Fundamentals of Project Management

Copyright © 2012-2019 Stephen
Ellwood, ETC BVBA
Stephen Ellwood, May 2015
Fundamentals of Project
Management

Ellwood, ETC BVBA
Agenda – Part 1
• Introduction, Course book 1
• The 6 Questions (Chapter 3)
• What are the Fundamentals?
• Making a Plan

Ellwood, ETC BVBA
Project Management is?
• A problem scheduled for execution
• A way to efficiently organise a team
• A way to manage technical [and commercial] risk
• The essential glue that holds a team together
• The difference between profit and loss
• From sport we learn that a group of talented individuals can
be beaten by a team of less talented players
– Part of the reason for this is that the team have a plan

Ellwood, ETC BVBA
Course book 1:
Fundamentals of Project Management

Ellwood, ETC BVBA
What is the Purpose of Project
Management?
• Management is not doing, so management is overhead, right?
• No! – we manage people to ensure:
– We are doing the right thing at the right time
– we are doing neither too little or too much
• Most importantly we are managing risk
– Not all risks can be seen at the beginning, some unforeseen things will
happen during the project but we still have to bring the project in on
time!

Ellwood, ETC BVBA
Project planning flow
Sales
Process Definition Design Verification Finishing
Risk
Planning
Understand
Customer
requirements
Win the business
Reflect the
customer
requirements
Identify and
reduce the risks
 Execute the
design
Monitor Progress
Take Corrective
actions
Verify the design
Take Corrective
actions
Document Design
Deliver to
Customer
Team
Size

Ellwood, ETC BVBA
The 6 Questions (Chapter 3)

Ellwood, ETC BVBA
The 6 questions of project management
1. What must be done?
2. How should it be done?
3. Who will do it?
4. By when must it be done?
5. How much will it cost?
6. How good does it have to be?

Ellwood, ETC BVBA
1. What must be done?
• The customer is the source of information
• Ask yourself:
– Are the customer requirements complete, unambiguous, actionable?
– Is the application (and its implications for critical specifications) clear?
• Your First Goal:
– Restate the problem within your frame of reference
– Reflect the customer requirements back to them as a functional
specification that they can sign off

Ellwood, ETC BVBA
2. How should it be done?
• This will normally be the easiest question to answer as it
should reflect your company’s expertise
• But, beware:
– Don’t do things “because you can” if there
is a better solution for the customer
– Do not go for the obvious
method/topology/process too quickly
• investigate alternatives

Ellwood, ETC BVBA
3. Who will do it?
• Pick the best team you can based on the constraints:
– Are the people with the most/right experience available?
– Which parts are least well defined? – put your best people here
• If a task is low risk, consider a person with less experience to
use it as a learning exercise
– People are happier when they are learning new things and will work harder
for you
– Your total team experience will be increased
• Take personality traits into account
– Some people like the detail, some people like solving difficult problems etc
– Team diversity can help overcome project obstacles

Ellwood, ETC BVBA
4. By when must it be done?
• Be aware that ASAP is not always the best answer
– The sooner the project is complete the quicker you get paid and the quicker
you can get a new project
– The customer always wants it ASAP – there is no downside for them in
wanting this
• However
– Doing things quicker requires more parallelization of tasks and a larger team
• Larger team means higher cost due to higher communication overhead
• More parallelization leads to more timescale risk as many tasks have to converge
together at the end
• Larger teams means fewer concurrent projects and more risk that a failing project
has knock on effect to other projects
• Try to find the “natural” (i.e. lowest effort) project plan then
sell that to the customer

Ellwood, ETC BVBA
5. How much will it cost?
• Customers usually want to pay a fixed price
– By the time Project Manager is given the project the cost has probably been
agreed
– A time budget in hours is given to the Project Manager when the project starts
• Customers often try change the spec’ after a project starts
– Make sure that the scope of the project is understood and look out for
“optional extras”
– Make agreeing the specification the first goal of the project and get the
specification signed off

Ellwood, ETC BVBA
6. How good does it have to be?
• Be Aware
– A good team player wants to do a good job
– Which specs are critical and which are goals?
• Do your homework
– Negotiate early with the customer to relax specifications as much as
possible
– Use prototyping to investigate the solution space
– Study feasibility to spread specification risk as widely as possible
• Not one supercritical task and lots of easy tasks

Ellwood, ETC BVBA
C = f(P,T,S)
• Cost is a function of Performance, Time, and Scope
• Which variables are fixed and which can you change
– You cannot create a plan when you have no freedom to create
• C, P, T & S should not all be fixed going into the project
– You will be held responsible, so you must have the authority to decide
• Your company has to operate in a competitive environment
– You will be asked to justify the budget
• Expect a lively debate if you propose a big increase to the budget
– The lower the cost the higher the profit
• We are all responsible to make money for our company

Ellwood, ETC BVBA
What are the Fundamentals?

Ellwood, ETC BVBA
Project planning, according to the book
1. If you have no plan, you have no control
2. The people who execute the plan should help plan it
3. Have the plan signed off in a meeting, not through the mail. A signature
from a contributor is a commitment, not a guarantee
4. Use exit criteria to determine when a milestone has actually been
achieved and the project is ready to proceed to the next step
5. Require signatures for changes in scope to alert everyone as to the
impact of the change on project costs, deadlines, etc
6. Risk analysis is part of planning. For every risk identified, develop a
contingency plan, when possible

Ellwood, ETC BVBA
1. If you have no plan you have no control
• Failing to plan is planning to
fail [Benjamin Franklin]
• Make the most realistic plan you
can and then stick to it
– “You can’t schedule creativity” .. but
we must pretend we can (page 32)
• Aim to reduce project risk as soon
as possible
– Front end loading
• If the plan is not working, don’t
abandon it – fix it!
With a
plan
Without a plan
Front end loading =
Take the pain early
(see also Fig. 1-3)

Ellwood, ETC BVBA
2. Those who execute should help plan it
• Good Judgment is the result of experience, but everyone’s
experience is different
– The time taken to do a task depends on the person who does the task
• Discuss the estimates with as many people as possible
– Getting multiple estimates helps to reduce estimate uncertainty
• It is unfair / self defeating to force aggressive estimates onto
people
– Its much less likely that people will strive to meet timescales they know are
impossible to achieve
• Be aware that the person who might perform the task most
efficiently might not be available
– Creating specialists is a double edged sword
– People like to learn so like variety

Ellwood, ETC BVBA
3. Have the plan signed off ..
• Once you have buy in from the team you can ask for
commitment:
– Signing off the plan shows a clear decision to “go for it”
– Print the project plan on an A0 sheet and hang it up to remind
participants of their commitment
– Encourage/expect project members to stand by their commitment

Ellwood, ETC BVBA
A commitment is a promise, not a
guarantee
• Some risks, even though known before hand, will be
unavoidable
• Unforeseen issues will occur, some beyond your control
• When forced, be prepared to modify/adapt the plan
– Be proactive and ready with a contingency plan
• The “Tom Tom” Algorithm

Ellwood, ETC BVBA
4. Use exit criteria
• All project activities result in a (single) deliverable
– Make sure the deliverable is well specified before the activity starts
• Activities with no deliverable are not part of the project
• Some deliverables build on other deliverables
– E.G. fabrication cannot occur until the design is finished
– By charting deliverable interdependencies we come to a Gantt chart
• Some deliverables have more significance than others
– Mark important deliverables as milestones, track them separately
– Payment milestones should be highlighted/have deadline dates attached to
them

Ellwood, ETC BVBA
5. Control Changes - internal
• “Control” in the context of projects
implies acting on information
– If you don’t react to a deviation, you are monitoring not controlling
– Compare progress to plan so that action can be taken to mitigate deviations
– A project will stay in control only if all team members stay in control of their
own work
• Project working times should be recorded daily
– People who capture their progress weekly rely on memory to estimate project
work. Such data makes for poor future estimating
• External audits help the team to improve performance
• The effort used to control a project should be
proportional
– E.G. Don’t spend €100 to purchase a €3 battery

Ellwood, ETC BVBA
5b. Control Changes - External
• The customer will always want more
– The more time passes the more time the customer has to think of
problems / improvements
– “Moving the Goalposts” makes it difficult to score
• Reflecting requirements back to the customer anchors the
goal posts and reduces the risk of changes
– It forces the customer to buy in to their own specification
– The customer must think about potential issues earlier (when there is still
time to make changes)
– It creates an anchor making changes more difficult for the customer
• Changes can still happen
– We remain in control
– The customer knows that extra functionality means extra cost

Ellwood, ETC BVBA
6. Manage the Risk
• Decision making is the fundamental skill in all Project
Management
– “Decide” comes from the latin meaning to cut-off i.e. a decision cuts off
other possible choices
– Making a decision too early / based on too little information increases
the likelihood that the optimal decision / path is missed
– Making a decision too late creates additional effort to pursue multiple
strategies, some of which will later be cut off
• By making decisions you exercise control over the project
– You cannot eliminate risk, you can only
follow a path that minimises it

Ellwood, ETC BVBA
6b. Manage the Risk – Front End Loading
• Front End Loading means taking the pain early so as to reduce
the pain later
• “Front-end loading is including robust planning and design early in a project's lifecycle (i.e.,
the front end of a project), at a time when the ability to influence changes in design is
relatively high and the cost to make those changes is relatively low. It typically applies to
industries with highly capital intensive, long lifecycle projects. Though it often adds a small
amount of time and cost to the early portion of a project, these costs are minor compared to
the alternative of the costs and effort required to make changes at a later stage in the
project.”
– Wikipedia

Ellwood, ETC BVBA
Front End Loading in Practice
• Spend quality time at the beginning to plan in detail
• Create prototype designs to explore high risk items
• Defer risky decisions until a critical mass of evidence appears
– Is this the correct choice given the constraints?
• Do NOT start the easy tasks first just because you can

Ellwood, ETC BVBA
Making a Plan

Ellwood, ETC BVBA
Making a work breakdown structure
• A WBS (chapter 4) subdivides a project into tasks
– A task has one or more inputs /dependencies and one output
– Smaller and better defined tasks are easier to estimate
• One cause of project failure is that something is forgotten
– Listing all WBS inputs helps prevent this problem
• A WBS does not show the sequence in which work is done
– That comes later from analyzing the inputs

Ellwood, ETC BVBA
Estimating Task Durations -1
• Estimation can only be done on the basis of experience
– Use experience (of something similar) to estimate new tasks
• Direct previous experience will give the best estimate
– Breaking things down will always increase accuracy (but takes time)
– Averaging estimates from multiple people increases accuracy

Ellwood, ETC BVBA
Estimating Task Durations -2
• If you cant estimate time directly, estimate complexity and
then multiply by a figure of merit based on past experience
– E.G. a document of 25 pages will take 5 days @ 5 pages per day
• An estimate should always be qualified:
– What is its accuracy?
– How was the estimate prepared?
– What assumptions / limitations does it contain?

Ellwood, ETC BVBA
Scheduling tasks to make the plan
• Schedule task based on their required inputs and outputs
• Schedule at a level of detail that can be managed
– No task should be scheduled with a duration much greater than 10
days. Subdivide longer tasks to achieve this objective

Ellwood, ETC BVBA
Activity on Arrow / PERT
• Activity-on-Arrow diagrams allow an easier assessment of the
impact of a slip on a project than do Gantt charts
– Unfortunately MS Project only makes Gantt charts
– Show the critical path through the project
• PERT charts make it easier to show risk on a project plan
– PERT Estimates include min, max and average durations
– PERT is not supported by MS Project

Ellwood, ETC BVBA
Analyzing the Critical Path
• MS Project (when asked) will show activities with zero slack in
red
– The critical path appears as a red line from start to finish of the project
• Look carefully on the critical path for high risk tasks
– Try to break these down to parallelise them wherever possible
– The best plan includes a contingency for all risks in the form of slack

Ellwood, ETC BVBA
Agenda – Part 2
• Introduction to the Scrum Framework
• Course Book 2
• Project Management Tricks and Tips
• Q & A

Ellwood, ETC BVBA
(and it’s application within Semiconductor
development)
The Scrum Framework

Ellwood, ETC BVBA
The Waterfall Model
• The “Normal” planning model of the last
two hundred years
• Project flows from requirements through
to acceptance
• Beware: Water does not flow well uphill!
– Notice feedback/iteration loops
– The bigger the step back, the more energy
is required
• Good for development
– Less efficient for research

Ellwood, ETC BVBA
The Agile (Scrum) Model
• The new (old) paradigm for project
management
• Break problem into small pieces
and iterate on them
• Beware: If you don’t choose the
waterfall model, you are using the
Agile model
– Software releases are free,
Tapeouts are not!
– We don’t have time to do it
twice, we only have time to do
it right.

Ellwood, ETC BVBA
Is Agile really that different from Waterfall?
• The short answer is no, but it does have a different Rhythm:
– In Waterfall the previous stage must be completed before the next
stage begins
– In Agile the project is broken down into several mini projects which
deliver functionality in layers
– Both require a “definition of done” to know when an activity is
complete
– Both aim to reduce risk by front end loading
• Waterfall does it by specifying in detail
• Agile does it by layering and doing the hard things first

Ellwood, ETC BVBA
The Agile Rhythm
• Scrum has roots as far back as 1993
• The Agile Manifesto came about in 2001
• Newer, cheaper technology made Agile
possible
Source: Grassy Fork Software 2014

Ellwood, ETC BVBA
Agile Terminology
• Agile terminology
– Epics
– User Stories
– Acceptance Criteria
– Tasks
– Product Backlogs
– Sprint Backlogs
• Some will say that Agile does not have requirements
– This is incorrect, Agile does have requirements
– Named differently, timed differently, organized differently

Ellwood, ETC BVBA
Waterfall-Scrum Relationship

Ellwood, ETC BVBA
Course Book 2

Ellwood, ETC BVBA
The Art of Doing Twice the Work in Half the Time
• The author is the Co-creator of
Scrum, an example of Agile
methodology

Ellwood, ETC BVBA
Why Agile is better (according to the book)?
• “The Waterfall model is broken because you cannot plan
something you don’t understand” (Chapter 1)
– “We can only give you an estimate after we have done it” versus “You
can’t schedule creativity .. but we must pretend we can”
– More realistically, it prevents you from falling down “rat holes” as soon
as the project starts
• Plan, Do, Inspect, Act instead of plan (=guess) and execute
– Understand that creativity is iterative, not linear
– To Act is to make decisions – to stay in control

Ellwood, ETC BVBA
Chapter 5: Strive for Flow!
• Choose the smoothest, most trouble free way to get things
done
• Multitasking makes you stupid
– Few people can easily pickup a task where they left off, they need to
get back up to speed
– Do one thing until its done before moving on to the next thing

Ellwood, ETC BVBA
• 90% of activities in a project will come
in roughly on time; the other 10% will
be more than 2x over budget.
• To finish on time we need to generate 10% contingency
within the 90% to allow for the overrun in the 10%
• We do this by seeking out and eliminating waste
The 90/10 Rule of Project Planning

Ellwood, ETC BVBA
Agile/Lean/Scrum : Cut out Waste
Learning from The TOYOTA Production System

Ellwood, ETC BVBA
Waste is a crime (Chapter 5)
• Half Done isn’t Done at all (page 94)
– A working schematic simulation at nominal PVT means nothing
• Do it right first time (page 97)
– When there is a problem, the team swarms around it to find the solution
• Working too hard makes more work (page 101)
– Mistakes made when working under pressure make tasks longer
• Don’t be Unreasonable (page 107)
– Challenging goals are motivators, Impossible goals are just depressing
– Enough with the Stupid Policies that slow you down
• Do not deviate from the QMS, do suggest improvements to the QMS
– Don’t be an Asshole
• A colleague is not there to be your friend, but not your enemy either

Ellwood, ETC BVBA
Scrum style Planning/Estimating
• Estimate relatively, not absolutely
– We are better at estimating relative complexity than absolute
complexity
– Use Velocity to turn relative estimates into man-hours
• Estimate based on the Fibonacci Sequence
– We are better at perceiving jumps than linear numbers
• Use Planning Poker to get better estimates
– Avoids “anchors” from throwing off the estimate
– Several minds make better estimates than one

Ellwood, ETC BVBA
Other little Gems from course book 2
• Fail Fast So you Can Fix Early (Chapter 1)
– Do the hardest blocks first
• Blame is stupid (Chapter 3)
– Don’t look for bad people, look for bad systems
• One Meeting a Day (Chapter 4)
– Catch issues before they become problems
– 15 minutes per day is enough to stay in control
• Secrecy is Poison (Chapter 7)
– Obfuscation only serves those who serve themselves
– Don’t shoot the messenger; bad news is better than no news

Ellwood, ETC BVBA
Final thoughts - Shi Ha Ri
– doing twice as much in half the time
First, learn the rules and the forms, and
once you have master them, make
innovations. Finally, in a heightened
state of mastery, discard the forms and
just be – with all the learning
internalised and decisions made almost
unconsciously (page 39)

Ellwood, ETC BVBA
Tricks and Tips

Ellwood, ETC BVBA
Teams don’t just happen – they must be
built!
• Having the entire team participate in planning helps start the
team-building process
• You can’t always choose teams that like each other
– Putting people together because they like each other will not be the
best way to achieve the project goal
– Teams with diverse psychometric profiles will perform better than
“clone” teams

Ellwood, ETC BVBA
Forming, Storming, Norming, Performing
• The leadership style changes during the team’s
development.
– In the Forming stage, it is directive
– In Storming, it requires influence
– At the Norming stage leadership
is participative
– In the Performing stage
it is delegative

Ellwood, ETC BVBA
The right stage at the right time
• At project kickoff (Forming), define the roles of each team
member
• Harness discontent (Storming) to identify risks early in the
Definition Phase
• Lead the team in transitioning into the design phase
(Norming)
• Use constructive feedback to keep project momentum
(Performing)

Ellwood, ETC BVBA
Project Control
• There is a tendency to sacrifice quality when deadlines are
difficult to meet
– Having well defined quality / performance objectives at the start
allows you to deviate from strength – knowingly making compromises
– This is what the Compliance Matrix is for
• Acceptable variances can be determined only through
experience
– Empathy requires understanding
– Try to be “bullshit proof”

Ellwood, ETC BVBA
Be ready to mediate when conflicts occur
• So-called personality conflicts are often just the result of poor
interpersonal skills
• Conflict is good if it raises a real project issue
– The trick is to deal with the issue, get buy in and move on

Fundamentals of Project Management

More Related Content

Similar to Fundamentals of Project Management (20)

Recently uploaded (20)

Fundamentals of Project Management