14

Estimation games

We assume a lot when we make estimations, and everyone thinks differently, so what one person defines and estimates can be very different. The exercises in this section are great to do with your team, your manager or your customers to help you gain a shared understanding of and consensus over the metrics in order to define what is involved in the activity and what the definition of success looks like.

These estimations provide valuable metrics for measuring progress and how accurate our estimations are by comparing them to the actuals. Many tend to under estimate the time it will take to complete a task. Honestly, I think if I admitted how long some things actually might take, I probably would not start them. It is, perhaps, a type of optimism of getting it done quickly and easily, and a tendency to think optimum rather than actual.

There is an old adage in the technology sector that says, ‘Estimate and then double it’. Today, some of the largest software companies in the world ask their developers to estimate how much work they can achieve in a given timescale and then they halve this workload, as this is the realistic actual work they will achieve. In the world of software development, there is a high rate of uncertainty and change. If you work in a high-growth business or a sector that is impacted by technology (that is just about all of them, by the way), then this extreme rationalisation of your estimates may be necessary.

By recording these metrics, you can measure how your estimates compare to your actuals and learn to adjust your workload accordingly in future sprints of work. You will also learn when your estimates are likely to be more or less accurate, given the nature of work.

Business as usual will tend to vary less, whereas new activities and change projects will include a degree of uncertainty, but you can be aware of how much you tend to over- or under estimate when estimating the unknown and adapt your rankings accordingly.

Scope creep

Scope creep is a term used to describe the addition of unplanned extra work, once a project has been started. There is only so much that can be defined and planned upfront: as an activity progresses, we learn more about it and so our decisions may change; we might identify that there are elements that were not taken account of at the beginning; or that, due to a change, additional work is needed in order to meet new needs resulting from the change. However, scope creep does not come only from changing our minds along the way; a primary reason for scope creep comes from underestimating how long an activity will take, how much it will cost, how much thinking time may be needed, and additional actions such as rework or hidden costs taken into account that had not been included in initial budgets and estimations.

We can see what work we are achieving and the work that we have planned versus what we did that was unplanned and unpredicted. It is possible to refine forecasts by reviewing previous metrics based on work that we have already completed, and so predict our productivity and manage our performance better in the future. For example, if we become aware that on average, 20 per cent of our work is unplanned, we can build this in as slack in the future in order to take that into account and provide us with the time we need to respond to these ad hoc work requests.

By talking through activities with your team, your manager, your customers and others, you can help to improve communication and gain a shared understanding of how the work breaks down, what is involved, how long that might take and, most importantly, its value.

This is especially valuable if you are working as part of a team and for communicating with your managers and customers, which invariably we are. It enables you to gain better clarity and expectation of what is involved in each activity, as well as a shared understanding of the metrics and priority of each activity.

Estimating helps to give a better understanding of the scope of the work to be undertaken and the impacts and metrics involved. Analysing estimations helps to gain a better understanding of workload and ensure there is understanding across a team about the nature, scale, value and difficulty of jobs, which means there is shared consensus about the work to be carried out and what is likely to be involved.

Working as part of a team and organisation means that each of us is dependent on others both internally and externally. Often, we receive work from one person and then, like a relay race, the baton is then passed to the next. Knowing more about these related activities helps to understand and improve upon working relationships. By estimating the size and time needed to deliver, we can establish what is involved, which will inform the decision-making process when work is being selected and prioritised in the future.

Estimation games

Estimation games are a great game to play with others, especially if there is a need to justify the time that an activity may take and clarify what is involved. The game helps everyone involved to reach a consensus and shared understanding of the work involved and its expectations. This game is derived from the original estimation and planning software-based game Planning Poker™, developed by Mountain Goat Software (used in SCRUM and XP Software Development and first defined by James Grenning in 2002 and trademarked by Mike Cohn of Mountain Goat Software), and has been adapted and evolved for more general use.

The practical outcome of the game is to agree on how the task will be carried out and how long this is expected to take.

• As a group, agree a scale of estimation that you are going to use – for example, one point might equal one minute, one hour or one day, depending on the general size of activities involved.
• As a group, select an activity that has been captured on a sticky note and have a few minutes’ open discussion.
  • What elements does this activity include?
  • Who will carry out the activity?
  • How is it going to be approached?
  • Are there other dependencies?
  • When will the activity be done and fit for purpose?
• Privately, each member of the group writes down clearly their estimation of how long this will take on a new sticky note or a piece of paper. Once everyone has written their estimate, all players show their estimate at the same time.
• Each player then takes it in turns to explain their estimation, what it includes and/or does not include and their reasons for the points they have allocated. The group discusses the variation and the reasons and agrees on a better definition of that task.

Reasons for differences in estimations:

• Different people doing the task, due to different levels of experience, can help to see who is best to do the job, based on how long it would take them, although there is also capacity to consider.
• Different ways of completing the activity.
• Definition of success, or the definition of ‘done’ for an activity is different – perfection versus good enough.
• Different assumptions of the elements of the process. For example, some may not feel that pre-work or post-work is included and estimate only the actual task.
• Waiting time gets included – the task is estimated in terms of how much time will pass rather than how much time it will take. If there is a waiting time within a task, it is a good indicator that that task needs to be broken down into multiple tasks that have dependencies and can be actioned when needed, rather than staying in an ‘in progress’ or ‘waiting’ status for extended lengths of time.

If estimations match, then this estimation is deemed to be the estimated time that the activity will take and it is agreed, as consensus has been reached and the next item is selected.

If the estimates vary, the variations are discussed and then a second and third estimation game can be played, if needed, in order to reach a level of understanding where consensus can be reached.

If difference continues, and an average is not agreeable to all in the group, ‘park’ the activity for estimation later in the meeting, as a review of other items may aid the decision. Alternatively, it may be agreed that further information is needed before a considered estimation can be made and its size is set.

As an activity is progressed and completed, actual time spent can be captured and the total time spent compared with the estimate when reviewing in order to see if the estimations made were good, or if they had been under- or overestimated significantly. The reason for this can then be established and acted upon in the future.

Estimation games are a great way to help the team gain consensus over what is involved in activities and how long they will take to complete. It helps everyone to gain a better understanding of the level of work that is required and gain a broader understanding of the scope of the work.

Packs of playing cards are available online which have numbers for estimation on them: for example, series of numbers such as 0, 1, 2, 4, 8, 15, 40, 80, 100+. These packs often include other cards, such as ?, for when a question needs to be asked before an accurate estimation can be made, or a */joker to challenge the activity. The 100 + card represents a task that is too big to be estimated accurately and needs to be broken down further.

Cup of tea

Here is a simple example to demonstrate how estimation can vary greatly. It is great to play with others as an initial estimation when first introducing the game.

I have used this example time and time again with teams and workshop groups and, every time, the outcome surprises its players and helps them reach a new level of awareness and understanding about how we all estimate and assume differently.

Using a points scale of 1–100, where 1 point equals 1 minute, ask the group to estimate the task, ‘Make a cup of tea’. Ask them to write the number of points they estimate and, when everyone is ready, reveal estimations at the same time.

The results likely will vary from 1 to 15, and there will be a number of reasons why:

• 1 – Make a cup of tea for myself, assume I am in the office and have an instant boiling water tap and all ingredients to hand.
• 4 – Make a cup of tea for everyone, assuming at I am at our current location, knowing that I have an instant boiling water tap so it will not take long.
• 6 – This is how long it takes to make and drink a cup of tea.
• 8 – Make a cup for everyone and assume that not just tea is wanted, so it might take a while.
• 15 – Make a cup of tea for myself, check messages, have a chat in the corridor.

Notice how each estimation has differing assumptions about what the task involves and whether resources are available. Some add additional tasks they feel are part of that process, and some expect change and therefore account for that time. Some assume automatically that it is for them, others include everyone in the group. Almost always estimation is made based on how long it would take them to complete the task.

This simple exercise highlights the realisation that we all think differently and, if estimating making a cup of tea can vary so greatly, then the jobs we do in our work and our understanding of what is involved are also likely to be varied.