≡ Menu
Pawel Brodzinski on Software Project Management

How Much Time You Add Value?

How Much Time You Add Value? post image

I was running a workshop on estimation recently. One of things we discussed was how much time people effectively spend on working on tasks which are assigned to them. How much time do they spend creating value?

Let’s take average software developers. What do they do? They code. But not all the time. They probably have some kind of warm up when they come to work. Then standup. And a couple of coffees. A meeting here and a meeting there. Handful of emails requiring attention or answer. A foosball match to break monotony of a day. A bunch of people asking about something either coming to a desk, or pinging through instant messengers. A couple of phone calls.

And don’t forget that each time after we lose focus we need time to recover. We need time to find out where we were and what exactly we were doing.

So again, how much time do we spend effectively creating value?

And another question: when we are estimating, what kind of effectiveness do we assume? Naive, but pretty common, approach is 100% even though we actually know it isn’t even close to such value.

Then, we are surprised that a project was underestimated. Again. What a shame. We didn’t see that coming.

Because we don’t learn. We don’t even try. Actually this is data we can gather pretty easily. Why don’t we do that then?

Yes, I pretty much expect the results would be surprising for many. And yes, I’m sure it would help you to make your estimates better. Not that I think that estimation itself is crucially needed, but that’s completely different story…

in: project management

7 comments… add one

  • Marcin Floryan January 10, 2012, 1:41 pm

    Very intriguing thoughts. Of course we never run at 100% utilisation and yet so often we base our plans on exactly that assumption. What’s really interesting, is that not only we would never be able to create value all the time, but we probably will never know up front how much of our time will effectively be devoted to value creating activities. We must also remember that it’s not even desirable to leave no spare capacity.
    As to planning itself, how much value does estimating bring, perhaps this post is a very good reminder that we have to reconsider if estimating has any value at all. I don’t think it does. At least not in most circumstances.

  • Pawel Brodzinski January 10, 2012, 1:50 pm

    @Marcin – Yes, we will never know exactly how much time we “waste” on non-value adding activities. However, knowing that historically it was, say, no more than one third or whatever is true in your case will definitely make you look differently at you capabilities.

    This is by the way something which comes to me every time when I discuss estimation: given that you measured a metric (pretty much any metric) in past projects, what makes you assume that in the next project you will do significantly better than you did in the past. The discussion gets more difficult when you have no historical data, thus this post.

  • Zsolt Fabok (@ZsoltFabok) January 10, 2012, 5:04 pm

    I tend to feel that the context you described is unique and observable only among software developers. For example factory workers have 5 minutes break in every hour and they work hard in the rest of the time. The same goes for people in the shops, in the public transport, at construction sites everywhere. It seems that the only place where people can do anything besides work and got paied is software companies ;-) developers are pampered unfortunately :-(

  • Pawel Brodzinski January 11, 2012, 1:21 am

    @Zsolt – It is true that software development industry is different than many others, but I wouldn’t say we are so special. Think of movie industry or professional sports – they are even more pampered than we are.

    However if you bring manual work as an example – in this case cost of context switching is lower. After all it is manual exercise, not knowledge work, so it is different.

  • Paul Klipp January 12, 2012, 3:44 am

    The way we did it in the early days (2004-2009) was to always estimate in ideal days and then track actual task cycle times and apply the multiplier to our estimates. We found that the multiple was very consistently in the neighborhood of 1.7. That way, we weren’t estimating how much time we spent adding value; we knew based on lots of past data.

  • Pawel Brodzinski January 12, 2012, 10:04 am

    @Paul – I would say that 1,7 is a pretty good result. I actually often propose similar approach when we discuss estimation – instead of trying to guess, just measure how you’re doing. Then not only will you be able to get pretty good result but also support it with facts.

  • Zsolt Fabok (@ZsoltFabok) January 14, 2012, 11:24 am

    @Pawel, I agree that it is hard to compare labour work with knowledge work, however from the employer’s perspective not doing something is kind of the same. Nevertheless, I don’t see a reasonable amount of time away from coding as waste. Programmers tend to see for example meetings as unnecessary interruption, but I believe that these small interruptions help them to think outside of the box, and come up with something different, maybe something better. I got my best ideas after project meetings, because I got away from my computer and had some to think and got some input from the other participants.

Leave a Comment