There’s a lot of interesting talk and thinking, going on under the heading of #NoEstimates. Woody Zuill and Neil Killick are two of the most vocal proponents. The basic idea, as I understand it, is that it is possible to do small chunks of work incrementally, leading as rapidly as possible to a desired shippable product, and that when you do that there is no need to do much of anything in the way of estimating stories or the project.
This is a great idea. I agree with it entirely. I first heard of this notion early this century, in the work of Arlo Belshee. Joshua Kerievsky also spoke to this notion, about the same time.
There are many reasons to support this idea. Lean principles generally lead to a continuous one-piece model of production flow. Kanban approaches generally limit work in process, and for small teams the ideal limit seems to be one. Neither of these approaches has any need to estimate how long something will take. Instead, they measure how long things take, in “cycle time”, and use that to make such predictions as are necessary.
Chet Hendrickson and I have been supporting this notion for ages, starting somewhere around the time of my article on “running tested features”, back in 2004.
It’s a darn good idea. Working in continuous flow, without story estimates, is a great way to work. I think most everyone should try to get there.
Raising this banner again is great. I’m here to point to some issues with following the banner, and to invite its proponents to help resolve those issues.
We need help, becuase there are some very serious issues with the way estimates are used in software development. I discussed some of these in my February article in Pragmatic Programmer Magazine. Estimation is very difficult, perhaps impossible, and often misused.
Estimates are difficult. When requirements are vague — and it seems that they always are — then the best conceivable estimates would also be very vague. Accurate estimation becomes essentially impossible. Even with clear requirements — and it seems that they never are — it is still almost impossible to know how long something will take, because we’ve never done it before. If we had done it before, we’d just give it to you.
Estimation is difficult, perhaps impossible, and we would be well off if we could avoid it. That’s why the #NoEstimates idea is so compelling. That’s why I hope they’ll try to help people move forward.
Mind you, estimation, in certain forms, is not always impossible. It’s possible, under some circumstances, to do pretty well. In addition, sometimes people really do need to know about how much something is going to cost before going ahead.
These needs, and some ways to meet them, are talked about in my April article in Pragmatic Programmer Magazine.
To apply #NoEstimates in their lives, people who hear the idea need to fit the notion into what they already know. With enough motivation, they will figure that out for themselves. That’s why raising the banner is good. What would be better would be if #NoEstimates could offer more help.
And estimates are often misused. Estimates are often used as a bludgeon to try to get programmers to work faster. This leads to unhappy programmers and to poor software. No one wins.
In addition, a focus on estimation often leads to a fascination with completing some fixed scope of work. But the most effective way to do things is to manage scope very carefully, to get the best scope of work accomplished at every point in time. Estimation often militates strongly against doing this. The result is “Weak Agile” and inferior results.
Getting to #NoEstimates would fix this. That’s why it’s good that it’s being shouted from the ramparts. What would be better would be more help.
For all these reasons, #NoEstimates is a great way to go — when you can. I believe it is the best known way to work, and believe it is a point at which Agile teams who continue to improve will ultimately reach.
It is right and appropriate for the #NoEstimates people to be pushing this idea, as others before them have done. More power to them.
I do have concerns mapping #NoEstimates into what’s going on in projects right now. I’ll write them down here so that the #NoEstimates people will know what they are. They may choose to address these concerns, or to ignore them. Either way is OK.
The name #NoEstimates is extreme. That could be OK. It could even be good.
I really mean that. I’m an Extreme Programming guy. That name has Extreme written all over it. But that’s not all:
Early on, we used to say that no line of code should be written without a failing test that required that line in order to work. We didn’t say most lines of code should have tests, or that you really ought to think about having tests. We said NO LINE OF CODE WITHOUT A FAILING TEST.
We didn’t even know how to do that in every case. We were just saying never to be satisfied if you didn’t get there.
Similarly, teams should never be satisfied until they get to #NoEstimates. However, they will have some troubles and I hope that the #NoEstimates guys can help out.
What is meant by the word “estimates” anyway? Is “will be released around Christmas” an estimate in the #NoEstimates context? How about “Wow, I think it could take us a month to do that. Let’s try to break it down.” Is that an estimate? Should we therefore not do that thinking?
We need some help here in understanding what the #NoEstimates idea really is.
How to apply #NoEstimates isn’t entirely clear. Does it really mean that all estimates are bad? If not, which ones are OK? How can we tell the difference between an estimate that’s useful enough that we should do it, and one that is pernicious and never should be done?
People trying to get to #NoEstimates will need to work that out. I’d like the #NoEstimates guys to help us with it.
Does #NoEstimates always work? What if all our customer has so far is some big ideas? What if he doesn’t have much money and wants to decide whether to go ahead? What should a #NoEstimates proponent do in that situation?
What if our product needs to cover five areas, at least reasonably well, before we can reasonably release it? Can we get some sense of whether we’ll have them covered by October or whether it will take until next February? Can a #NoEstimates team help with this kind of concern?
Readers of this excellent #NoEstimates idea need help. A request for estimates from two or more levels up amounts to an absolute need. Even if we agree that getting to #NoEstimates is the right thing to do, how do we get there?
What are good ways to wean an organization away from estimates?
If we must estimate as we move along toward #NoEstimates, what are the best ways to do so? What are ways to provide information that management wants, with reasonable accuracy and in a way that leads them to discover that they can live without conventional estimates?
How can we avoid misuse of estimates? Are there ways of doing it safely, or of providing estimate-like information that cannot be used in a harmful way? If not, what are the best ways to protect ourselves, or our people, against misuse of the information we provide? And, again, how can we work out from under this kind of situation?
#NoEstimates is a great idea. It has been around for a while, and it’s a good thing that it’s getting a boost.
There are serious issues involved in moving from where people are to #NoEstimates. People need help.
It’s great that the #NoEstimates gang are raising the banner. If that were all they did, it would still be great, because other people will try, learn things, and write up what they learned. Progress will be made.
What would be even more great would be for the #NoEstimates gang to start digging more into issues like those raised here, because that way we will be able to progress faster.