Objects of Desire 2000

1
Objects of Desire 2000

• Karen D. Boucher
• karen_at_standishgroup.com

2
The Standish Group

• Since 1985
• Headquarted in Massachusetts
• Primary Research Based
• Surveys, Focus Groups and Personal Interviews
• Customer Focused Research
• Analyst to Client Ratio Max 1-15

3
Agenda

• Why are we here?
• How far have we come?
• What have we learned?

4
New Starts Applications(in thousands)
5
Technology is not the business, it is the
delivery of the business
6
Case Study The Bank

• Bank Tellers
• Types of services Withdrawals, deposits, balance
  inquiry, funds transfer, etc.
• Customers limited to bankers hours
• Expensive and time consuming for the bank
• How did they deliver People and Applications

7
Case The ATM Bank

• The ATM Machine
• Types of Services Withdrawals, Deposits, Balance
  Inquiry, Funds Transfers
• Self-service for customers
• Cheaper delivery of service for the bank
• How did they deliver?
• New Applications

8
Case Internet Banking

• The Internet Client
• Types of Services Deposits, Balance Inquiry,
  Funds Transfers, Loan Services
• Self-service for customers
• Cheaper delivery of service for the bank
• How did they deliver?
• New Applications

9
What can technology do?

• Saves the business money
• Delivers new services faster
• Delivers information to partners faster
• Shows the value of the business to more customers

10
Technology will continue to change

• How do you maximize the potential, minimize the
  risk and decrease the cost of this change?

11
Objects on the Rise
103,600 OO Apps.
12
The Object is.Faster, better delivery of the
business
13
Why this study?

• Reality check
• Assist organizations stuggling with implementing
  successful OO applications
• Determine what tools / techniques / procedures
  assist OO development groups

14
Methodology

• First study in summer 1997
• Second spring 2000
• 34 / 61 Personal Interviews

15
1997 Versus 2000

• Difficult to find users
• High reuse numbers - the norm
• ROI needed to be proven
• Language choice difficult decision
• Tools difficult to find
• Skills difficult to find

• Easy to find users
• Reuse?
• No justification required
• JAVA and C (of course)
• Tools in abundance
• Skilled staff still difficult to find

16
As far as Im concerned, I know the benefits of
it and Im not worried about the return on the
value. I know it comes.
17
Technology Advancement from 1997 to 2000

• Language Maturity
• Longevity of Java not in question
• Speed no longer an issue
• Standards have evolved
• Tools have evolved
• ORBs ubiquitous
• Development tools follow OO methodology
• Analysis and design tools
• Testing tools

18
We were tracking the technology for at least six
months, but it wasnt until Java hit the market
that we started to really do things with it.
19
Language Choice
Positive Rating
Negative Rating
20
Barriers to Reuse
21
Barriers to Reuse

• Change of mindset

22
Programmers are a different animal. They want
to reinvent the wheel every five minutes. They
can do it faster, better, stronger themselves.
23
Barriers to Reuse

• Change of mindset
• Rush to deliver

24
Theres no way to get around the fact that there
is a learning curve and that first project is
going to take significantly longer than
traditional development.
25
Barriers to Reuse

• Programmers
• Rush to deliver
• Just a language?

26
For our first few projects the problem (with
getting reuse) was that we were just programming.
Even though we were using an OO language, we
were still just programming.
27
Barriers to Reuse

• Programmers
• Rush to deliver
• Just a language?
• Project of the moment

28
Im a big believer in reuse, Ive seen it work
and I know that if you put in the time up front
it pays off. But sometimes the alligators are
nipping at your ankles, its difficult to think
about how much better things are going to be
tomorrow if you battle it out.
29
Lessons Learned

• Experience Counts

30
In my experience, experience counts.
31
Lessons Learned

• Experience Counts
• The more OO projects, the more reuse

32
Averaged 67 increase in reuse for subsequent
projectsDirectly leading to decreased
development time / cost / manpower
33
Lessons Learned

• Experience Counts
• The more OO projects, the more reuse
• Invest the time up-front

34
Some of the stuff we made was specific to the
project rather than making it more generic and
bringing outside of the objects things that were
specific to the project. So when we went back to
use them they had to be modified to a greater
degree than if we had done more investigation up
front.
35
Third time is a charm.
36
Lessons Learned

• Experience Counts
• The more OO projects, the more reuse
• Invest the time up-front
• The right tools analysis / design, middleware
  frameworks

37
You cant just throw every thing that youve got
into one box and have it work together. Youve
got to integrate it in some manner.
38
70 of code for a typical application is
infrastructure code
39
The Secret is in the Infrastructure

• Application Infrastructure Determines
• How users access data, other applications and
  share knowledge
• How fast this can be done
• How secure this is
• How available the services are
• Etc.

40
Standish research shows that those projects which
include thecreation of custom middleware have a
zero chance of being completed on time and on
budget.
41
Lessons Learned

• Experience Counts
• The more OO projects, the more reuse
• Invest the time up-front
• The right tools analysis / design, middleware
  frameworks
• Incentives / penalties program

42
Somehow youve got to promote reuse, and it may
takes threats or bribes to make it happen.
43
Lessons Learned

• Experience Counts
• The more OO projects, the more reuse
• Invest the time up-front
• The right tools analysis / design, middleware
  frameworks
• Incentives / penalties program
• Library (repository) and librarian

44
Once we really started increasing the number of
OO projects it was an absolute requirement to
have a central repository.
45
Recap

• Theres no substitute for experience
• Take advantage of the tools available
• Institute programs to promote reuse

46
The Future...

• Component Marketplace
• Framework Market
• Modelling Tools Market

47
Questions?