Thursday, June 30, 2005

Negotiate the business requirement and make the scoping decision


As a product manager and a business analyst, you are going to
present the suggested scope to your development. Do you see
this is a "negotiation" process? Do you see that typically the
product management wants to get the items IN and the
development tries to get the scope OUT? Do you see that you
ever need to ADD MORE into the scope so you can GET MORE
after the "negotiation"?

I do not have a clear answer for all the above questions, but I
can see that these are two models, two types of assumptions,
and two different processes:

1. People playing different roles sometimes need to play the
devil's advocate. The best decision is reached via the process
of negotiation among the people representing different interests.
In the software requirement scoping case, the product manager
represents the "requirement" side, or the "demand" side, and
the development represents the "delivery" and the "supply"
side. The assumption is that we cannot avoid people having
different interests. Instead of trying to solve the conflict of
interests within one's mind, disputing, debating, discussing,
negotiating, and finally reaching an agreement is the way to go.

2. Defining the scope can follow the standard decision making
process and can be scientifically done. The assumption is that
there is really no conflict of interests. Both product management
and development are toward to the same goal - delivering the
right solution timely to satisfy the targeted customers. Making
profit for your organization in a short run as a mean to maximize
the shareholder's value in a long run. The customer satisfaction,
effort, time to market, risk of delivery, avoiding rework, and
quantity should all be considered. They can be considered in a
systematic way as an individual or a team effort. A product
manager who propose the scope should consider both side or
even more factors as mentioned above. The reasoning and
the justifications are discussed with the development for
ensuring the completeness of proposal.

I do not have an answer. I can see both models explain the real
world situation. What do you think?


Monday, June 27, 2005

ERP and Cost Control

Today is the first day of the AACE annual meeting. I attended the keynote from Joel, Primavera and several good technical sessions.

One of the technical meeting I attended is about the Project Cost Management and System Integration. It discussed what is required for an integrated cost control solution. They described that it should allow you to have best-breed of the following systems, while you should not compromise your cost control :
  • Accounting
  • Change Management
  • Estimation
  • Scheduling
  • Procurement
  • Time Reporting
  • Contract
I agree with most of what they said about the cost control system. The presentation provided a very good overview of what project cost control system is about, and reach a very similar conculsion as one we did several week ago for my boss.

The most interesting topic in the presentation is that it compares different solutions:
  • Spreadsheet based
  • Cost Loaded Scheduling
  • ERP
They concluded that none of these solutions is ideal. You should really have a separate, dedicated cost control system. I will agree with their conculsion if their assumption about ERP are true:
  • ERP offers a project accouting solution, not really a project cost control solution.
  • Project Accounting is concerned with expenditures and based on GAAP.
  • You need to collect data and report against your chart of accounts, a single, rigid, non-project specific structure.
  • You can only have a single WBS, which is in the cost accounting system and is not controlled by the project cost control people.
  • ERP does not offer robust budgeting and forecasting functionalities.
  • ERP does not provide the end user customization.
  • ERP provided good "online" reports, but poor "printed" reports.
  • ERP does not do trending or what-if analysis.
These may be true for some ERP systems or for the situation you may face 5 to 10 years ago. Unfortunately, most of them may not be true today. At least for the application I am being involved in now.

I am appreciated the analysis they did. It is still an outstanding presentation. It highlight what the ERP vendor should work on. I do strongly believe that customers are looking for an single integrated system that can address the needs described earlier. If they cannot find a single system from a single vendor, they will take a best of breed approach.

If you are interested in what they said in the presentation, take a look at their Project Cost Management Solution White Papers. They are very well written with good logic and reasoning.


.

Sunday, June 26, 2005

AACE Training and CCE/CCC Exam

I attended the four day training courses and took the Certified Cost Engineer exam in the AACE conference.

The training is for preparing the cost engineer exam, which covers several topics, such as cost estimating, cost control, project scheduling, project management, and engineering economics.

This is the most tough certification exam I have ever attended. It is cross several disciplines and is very very mathematical.
  • It tested the probability of the normal distribution. Fortunately I learned that in my undergraduate as a major in statistics.
  • It tested "computer operation" terminology. It should be an easy one, but it is actually not. AACE has it own definitions for the terms, such as "byte", "operating system", "mainframe system", etc.
  • It ask people to manually do the CPM scheduling using the PDM method. I learned this in the project management course for the PMP certification. The one tested in the CCC exam is more closed to the practical situation. It asked you how much you will lost if the task X is delayed for the N days. You need to do forward pass, backward pass, identify the critical path, etc.
  • You need to know the different estimation classes. AACE has its own definition while ANSI has another. Each class is for different purposes and is under different range of details. This may be common sense for cost estimators, but definitely not for someone who do not do this everyday. You can use different method for different class. You have to memorize all that if estimation is not your daily job. You also actually need to do a lot of mathematics calculation in the cost estimation exam.
  • The EVM analysis is the most easy one in the cost exam. All simple number calculation: SV, CV, CPI, SPI, and EAC. The hard one is to do the calculate for the productivity analysis. I almost miss this one. It asked questions about when you should which contract type. It is a very practical situation. I gave up this one since the answers are all very unclear. I selected to work on the mathematics instead.
  • Out of the four parts. the engineering economics is the hardest part. I took some financial courses in the past and have been taught the different profitability analysis or the investment evaluation methods several times. The CCC exam has the most difficult question and requires a lot of real world knowledge. You need to know MARCUS as the depreciation method required by US IRS. When you do the cash flow analysis, it asks you to consider that deprecation, which may affect your income tax, and income tax saving should be considered as the cash flow. I have not seen this complicate question in any exam. Meanwhile, the property tax is your out flow. You need to compare buy vs lease situation. Financial calculator is required.
  • Some of the exam questions are related to manufacturing, but most of them are engineering and construction related. I have never heard "battery limit" area. It mentioned in the exam and I missed that one.
Before you are siting on the exam, AACE requires you to send them a technical paper. Getting certified as a cost engineer is really difficult.

Saturday, June 25, 2005

Value Engineering for Software Construction

Many years ago, I lerarned that the "Design Pattern" concept was borrowed from the building architecture field to the software architecture field. I found that it was interesting to see how the cross field interaction can bring the good idea.

This week, I am in the AACE conference in New Orleans, LA. AACE is an association for cost engineers, many of who come from the construction industry. I learned that in construction, people are very concerned about "getting the great value possible for every dollar spent on captial construction". A set of "Value Engineering" or "Value Analysis" tools and techiques were thus developed to address the needs. Basically it is used to select the best design concept, which can yield least lifecycle costs. The concern is not just to gain the cost saving for the project during the construction phase, but for meeting operatability, maintainability, reliability, durability, and other criteria from the owners.

In software construction, we have the same concern. John Wookey called it "Superior Ownership Experience". Can we borrow the value engineering methodologies to the software application architecture and construction process?

Tuesday, June 21, 2005

Business Requirement vs. Product Requirement

Are these two separate requirement documents? Some people believe so while others do not.

I feel that we should separate these as two separate deliverables.

The business requirement document is sometimes referred as the marketing requirement document (MRD). It discusses the requirements at a very high level from a business perspective. It describes the problem that need to be resolved, the market opportunity, and sometimes the competitor offerings. It justifies if a development project should be initiated. The business requirement document is not tie to any given development release and should avoid using the product specific terms. Sometimes the requirements are described in a such neutral way that you cannot tell which module or applications will be impacted. Actually, the business requirements is more like a RFP from some prospects. It is a user view of the solution.

The product requirement document is sometime referred as the product requirement specification or functional specification. It describes the expected functionality from a product. It is an outcome from the requirement analysis process. It interprets the business requirements by mapping and describing them using the product terminology. It is the document to communicate the business requirement to the product stakeholders. A product requirement document should be written with a product release in mind although not all features identified need to be scoped in. The key point is that the detail scoping decision can be made based on the product requirement document. It is the product requirement document identifies the scope of a development project. In the PMBOK, the product requirement document can be regarded as the scope statement for initiating a development project. It defines the features and is the product view of the solution.

At some stage, the product requirement should be baselined for functional design. It should be subject to the change control process for scope control. Scope creeping can thus be avoid. However, it worth mentioning that the requirement analysis is a iteration process. You should not be too rigid on changing the product requirement during design phase. You should not feel hesitate to contact customers to clarify the requirement.

Here is a very good document I found on the web.

Monday, June 20, 2005

Structure mapping

Moved to my new blog site: dylanwan.blogspot.com - Work Breakdown Structure Mapping

Chinese

中文嘛A通

Link from google

This is a link to google.

Fwd: Try gmail


Gmail allow me to post from the Internet without using netscape.

I like it.

Center

Left

Right
Highligh this

Make Bigger.

or small.


Saturday, June 18, 2005

reasons to start a blog

Mia vita: Benvenuto!!

Current Visitors

I should study how to add the "Current Visitors" section.
Sexy Pix Mega Post: Oscar HQ pictures: "Current Visitors"

this is a chinese blog

��TrapNest�� Bilingual Version

this one has a good format too.

a very plain format - northern apostolic mom

Try fonts and other formatting.

Yahoo! did not do the HTML link well.  I can add a link from the netscape email client
  • one
  • two
  • three
How about this

I am using the fix size font.

And indentation
this text should be indented.
:-) ;-) :-D

Try a table as well

col1
col2
cell1
cell2

wip

Hope you can see the image.




Try from Netscape

Here is another link to oracle.

Wednesday, June 15, 2005

Another test

Test

Sunday, June 12, 2005

I want to add a link

"Link" from email is not working.
Here is a link to www.oracle.com

Can I use a hyperlink?

Here is the <a href="http://www.oracle.com">link</a> to
oracle.com.

Try one more time

I do not like to see "yahoo".

Try to post from an email

This message is posted from an email.
.

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com

Start Here

This is my first posting. Just a test.