R.D.Fernandez et al - Software rates vs price of function points

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

Raúl Fernández

SW Productivity Consultant

Insert here a picture

An Updated Cost Analysis

2IT 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


BackgroundOver 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.


BackgroundThe basic numbers of the 2012 sample were:

•10 Clients

•14 Providers

•3,405 Projects

•196,356 UFP

•2,168,192 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

0

1 0

2 0

3 0

4 0

5 0

6 0

7 0

8 0

9 0

1 0 0

1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 2 3 2 4 2 5 2 6 2 7 2 8 2 9 3 0 3 1

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


Background

0

1 0

2 0

3 0

4 0

5 0

6 0

7 0

8 0

9 0

1 0 0

1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 2 3 2 4 2 5 2 6 2 7 2 8 2 9 3 0 3 1

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

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

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

CLIENT-PROVIDER

MULTI-PROVIDER CLIENTS

ONE-PROVIDER CLIENTS


ObjectiveLEDAmc 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 benchmarkingThe 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?


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

2014 vs 2012 benchmarking

CLIENT-PROVIDER

The rates have remained almost the same.


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



Apparently there is no relation between FP price and rate…

CLIENT-PROVIDER

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

Rates vs FP Price in 2014


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



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


P1 P1P2 P2P3 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.


INPUTS (2013)


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

21



TOTAL 1 2PROV 1 PROV 2

INPUTS (2013)

Case studies

TOTAL 1 2

� 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


21

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)



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


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.


INPUTS (2013)


Case studies

TOTAL 1 2 3 4

1. New provider with

lower rate

2. In-house development

rate

1 2

INPUTS (2013)


PROV 1 PROV 2 PROV 3 PROV 4


TOTAL 1 2 3 4

Case studies

TOTAL 1 2 3 4

1. New provider with

lower rate

2. In-house development

rate

1 2

� Logical behavior of FP

price

� In-house

developments at lower

FP cost than in

outsourced scenario

INPUTS (2013)





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


� FP price controlled

� Outsourcing process controlled

2014 situation



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


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.


INPUTS (2013)


TOTAL 1 2 3 4 5 6

Case studies

� Renegotiation of rates

Ratio 2012: 1.1

Ratio 2014: 1.0

INPUTS (2013)



TOTAL 1 2 3 4 5 6

TOTAL 1 2 3 4 5 6

Case studies

� Renegotiation of rates

Ratio 2012: 1.1

Ratio 2014: 1.0

� FP cost control

Ratio 2012: 2.3

Ratio 2014: 1.7

INPUTS (2013)



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


� Uniformity of FP price

� 9.5 million euros saved

2014 situation



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.

R.D.Fernandez et al - Software rates vs price of function points

Business

highest fp price

higher fp price

lower rates

lower productivity

confidence conference

development projects

software development

productivity models