Estimate work in five simple steps with kanban! Are you planning to start using this project management tool? The task may appear much harder than what you’ve expected. Follow the tips below to learn how to estimate work in 5 simple steps with kanban!
Static or detailed estimates are usually not needed for new features when working with Kanban. If your team struggles with estimating, it’s probably time for a change. A lot of teams have been using traditional methods such as story points or t-shirt sizing. These can be adapted into more clear and understandable user stories that just describe the functionality without having any specific units of measure. In user story mapping, for example , each card represents a user story. In the image below, you can see how user stories are described as a comparison between a minimum and an ideal outcome.
In Kanban, estimation is done by comparing the current state of work with the desired outcome . When estimating in Kanban, it should be expressed as how much closer a certain task will bring a team to its outcome compared to other tasks on your board. Use your best judgment and avoid placing estimates on each individual task if there’s no clear difference between them at first sight. You can even use different colour cards for prioritizing tasks so that their progress will be more visible from outside of Kanban. To estimate card size you can use the Fibonacci sequence: 1, 2 ,3, 5, 8, 13, 21 … which you can read more about here .
Ideal outcome.It is important to know that there are no perfect numbers in Kanban for estimating tasks. The point is not how close your estimate will be to the final result but whether it shows progress towards the outcome compared to other tasks on your board. We recommend trying micromanaging first without any measurements of metrics and only then start micromanaging by introducing methods like t-shirt sizing instead of exact values or time tracking etc.”
“Kanban estimation done through comparison to an ideal outcome acts as a measure of momentum. Kanban estimates are made by measuring distances between where work currently is and where one hopes it could go. The object isn’t perfection. The object is making progress. Estimates are made by answering the question, “Where do we have to get so that all stakeholders are happy? If your team can’t answer this question yet, don’t worry about it right now.
You see, at first I wanted to share with you one of my favorite articles that explains how estimation works in kanban boards. You can find an excellent presentation here . It’s all very logical and clear. But then I started reading comments under the article and suddenly realized that not everyone reads about kanban estimation without having some problems with it. That’s why I decided to write a post devoted specifically to this topic. Let’s go!
First of all, I’d like to say that kanban estimation is something that doesn’t require any complicated math. It’s not about some complex calculation with standard deviation or weighted average . But still some people have difficulties understanding it. They find it hard to give a reasonable answer when estimators ask them for WIP limits based on historical data. If you belong to these people more likely than not, don’t worry too much about it yet again – try estimating your tasks in one column first and see how it goes (the method described below). After a few iterations, you’ll naturally get better at it and will be able to estimate accordingly even for new tasks in several columns simultaneously.
However, if you feel confident enough in your estimation skills and you want to make them even better, read on.
Some time ago I was asked by a reader of my blog to share the method described below. I said it would be easier to model as an example. We’ll use a simple task with two steps – this is a very typical setup for Kanban style tasks (if you’re not familiar with kanban, I suggest reading David Anderson’s book).
The task card. Note how we can trace both column widths from WIP Limit column into Task Estimates one (the red arrow): Estimate each step width separately This is actually the simplest thing that works pretty well for most people who don’t feel comfortable with estimating as a whole (and this is actually pretty common as we’ll see in a moment). In the picture above, I’ve drawn dotted lines from one column into the other to separate these two estimation tasks.
The right hand side of those arrows shows how widths are traced into Task Estimates or into WIP Limits if you like. If Moving Averages (explained below) are used, you can also find one dotted line going from Task Estimates to WIP Limit and another one going from the Estimated At column downwards.