Depends on:1. Staff experience
2. Project novelty
3. Consequences of estimating errors
Project estimation is a complicated process, and for many types of projects
it is not done with much accuracy. Doing it well involves the contributors
who will be responsible for the work and the project leader,
and may also benefit from inputs of experts, sponsors, and stakeholders.
Regardless of exactly who gets involved, precision and consistency
rely on good process and metrics,
Establishing Objectives
=======================
Before any detailed, activity-level estimating can be done, there will be
high-level estimates for the project as a whole established by the project
sponsor and stakeholders involved in initiation. If the assumed cost and
duration are not consistent with the expected benefits and value, no
project will ever exist. These estimates may be based on some analysis,
or they may be wild guesses pulled out of the air; whatever their basis,
these initial estimates should be treated only as general goals. Although
they may also represent some known constraints, the initial top-down
estimates can be used only as provisional starting points. Setting a realistic
project baseline requires thorough planning, analysis, and detailed
estimates developed by the team responsible for the work.
Involving Activity Owners
Project planning relies on good requirements definition and scoping,
involving the stakeholders. The resultingscope
for the project provides the basis for a work breakdown structure,
which at its lowest level defines the tasks that must be done to
complete the project. Each of these tasks will need both a duration and
a cost (or effort) estimate, and the best source for the initial bottom-up
estimates will be the assigned activity owner plus any other contributors
involved.
The first consideration for estimating will be the method chosen
to do the work. Effort and duration estimates depend strongly on the
approach and process to be used for the work, so once this is clear,
bottom-up estimates are possible. Other considerations for the initial
estimates include potential risks and any remaining unknowns, such as
any contributors who will be involved but are not yet assigned.
Involving Others
=============
For work where there is ample precedent and a lot of experience, the
estimates provided by those who will do the work may be more than
adequate. For work that is novel or where you lack experience, you may
want to get outside assistance. Involving peers in your own organization
or networking with contacts in professional societies may provide useful
information. Even getting quotations or proposals from outside service
providers (whether you are considering outsourcing the work or
not) can provide potentially useful insight. Again, additional considerations
will include risks and unknowns, such as just how comparable the
experiences and metrics developed elsewhere will be to your particular
situation.
Adding Your Inputs
===============
The final arbiter of project estimates needs to be you, as the project
leader, because ultimately it’s your project. In examining the initial estimates,
be skeptical of any that appear either too conservative or too
optimistic. Ask questions about the basis for the cost and duration estimates,
and probe to uncover exactly how the work is to be done. If you
are not convinced that the people providing the estimates know enough
about the work, ask more questions. Especially if the estimates seem
too small, challenge the owner to break the work down further (whether
you intend to track at that level or not).
Compare the duration and cost estimates, and test if they are consistent
given what you know about how the work will be staffed. If they
are not consistent, provide guidance for adjusting the cost estimate, the
duration estimate, or both. If staffing commitments for all of the work
are not in place, collect range estimates that will be refined later when
the skill sets of the contributors are known. If there are known risks
associated with the work, include an appropriate margin for reserve (to
be managed by you for the project as a whole, not to ‘‘pad’’ activity
estimates).
Finally, compare the bottom-up detailed estimates of cost to the topdown
objectives for project expense to see how much trouble you are
in. Consider options that might make sense for bringing costs more into
line with expectations if the variance is significant. Use the duration estimates
to build a preliminary schedule and test it against your project’s
timing goals. Again, if there are substantial differences, revisit options
that might realistically reduce durations. In all cases
where your estimates exceed your initial constraints, document the
information that you will need to negotiate a realistic baseline with your
sponsor before formally committing to a fixed project objective.
Regards
Kshitij yelkar
www.yelkar.com
