Software Rates vs Price of Function Points An Updated Cost ...IT Confidence 2014 – October 22, 2014 2 Software Rates vs Price of Function Points: A cost analysis Goals of the presentation

http://itconfidence2013.wordpress.com

Software Rates vs Price of Function Points

2°International Conference on

IT Data collection, Analysis and Benchmarking

Tokyo (Japan) - October 22, 2014

Rafael de la Fuente

Founder and CEO of LEDAmc

Dácil Castelo

Productivity&Estimations Area

Director

Raúl Fernández

SW Productivity Consultant

Insert here a picture

An Updated Cost Analysis

2 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com

Software Rates vs Price of Function Points: A cost analysis

Goals of the presentation

Nowadays, in Software Development Contracts the key elements determining the price are: the Rate and the Effort. The possibility to measure the quantity of software produced (the size in Function Points) allows us to assess whether there is a logical connection between the price of the projects and the software actually produced.

The main goal of this presentation is to show what happens with the FP

price in several scenarios:

G1. Implementing estimation models.

G2. Implementing productivity models.

G3. Introducing new competitors in a productivity model.

G4. De-localizing providers.


• Background

• Objective

• 2014 vs 2012 benchmarking

• Case studies

• Conclusions

Agenda


Background Over the last seven years, LEDAmc has managed the productivity of more than 12,000

development projects of 10 significant clients in Spain (mostly telecommunication and

financial companies). They were mainly small enhancement projects.

The main goal of the measures is to control large contracts of Adaptive Maintenances,

which involves the highest percentage of our clients’ development budget.

In 2012 LEDAmc developed a benchmarking study with some of these projects (we

selected 3,405). The purpose of the study was to analyze the relationship between the

rates of the different software providers and the price of the Function Point that they

were offering their clients. This study was presented:

• In the UKSMA conference in October 2012

• An updated study, in the IT Confidence Conference in Rio, in October 2013.


Background The basic numbers of the 2012 sample were:

• 10 Clients

• 14 Providers

• 3,405 Projects

• 196,356 UFP

• 2,168,192 Man Hours

• 69,926,907 Million Euros

And the main conclusions were:

Pressure to lower rates ends up in lower productivity and higher FP price.

There is not a logical relationship between Rates and FP Price.

The performance of the providers changes dramatically among clients.

Clients with only one development provider have the highest FP Price.


Background

CLIENT-PROVIDER

R

A

T

E

F

P

P

R

I

C

E

0

10

20

30

40

50

60

70

80

90

100

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 31

Pressure to lower rates ends up in lower productivity and higher FP price.


Background

0

10

20

30

40

50

60

70

80

90

100

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 31

The same provider for different clients, offers significant differences in FP price.

R

A

T

E

F

P

P

R

I

C

E

CLIENT-PROVIDER


Background

0

10

20

30

40

50

60

70

80

90

100

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 31

F

P

P

R

I

C

E

The price of the FP with only one provider is higher.

CLIENT-PROVIDER

MULTI-PROVIDER CLIENTS

ONE-PROVIDER CLIENTS


Objective LEDAmc has been continuously working for most of those clients, implementing Software

Development Productivity Models over the last two years.

Our productivity benchmarking data base has grown: It has more data for the same

clients and new clients have been included.

The main goal of this presentation is to show what has happened in these two years with

clients who have worked hard to improve their software development productivity.

We will first explain the general evolution of the Function Point Price for those clients,

comparing 2012 and 2014 benchmarks.

Afterwards, we will focus on three case studies that are especially significant.

Finally, we will draw conclusions on the economics of software management, based on

our results.


2014 vs 2012 benchmarking The 2014 sample were:

• Clients that were in the 2012 study and have implemented some estimation or

productivity control

• 4 Clients

• 12 Providers

• 883 Projects

• 44,897 UFP

• 415,768 Man Hours

• 15,428,772 Million Euros

What has happened with the FP price?


0

10

20

30

40

50

60

70

80

90

100

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

2014 vs 2012 benchmarking

R

A

T

E

CLIENT-PROVIDER

The rates have remained almost the same.


0

10

20

30

40

50

60

70

80

90

100

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18


The FP price is more homogeneous in 2014 than 2012.

CLIENT-PROVIDER

F

P

P

R

I

C

E



R

A

T

E

Apparently there is no relation between FP price and rate…

F

P

P

R

I

C

E

CLIENT-PROVIDER

0

10

20

30

40

50

60

70

80

90

100

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Rates vs FP Price in 2014


0

10

20

30

40

50

60

70

80

90

100

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18


.. but if we group the data by client, the results look different.

CLIENT A CLIENT B CLIENT C CLIENT D


R

A

T

E

F

P

P

R

I

C

E


0

10

20

30

40

50

60

70

80

90

100

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18


There are big differences in the FP price for the same provider with different clients.

CLIENT A CLIENT B CLIENT C CLIENT D

R

A

T

E

F

P

P

R

I

C

E


P1 P1 P2 P2 P3 P3 P3



Conclusions

The results are homogeneous on a client basis because they have implemented

estimation or productivity control models that allow them control the FP price.

Because there is no transparency in software development and the public FP price does

not exist in the market, the FP price stills remains very different among the clients.

The natural differences of the systems architectures and processes should not imply

such FP price differences.

… now we will present some case studies highlighting the change that has been

produced for these clients who have introduced control systems.


Case studies

High FP price

Only one provider

No estimation control

Case Study 1 Competition among providers + Estimation Model

INITIAL SITUATION OF THE CLIENT (2012)


Case studies

During 2013, the implementation of a centralized mechanism of estimation control has

allowed to objectively contrast the costs offered by providers with corporate cost

strategies.

It includes functional sizing, as well as other sizing methods (for parametrization,

technical changes, test support,..)

Benchmarking studies helped establish a cost strategy and a target price for FP .

Introducing a new development provider fosters competitiveness among providers and

helps establish a FP price within market values.

Case Study 1 Competition among providers + Estimation Model INPUTS (2013)


0

10

20

30

40

50

60

70

80

90

100

TOTAL 1 2PROV 1 PROV 2

INPUTS (2013)

Case studies

1. The rate was only

reduced by 5%

2. New development

provider worked with

a similar rate

R

A

T

E

2 1



0

10

20

30

40

50

60

70

80

90

100

TOTAL 1 2PROV 1 PROV 2

INPUTS (2013)

Case studies

R

A

T

E

0

10

20

30

40

50

60

70

80

90

100

TOTAL 1 2

F

P

P

R

I

C

E

The FP price was

reduced by 44 %

Similar FP Price

among providers

1. The rate was only

reduced by 5%

2. New development

provider worked with

a similar rate

Case Study 1 Competition among providers + Estimation Model 2

1

PROV 1 PROV 2


Case studies

A good estimating process was used for all projects

44% reduction in FP price

More than 10 Million euros saved

FP Price within market values

Two competing providers in 2013 – Four in 2014

FINAL SITUATION (2014)



0

10

20

30

40

50

60

70

80

90

100

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 31

Case studies

Specialized providers

On-site development

High rates and very low FP price

Estimation and productivity control implemented

F

P

P

R

I

C

E


Case Study 2 De-localization of providers


Case studies

The client strategy was to proceed to the de-localization of the providers to a

typical outsourced model introducing new development providers with lower rates.

The client also decided to create an in-house development team in order to

preserve and extend the know-how.

Case Study 2 De-localization of providers INPUTS (2013)


Case studies

0,00

10,00

20,00

30,00

40,00

50,00

60,00

70,00

80,00

90,00

100,00

TOTAL 1 2 3 4

R

A

T

E

1. New provider with

lower rate

2. In-house development

rate

1 2

INPUTS (2013)


PROV 1 PROV 2 PROV 3 PROV 4


0,00

10,00

20,00

30,00

40,00

50,00

60,00

70,00

80,00

90,00

100,00

TOTAL 1 2 3 4

Case studies

0,00

10,00

20,00

30,00

40,00

50,00

60,00

70,00

80,00

90,00

100,00

TOTAL 1 2 3 4

R

A

T

E

1. New provider with

lower rate

2. In-house development

rate

F

P

P

R

I

C

E

1 2

Logical behavior of FP

price

In-house developments

at lower FP cost than

in outsourced scenario

INPUTS (2013)





0

10

20

30

40

50

60

70

80

90

100

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 31 32 33 34

Case studies Case Study 2 De-localization of providers

FP price controlled

Outsourcing process controlled

2014 situation

F

P

P

R

I

C

E



0

10

20

30

40

50

60

70

80

90

100

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 31

Case studies

Ratio of the rates: 1.1

Ratio of the FP Price: 2.3 F

P

P

R

I

C

E


Case Study 3 Controlling large maintenance contracts


Case studies

During 2012 the client implemented a productivity control for maintenance

development.

An annual productivity SLA for all the providers was established.

The client also implemented a productivity and quality providers ranking.

Every two months the client presented to the providers their position in the ranking.

Case Study 3 Controlling large maintenance contracts INPUTS (2013)


0,00

10,00

20,00

30,00

40,00

50,00

60,00

70,00

80,00

90,00

100,00

TOTAL 1 2 3 4 5 6

Case studies

R

A

T

E

Renegotiation of rates

Ratio 2012: 1.1

Ratio 2014: 1.0

INPUTS (2013)



0,00

10,00

20,00

30,00

40,00

50,00

60,00

70,00

80,00

90,00

100,00

TOTAL 1 2 3 4 5 6

0,00

10,00

20,00

30,00

40,00

50,00

60,00

70,00

80,00

90,00

100,00

TOTAL 1 2 3 4 5 6

Case studies

R

A

T

E

Renegotiation of rates

Ratio 2012: 1.1

Ratio 2014: 1.0

F

P

P

R

I

C

E

FP cost control

Ratio 2012: 2.3

Ratio 2014: 1.7

INPUTS (2013)



0

10

20

30

40

50

60

70

80

90

100

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 31 32 33 34

Case studies Case Study 3 Controlling large maintenance contracts

Uniformity of FP price

9.5 million euros saved

2014 situation

F

P

P

R

I

C

E



Conclusions

1. It is relatively easy to influence FP price. The main levers are:

>>> A centralized estimation control

>>> A development productivity control

>>> Several providers in competition

2. Working on those key elements the results are much better than working only

on the rates:

>>> The behavior of the unit cost of production (FP price) is more homogeneous.

>>> The cost of the development process can be controlled.

3. Diversification of providers with joint management results in a better uniformity

of the FP price.

Documents

Software Rates vs Price of Function Points An Updated Cost ...IT Confidence 2014 – October 22, 2014 2 Software Rates vs Price of Function Points: A cost analysis Goals of the presentation