…and what you can do about it
Why things take longer than you
think…
— Full service digital agency
— Established 1997
— 62 minds in London & Bristol
— Award winning
Clients
Why do things take longer
than we think?
Theme 1: Estimates don’t make projects
late
Estimating is relatively easy,
controlling projects is hard
Before
— crystal clear scope commitment
— estimate for this scope
— clarity of project/company/client objectives
During
— chase down risks
— chase down unknowns
Project Controls
How good at estimating are you?
Put a range which gives you a 90%
chance of including the right answer
Typical results
— With Black Bullets
Source: Steve McConnell
Theme 2: The Cone of Uncertainty
Time
Unce
rtain
ty o
f est
imate
4 x
0.25 x
1 x
Source: Steve McConnell
Time
Unce
rtain
ty o
f est
imate
1.5 x
0.67 x
1 x
Source: Steve McConnell
You can be more lucky
You can’t be more accurate
The cone does not narrow itself
You narrow the cone by...
removing unknowns making decisions
designing
Time
Vari
abili
ty in t
he e
stim
ate
4 x
0.25 x
1 x
Initial Concept
Requirements
Wireframes
Visual Design
Development
Completion
When estimating, consider where you are in the Cone
Re-estimate at key points
Time
Vari
abili
ty in t
he e
stim
ate
4 x
0.25 x
1 x
Initial Concept
Requirements
Wireframes
Visual Design
Development
CompletionZone of Hope –
20%
Cape of Fear
Cliff of Despair
Theme 3: Estimates and Commitments
Your estimate is wrong
Estimates inform commitments
Estimates Inform Commitments
Commitment
Estimate
Commercial
Appetite
Capacity
Fame
If you commit before the Cone has narrowed, consider your
strategy
Time
Estimating techniques
Fingers and thumbs
Fingers and Thumbs
Team Captain
— Read out each country
— Count 1,2,3
— Everyone shows their
estimate
— Discuss and agree
Relative Estimates
Relativity
— Arrange animal cards in order of weight
— Set reference as Antelope = 3 (not 3kg)
— Put number cards next to each animals
Affinity Estimating
Affinity
— Lay the number cards in order
— Set reference as Convertible =3
One at a time each person either…
— Takes and place a new card
— Moves existing card, explaining why (must move a card if
you don’t agree).
Theme 4: Planning Fallacy
Wishful Thinking
— People focus on the most optimistic scenario for a task
— Only affects things we estimate ourselves
— Anecdotally people miss predictions but make deadlines
— Anonymous predictions less affected
— Youth/inexperience
— Young industry
Countering Wishful Thinking
— Be aware of it
— Do best case then worst case estimates
— Get second opinions
— Treat variations in estimates as a positive insight
And when we carry out work
— Give yourself milestone and deadlines
Theme 5: The Three Cs
CountComputeCompare
Count
— visible early on
— easy to
understand
— about 20 things
Compute
— do the sums Compare
— against previous
projects
Historical data for project profiles
Phase % of typical budget Budget this projectKick Off 1% £800Phase 1 8% £6,400Phase 2 12% £9,600Phase 3 15% £12,000Phase 4 22% £17,600Phase 5 30% £24,000Phase 6 12% £9,600Total 100% £80,000
Budget for this project £80,000
Theme 5: To over or under-estimate?
The project might never start
(with your company)
If you over-estimate
Parkinson’s Law‘work expands to fill the time available’
If you over-estimate
If you under-estimate
Increased time and effort
— Emergency status meetings
— Dealing with the client
— Re-estimating
— For additional deployments
for phases
— Briefing and managing extra
resource
also…
— Weakens your position with
client
— Delays other related
projects
— Delays other client’s
projects
Extr
a eff
ort /
cos
t / s
ched
ule
100% Over-estimatedUnder-estimated
Brook’s Law‘adding manpower to a late software project makes it
later’
Finally…
Tips for what you can do about it
1. Good estimates without project controls are worthless
2. Count, compute, compare (Judgement is last resort)
3. Create project profiles based on historical data to compare against
4. Treat ‘high’ estimates as a priceless warning of trouble ahead. Do
something about it.
5. Be aware of your position on the Cone of Uncertainty
6. Beware the large cost of under-estimating
7. Be clear on what scope goes with the estimate
8. Re-estimate at key points and track progress
9. Beware planning fallacy
10.Give yourself deadlines
Nairn Robertson | @nairnski
nairnrobertson.com
Software Estimation:
Demystifying the Black Art,
Steve McConnell