How Did You Arrive At That Estimate?

As anyone who has ever been involved in estimating a large, unique and complex piece of work will attest to, estimation can be difficult.  So often, change initiatives are delivered in a volatile and uncertain environment. With shifting sands, estimates have a high element of uncertainty. We shouldn’t be surprised at this—even seemingly simple and repeatable activities involve uncertainty. Imagine driving to the airport: If your drive to the airport normally takes 2 hours, how long do you allow when you’re driving your family there for a vacation? Well, if it involves a trip on a busy and congested highway where there are often accidents you might take a view that it’ll take between 1.5 – 3 hours, and you’ll likely plan on this basis, or at least take this into account.  If there is uncertainty in something as seemingly mundane as a car journey, is it any wonder that estimating effort on a complex project is difficult?

 

Yet the real danger is not in the uncertainty itself—that is a given—it is that this uncertainty becomes hidden or opaque, and projects proceed based on an estimate that has been arbitrarily plucked from the air. The uncertainty goes unacknowledged and so often projects act as if it isn’t there at all. A difference of expectations emerges:

 

Person A: “What is your ballpark estimate for this”

Person B: “Well… you’re putting me on the spot here…. but around 100 days… probably”

 

What happens next? So often 100 days becomes the figure, it is seen as a commitment. It is treated like a contract that is set in stone, and if it isn’t met then often the “blame game” starts.  “People will just have to work harder. It is their estimate” an aggressive manager might say.

 

Show Your Working

Perhaps in some cases these types of conflict emerge due to a difference in expectation over what is being estimated, and the level of uncertainty attached to that estimate.  Reflecting on the example above, Person B may be talking about effort rather than duration… they may also be basing this estimate on previous project experiences and there may be some unstated assumptions.  They may, in their mind really be thinking “It’s around 100 days, perhaps with a range of -25%/+200%, as we don’t know enough yet”.  Because this isn’t stated, the level of uncertainty stays hidden, and too often “100 days” is taken as an undisputed objective truth, even though that was never the intent.

 

A great question to ask when presented with an estimate is:

 

“Can you tell me how you arrived at that estimate?”

 

This generally leads to one of two responses. Some people will explain, in detail, how the figure is derived. They’ll explain which technique they used (perhaps based on previous experience, bottom up, top down, rough order of magnitude, three-point, or any other technique). They’ll describe the contingency ‘buffers’ that are worked in, will highlight assumptions and risks, and will give an approximate tolerance level.  This is useful, transparent, and allows challenge (“are you sure 10 days is long enough for that task? It feels a lot bigger than that to me, can we talk about it?”).  These types of estimates are still best guesses based on the information available. They can still be wrong, but this doesn’t mean they can’t be useful. They can be regularly updated as the horizon becomes clearer.

 

The second type of response is a blank stare, followed by deflection. If someone can’t describe how they arrived at an estimate, then this probably means it has a very wide tolerance indeed. Again, this doesn’t mean the estimate isn’t useful—early in projects very rough sizing can be insightful—but it is important that we know the limitations.

 

Estimates Have Error Bars

The reality is that all estimates have an inherent level of uncertainty associated with them, and if we are to use estimates it’s important that we find a way of showing that uncertainty.  We should talk about tolerances, ranges or certainty levels. If, for example, we provide a figure but state we have a lack of information so can only estimate to within an approximate order of magnitude (x10), a sponsor may take an informed choice to invest a small amount in some discovery work to better scope the situation.  We can then test, learn, update the estimate but more importantly find out more about the viability and feasibility of the project.

 

Whilst people don’t like hearing about uncertainty and risk, it is still crucial that we find ways of communicating it. What is often really needed in these situations is a cultural change that embraces uncertainty, experimentation and organisational agility. And as change-makers, we have a large part to play in cultivating this cultural change.

 

---

What are your views on the topics in this post? Do you have any tips, perspectives or anything to add?  I’d love to hear your thoughts.   Please add a comment below, and let’s keep the conversation flowing! 

If you’ve enjoyed this article don’t forget to subscribe.

---

About the author:

Adrian Reed is Principal Consultant at Blackmetric Business Solutions, an organisation that offers Business Analysis consulting and training solutions. Adrian is a keen advocate of the analysis profession, and is constantly looking for ways of promoting the value that good analysis can bring.

To find out more about the training and consulting services offered at Blackmetric, please visit www.blackmetric.com