Topic 3.5: Estimating cost and time
If you under-estimate the time and resources required to deliver a task or project, then (as per the triple constraints) you may have to sacrifice either scope or quality to get the project delivered.
This may be difficult to explain to stakeholders, especially if you have a fixed budget and/or delivery date.
Conversely, if you over-estimate the time or costs required, you penalise the organisation by locking away valuable resources that could be spent on other projects or operations.
Even though it might be your best intention to return unspent requirements, Parkinson’s Law suggests that you will still find a way to spend them!
Alternately, at the initiation stage, your project may be rejected on the grounds that it will take too long or be too expensive.
If you are tendering competitively for project work (even within your organisation), this means your over-estimate could cost you the job.
Deliberately over-estimating time or costs is a technique colloquially called padding, and we will discuss the strategic risks of this later in the course.
For now, though, at the business case level, we would like to be as accurate as possible in estimating the time and costs required to deliver each of our three, four or five identified options.
So how accurate do we need to be?
It is widely held that a 20 percent margin of error is acceptable at this stage; however, your organisation may have its own policy in this regard.
Plus or minus 20% means that we will tolerate our business case cost estimates being 20% ‘wrong’ either way, but any more than that is considered unreliable for decision-making.
Now 20% seems like a big margin of error; however, keep in mind that all we are doing at the business case level is comparing options so as to arrive at a preferred project solution.
When we take this option to project planning – the next stage in the project’s life – we will revisit our cost estimates and refine them to within 10%.
However this is a much more detailed (that is, expensive to undertake) process, and more effort than it is worth for business case decision-making, as we shall see.
Note, though, that 20% margin of error is a refinement of the ±30-50% we were happy to accept at the opportunity definition / project concept canvas stage.
So what are some of the common mistakes people make when estimating, and how can we manage these risks?
Sources of error
Unclear scope / missed activities
The more detailed your scope of work, the easier it is to estimate time and resource requirements with precision. Conversely, a scope that is poorly defined or patchy reduces the quality of our estimates
Poor project record-keeping
As one very reliable source of time and cost data, past projects are usually our first point of call when estimating future work. This assumes, though, the data exists and is accurate.
Asking the wrong people
Don’t confuse people’s confidence in their own estimates with their actual ability – the two do not always align. Make sure you are consulting and appropriate weighting the estimates of experts.
Quality estimates take time to achieve. When rushed we are often forced to make assumptions in the absence of data. Each assumption will exponentially increase your margin of error.
We often assume best-case scenarios when making estimates, overlooking risk factors such as the natural tendency of people to have unproductive, conflicted or down-time, the costs of scrap, rework and spoilage, and the inevitable price fluctuations of goods and services. This is why real data based on past experience should be more valued than crystal-ball gazing.
Padding is the process of deliberately over-inflating estimates to manage unforeseen risk. Beyond the over-estimation risks cited above, we talk about the damaging cultural impact of padding later in this course.
Suppliers will also sometimes deliberately low-ball estimates (under-quote) to win work, charging for expensive variances later in the project when the cost of changing the supplier is seen as too prohibitive. Beware a price that is too good to be true!
Having a clearly defined and agreed approach to estimation ensures the consistency and reliability of estimation inputs and process (such as acceptable margins of error). This is especially important when you are relying on multiple stakeholders in this regard.
Use multiple sources
Broad stakeholder engagement is critical to estimation success. If you consult three stakeholders and get huge price or time estimation range, consult further and in more detail to reduce that confidence interval.
If you don’t define the scope of what you are estimating, assumptions will fill the gap. Each assumption will have its own margin of error that compounds on the others. If in doubt, dig deeper and ask.
Even the simple task of Googling estimates demands a good understanding of what is needed and how it integrates into the overall project design.
Allow normal (not best case) conditions
This is where real world data on past project performance is critical. Go beyond past project plans (after all, a plan is a collection of estimates), and look at how long things actually took and how much they actually cost.
Use consistent units
Especially in international projects, make sure everyone is talking in a common currency and agreed weights and measures (for example, miles or kilometres).
Assume task independence
When considering a whole series of activities together in one estimate, you can easily underestimate the nuances of specific tasks within the set, and try to force-fit individual activities into their higher-level estimate. You will see this is especially important when we get into project planning.
Estimating is more than just a task; it is a process that involves:
Agreeing a methodology
Clearly defining the scope of what you are estimating
Assigning responsibility to the right people
Appropriately engaging stakeholders
Validating and documenting the estimate, then
Consolidating unit estimates into a whole.
Don’t skimp on this activity – you will (literally) pay for it later!
Although optimism can play havoc with estimates, it isn’t necessarily a bad thing.
The core principles of optimism bias are outlined in the following TED Talk by Tali Sharot, a cognitive neuroscientist at University College London.
We will return to bias later in the course; for now though, think about the implications of optimism generally for the management of stakeholders and projects.