th index

Time

Services

Products

Grow

th index

Time

Services

Products

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

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

33Product profitability = (product sales – (product cost + R&D)) / product sales

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

Mean Profit Contribution by Revenue Type & Firm Age

-140%

-120%

-100%

-80%

-60%

-40%

-20%

0%

20%

40%

60%

80%

1 3 5 7 9 11 13 15 17 19 21 23 25

Age

Mea

n Pe

rcen

t Con

trib

utio

n

Product Profit MeanService Profit Mean

Mean Profit Contribution by Revenue Type & Year

-10%

0%

10%

20%

30%

40%

50%

60%

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

Year

Mea

n Pe

rcen

t Con

trib

utio

n

Product Profit MeanService Profit Mean

34

Median Operating Income by BM and Year

-50%

-40%

-30%

-20%

-10%

0%

10%

20%

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

Year

Med

ian

Ope

ratin

g In

com

e

ProductsHybrid ProductHybrid BalancedHybrid ServiceTotal Sample

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

35

Market Value by Business Model

NOTE: Excludes Microsoft and Services group

Mean Market Cap by BM and Year

0

500

1000

1500

2000

2500

3000

3500

4000

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

Year

Mea

n M

arke

t Cap

($M

)

ProductHybrid ProductHybrid BalancedHybrid ServicesTotal

36

Survivor/Exit Comparison

* Means significantly different at 5% level

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

Survivors Exited Firms

Total Sales $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 *

T-test of Difference of Means for Product Firms

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

37

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% maintenanceRan random effects regression using all observations in which maintenance was broken out.Dependent variable: service marginsExplanatory Variable: maintenance as % of servicesControl Variables: age, size, market, yearMaintenance comes out significant with a coefficient of .53.

10% increase in maintenance as a % of service = 5.3% increase in service margins!!

38

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

39

0

100

200

300

400

500

600

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

#of companies

Public IT Services Firms in US

Projection

40

Revenue Breakdown (estimate)

0%

20%

40%

60%

80%

100%

120%

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

% service % product

Projection

41

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)

42

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

43

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

44

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?

45

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)