Jul 12

Relax. It’s just an hour of work

Just go with itHere’s one to think about and I would like your opinion or ideas on this one. I recently stumbled into an argument with a fellow coach. I told him that it’s not bad to direct or influence a team towards a certain baseline when it comes to estimating with poker. That it’s not bad to tell a team to pick the smallest story and give it a three and from there estimate the other stories. I usually help a scrummaster with this to just start a new team.

The ideas behind the three is something that Jeff Sutherland told me a long time ago during a training. He used a simple calculation.


The outcome is 2.33

The idea behind this was that if you take a day and then just take away a few hours and then also take away a little slack you get something that’s left namely the outcome of this calculation. 2.33 is near 3 and also near 2. But we just round it up to a big number in favor of the team. Ok keep this in mind.

The other thing is this. In the old day, and for some companies not doing anything Agile, a project manager approached a developer with something small and urgent. The “can you do this and how long will it take” question. In most cases a disturbance, a small fix. A developer would respond with the reply “I think an hour or so”. And we all know that the  ‘or so’ is the problem. A developer takes on a small job that will take an hour. But forgets to think about the fact that he has to drop the work that was running. Next the developer needs to find out what the work really is all about. He has to start doing it. Test it. Deploy it. Write documentation or something. Deliver it back. An hour work in most cases takes about four to five hours. If a working day has eight hours this takes a large part of the day. The rest of the time is lost in discussion, lunch, talks with colleagues and other important but non related work.

Are you getting where I’m going with this. The calculation presented by Sutherland wasn’t that bad. He thought that if you take the smallest job and assign it a three. This would be the same as giving it the fast one hour estimate. Knowing that it takes longer and almost most part of the day. So a three is almost a day work for one person.  Ok, you are probably starting to curse or getting uncomfortable with the idea that you can apply the scrum estimation on exact time. I know, I have the same feeling. But just stay with me a little longer.

There is a little truth in this way of thinking. It Is very dangerous to tell a team that a three is a day’s work. Before you know they start estimating with this and with time. Very dangerous so don’t say that. Just pose the fact, that they can pick the smallest item and assign it a three. That’s it. No more explaining and move one. But secretly and somewhere in the back of your head you know that you have helped them to establish a sort of baseline. You did what they have done so many times. You have asked them to quickly estimate a small work item that is much bigger than they anticipated to begin with but at the same time provided all the extra slack an time they need to get it done within a certain time limit.

The fellow coach believes that you must not do this. Just let the team do their thing. If they want to give the smallest item a eight or one or whatever. They can because this is agile thinking. You should not intervene. Ok, but we are not hippies who just let thing go their way. We try to help deliver a product. Sometimes products in large corporations where “Just do as you fee” is not getting you there. A little metrics and calculation is needed. Mind you this does not mean anything else. The team still picks the work they can deliver following their forecast, but you help them on the right track when it comes to estimating. And at the same time root their baseline based on simple human planning, or better. The lack that we humans have when it comes to planning. Scrum is not flower power stuff but seriously delivering a product and there is nothing wrong with a few certainties, even if those are based upon human factors and interactions.

But hey, it’s just an idea. Prove me right or wrong 🙂

Permanente koppeling naar dit artikel: http://agilethings.nl/its-just-an-hour-of-work/