what is worse: over or underestimating?

 

I work in IT consulting, so everything I am going to share will be solely based on what I have experienced in the last 12 years in the industry. Saying that I do agree with the author.

We often start estimating without having much information in the hand. And asking the client for clarification might not be an option. Because different consulting firms are given the same bullet points in the name of requirements, which are really high levels, and most probably clients also do not know much about the solution.

They are expecting the consulting firm to propose the solution and quote a price that will be within budget. And if the competing firms do not have any prior relationship, then the price would be the primary differentiator.

And often in this kind of situations, conservative estimation is the way to go, so that the firm gets the project and get the foot inside the door. And another reason is, due to lack of information firms often underestimate.

After the project starts, the implementation team faces various challenges, including not so responsive to other teams. Environment unavailability, test data unavailability, worst of them all, the team did not get a good amount of time to learn the business and understand the requirement, and due to that, the quality gets bad.

If you notice, estimation or in this case specifically overoptimistic estimate the planning suffers, design suffers at times, whole the team suffers to meet the deadline, and meet the quality standard.

In my experience because of this optimistic estimation or underestimation whole team burns out and reputation suffers profitability suffers too.

Now, say we have an existing relationship with the client, and after delivering the first project, we have to now estimate for the second project. This time we know what needs to be done. But the problem is, we are too scared to think that this time we will be able to steer through smoothly especially because we have already experienced those challenges before.

The team starts overestimating because the team is pessimistic and anticipating the same problems they faced in the first project where they were overoptimistic and underestimated.

Overestimation has a major issue, especially when we ask for an increased dollar amount. Say we delivered one project for $100,000, now if we overestimate and ask for $150,000 from the client, the client will never agree, because, for a client, now we have the knowledge, and more importantly we did a similar kind of work for a less price previously.

So, in my experience, underestimating is worse, because the team burns out of working extra hours to meet targets, it becomes difficult to even go back to the actual estimate if we start with underestimating. The worst part, often the quality of the deliverable suffers due to lack of time , and as a result organizational reputation also suffers.

Now, overestimating might cause us not to get the project, but the question was would I choose overestimation or underestimation, and I chose overestimation. At least the resources won’t be burnt out, quality would not suffer or company reputation would not suffer.

 

Is there a way to combat the tendency to do these things based on our past, negative experiences or our need to be optimistic about the outcome?

We can create a center of excellence, and bring in people who have experience with similar projects for estimation. We can leverage their knowledge to come up with a better estimation.

Or, we can do what we do in an agile scrum. Everyone in the team votes on the complexity of a task and comes up with an estimation that how much a task would take. This actually helps with 2 things, past experience and while discussing in the group on specific functionality, people get a better understanding of the requirement, so that they can provide a better estimation. The benefit of scrum is flexibility as the team continuously receives feedback from stakeholders ( Angeles, 2013).

These are a few methods we use in our IT services sector. I am not sure if we can use the same exact methods in other industries.

 

Among the author’s proposed approaches to mitigate estimations inaccuracies, which one you consider the most effective or Ineffective? Explain why.

In my experience putting together a group of people who have experience in similar projects to estimate the effort works fine. I mentioned that we follow agile scrum, where we too, breakdown work in Epic to stories and then stories to tasks.

While discussing the understanding level goes up and the estimation becomes easier and more efficient. This approach reduces risk and paints a clearer picture of the complexity of the task (Grifffin, 2015).

I have found reaching out to the client to be mostly ineffective. Mostly because most of the times, the manager we interact from client-side, might not have much idea about what needs to be implemented. And, it is simply not possible to make someone available to clarify all our answers, so, unfortunately, we need to make a lot of assumptions based on past experiences.

References –

Griffin, M. (June,2015). The Art of Creating Accurate Estimates. A List Apart. Retrieved from: http://alistapart.com/column/creating-accurate-estimates

Angeles, S. ( August,2013). What is Agile Scrum Methodology? Retrieved from https://www.businessnewsdaily.com/4987-what-is-agile-scrum-methodology.html