No Estimates: A Brief Introduction and Practical Uses
Estimation has become engrained into our daily lives. How long will it take me to get to work? What should I budget for my grocery spend this month? Literally on a daily basis, we surmise outcomes virtually from thin air. This behavior has infiltrated the workplace and now has become a cornerstone of project planning without which we feel we could never dream to begin our work.
As it relates to project management and software development, clients predictably lead with a few standard inquiries: How long will this take? And how much is this going to cost me? If you’re not familiar with the No Estimates movement, you may feel like, yeah, these are reasonable questions that I should be able to answer with some degree of accuracy.
While we can understand, as humans programmed to think we need estimations to make wise decisions, that a customer would want to know these projections from the get go, humoring this need, in fact, works to the detriment of how we go about managing our projects.
In the 1950’s, the Japanese began a movement focused on Lean principles. They found that keeping an inventory or parts and pieces needed for manufacturing was a poorly designed plan. It can be difficult to grasp at first, but they soon realized that having their materials supplied “just in time” was a more efficient and cost-effective use of their resources. It allowed them to utilize warehouse space for revenue generating activities rather than a store, and it enabled them to cutout unnecessary upfront costs.
No Estimates is another type of Lean system. Deciding which activities are slowing you down or wasting time and money on the front or back end as they relate to estimating practices and eliminating the need for them altogether are at the core of this method.
Why we’ve become addicted to estimating
We’ve quite literally become addicted to the practice of estimating. It provides certain levels of comfort to feel as though you know what to expect. The problem is that estimations are little more than guesses providing us with a false sense of certainty. Having financial projections allows for budgeting but in the case of companies working on software development projects, the return should be the focus, not the investment.
Why estimates don’t work
Estimates simply don’t provide the benefits we’re programmed to believe they do. There are far too many uncontrollable variables in the workplace that prevent 100% smooth project progression. Software development projects require multiple teams with varied skills and skill levels. Not to mention, rarely are we ever working on simply one project at a time. This being said, the act of transitioning a project to the next team is hardly ever a perfect baton hand off.
People take vacations. They take leave for personal reasons, and they take ill. These factors simply cannot be foreseen and can completely blow your estimations out of the water. Not only can setbacks like this push back estimated deadlines but there’s a chance you’d need to bring in outside sources to finish the job hence creating unforeseen uplifts in your estimated budget. So much for estimations.
Estimating comes with other dangers such as living up to those estimations. This can go either way, but keep in mind that once goals are set people adjust their workflow expectations to meet that exact goal. If expectations are set high, you could end up with stellar project outcomes, but you run the risk of falling short. Set estimated goals too low and you’ve sold yourself short.
Don’t forget the most unpredictable factor of all: the human factor. Our estimating abilities are undeniably tied to our personal emotions. For instance, if given a project and told that if you muck up and fail to meet your projections you could risk losing your job, chances are you’ll err on the side of caution and provide conservative estimates. Again, you could be selling the project short instead of saving your job. On the other hand, someone in a sales position could very likely inflate their estimations to give themselves more room to make a buck.
All things considered, we’re coming to see how estimations, while we think we need them, are little more than comforting and restrictions.
How to work with clients without giving estimations
To be clear, No Estimates dictates that we eliminate as much estimation as possible. There are circumstances where estimating will be necessary, but relying on estimations too much is a waste of time and energy and directly conflicts with Lean business practices. This is where forecasting comes into play.
Forecasting and Estimating sound similar. The main difference is that forecasting involves the use of known factors to surmise an outcome. Much different than estimating which is providing simply the best guess. Forecasting can be useful in working with clients that require some upfront “estimations,” but be mindful that past performance doesn’t perfectly predict future outcomes. As we’ve said before, unforeseen factors arise with employees and teams in the project stream.
Managing stories in slices
Your client has provided you some stories at the onset of the project. Tackling each story in its entirety at once may seem like the correct route, but really, it’s not. Breaking down stories into manageable and actionable slices provides you a way to report progress on a more frequent basis. The more frequently progress reports are needed, the more slices you’ll need to make into each story. Discovering which stories are necessary for your project is also important on leaning out your process and reliance on estimations.
An example could be made from a standard IT ticket submission by an employee:
“I need to be able to submit issues to IT so that a particular problem can be fixed which prevents me from working.”
Simple enough, and possibly too small to slice up, right? Wrong. Breaking this request or story down into several slices helps determine the amount of work required to fix it. Picking apart this particular story, we can decode the request into two actionable items:
- Employee needs to be able to contact IT
- IT needs visibility into a queue of issues waiting to be worked on
We’ve successfully limited the scope of our solution by identifying two features to focus on that will inevitably solve the issue. Slicing stories into multiple actionable parts helps us to actively manage projects.
Slicing up stories also helps with time management, a piece which helps to eliminate estimations. Breaking up stories into chunks helps to separate quick tasks to those which more time-intensive. This particular form of “chunking” enables your team to implement temporary solutions in a live environment, therefore, mitigating risk. It may be a few more steps, but overall it helps with operational implementation.
Some stories need to remain intact. Such stories should be given a maximum calendar duration a couple of months, of course, depending on the overall size of the project. Anything broken out of these should be given a maximum duration of a day. Any user-specific story should not be granted any level of advanced priority beyond a one-week timeframe.
Breaking stories out based on this practice enables you to determine how many of each category can be completed in a given time frame and provides accurate forecasting capabilities.
Track historical data from the start
Providing progress updates is essential in keeping clients who have not embraced the No Estimates theory happy and secure in your ability to complete the project according to their expectations. If data is not collected from the onset of a project, it can be difficult to piece together a before scenario due to changes in technology or staff.
For example, keeping track of your story deliverables (in slices, remember!?) can allow you to extrapolate how many RTSs you’ll be able to produce over the life of the project. This type of forecasting (not estimating) can be valuable to your relationship with the customer and goes far in demonstrating project progression.
Making the transition to No Estimates
Changing our way of thinking and the manner in which we do things is difficult no matter the task. So, we should expect nothing less from the implementation of a new project management process. Baring this in mind, cutting out estimates on day one of the transition is not realistic nor is it expected. There are, however, several things you can do to pave the way to a lean, No Estimates, project management practice.
Create story points
Basing your projections on story points rather than days and monetary costs is the first step in the No Estimates change. The benefits of story points are important and much more helpful than providing nothing more than a generic timeline. Story points help you to discover dependencies and complexity of issues. Story based estimation allows you to break a project up into predictable milestones.
Eliminate estimating task timelines
Trying to gather data points on story progression is counterproductive especially when compared to the overall time estimate to completion. Often we’ll see a story broken down into sub-stories with estimated timelines only to find that when they’re added back together, they don’t meet with initial expectations. Anecdotal progress updates are more useful to your overall completion goals.
Limit time to complete stories
Any story that starts at the beginning of one iteration that isn’t completed by the end of it points out a failing in your organizational structure. By limiting the time stories are allowed to be floating out there without resolution helps to point out these shortfalls and further your No Estimates program development.
As mentioned before, keep track of the time it takes to complete different types of stories. This type of information can become invaluable for future project forecasting.
Realizing when you’re late and what to do
Just because we’re aiming to rid ourselves of estimated time restrictions doesn’t mean that we’ve exempted ourselves from being off schedule or coming in late. One of the primary reasons for delivering a project behind schedule is because the scope was unrealistic, and the only way to remedy the situation is to have a serious talk with your client about reducing the scope of the project.
No one ever wants to admit that they’re not able to finish exactly what they’d set out to do, but it’s important to understand that if you followed the No Estimates theory and worked on the most important elements of the project first, it is likely that the client would be able to do away with certain items in order to get the project wrapped up. Maintaining flexibility and organizing the project so that it’s simple to see that each component or story may not necessarily be reliant on the next helps one to recognize that it is possible to deliver on some but not all of the original story requests.
It is quite common for the scope of a project to grow while it is still ongoing. If you’re worried about the scope being inflexible, the mere fact that additions can be made on the fly will tell you that it surely is not. Adjusting down as you go is a stellar way to keep a handle on the overall scope of the project. As you begin to understand the stories provided and begin to weave the solution, you’ll find that some of the stories don’t need resolution after all.
In the end…
We’re addicted to estimating. Estimating is difficult to get away from and hard to make people understand why it’s not only useless but also detrimental to our success in accurately setting expectations and creating optimal workflow efficiencies. So, what do you do? Changing the way an organization thinks is a mountain to climb and realistically can be impossible. That’s the truth. But, there’s nothing stopping you from implementing No Estimates practices within your team. Demonstrating the benefits of change is much easier than asking someone to quit cold turkey. Incrementally implement the No Estimates methods into your project management style and see for yourself just how superior they are to traditional estimation systems.
If you are interested in learning more about #NoEstimates you can buy the book NoEstimates, just click here.