Sunday, March 21, 2010

Value from the Scrum course

I wanted to reiterate views I have expressed elsewhere, I think.

My best guess is that a typical development team of 7 or 8, including the product owner and the scrummaster, costs about $1 million per year. Including all related costs. (In my opinion, every team should know their total annual cost.)

My best guess is that a typical development team can produce about $3 million per year in Net Present Value. Based on what they develop; ie, their share of the "earnings" from the product. This is discounted over the 3-5 year typical time horizon. (In my opinion, every team should be given an estimate, and some data after the fact, of the value of the work they will be doing, say, over a year. This concentrates their minds wonderfully.)

From a Business Value viewpoint, the goal of the Scrum course is to significantly increase the productivity of each team. So, let's assume the whole team comes to the Scrum course. And some related managers. Let's assume the team can double their productivity (their velocity) within one year. Let's assume in years 2 and 3, they continue to improve. So, saying we go from a NPV of $3 million per year to $6 million per year is not a stretch.

Let's say that there are other costs/factors that also contribute to the doubling (coaches, impediments removals, etc). So that the Scrum course, which teaches 3 teams, only gets 25% of the added value. 25% of $9 million is: $2.25 million. (Does the scrum course deserve more or less than 25%? You judge.)

Now, get out your spreadsheet and calculate the value for yourself, with your own assumptions. See XLS file here. Let me add that there are many other benefits, but let's ignore those for now.

By the way, a decent team (not a great team) in Jeff Sutherland's opinion should be able to double velocity in 6 months. Many do in 6 weeks. And some companies are seeing an improvement of 5x-8x for every team and it happens quickly.

A couple more things to say:
  1. The purpose of the course is not to convey explicit knowledge, although that happens. The key purpose is to get the people willing, wanting, and waiting to change. "It's the Tacit Knowledge, stupid." (To play with a political quote from the past century.) Without this change, which always happens somewhere else than the brain, no worthy improvement will take place.
  2. The team must be having fun. I won't explain that more here.
  3. The team must be working reasonable hours (40 or less in total).
  4. Thus, the way to velocity improvement is by impediment removal.
  5. And thus, managers must be removing some impediments (and letting the team self-organize) or the full amount of improvement won't happen.
  6. Finally, this is still not a silver bullet. Teams can be dysfunctional and projects can still be impossible. (The good news is that these serious problems can be identified very much sooner.)
Now, our challenge to you. Have you achieved a measurable and believable increase in your own baseline velocity? How much?

You owe it to yourself, your teammates, your company, your customers, and right now to the world economy, to get your team going. You can do it. If you feel you can't change your organization, first, don't believe yourself. "The culture" can resist one or two people. It cannot resist a fired up group that is right.

Still, if you can't change your organization, then you must change your organization. And then double there. You will be proud of yourself when you do it. You know you can. We know you can; we cheer you on.

Doubling is not enough, but like 12 lawyers at the bottom of the ocean, it's a good start.


Jeff Sutherland said...

It is very easy for new ScrumMasters who take my Certified Training to double the velocity of their teams in a few months. Many of the venture companies I could triple their velocity in three two-week sprints.

A typical team costs $100000 a month in the U.S., 100000 Euro a month in Europe. So they start saving at least 50000 Euro a month within a couple of months. So a team take a couple of months to get organized and spends the next ten months at double the velocity (a very conservative achievement). That's 500000 savings the first year.

Let's say they spend 2000 Euro to take the course. Then that is a 2500 percent return on investment the first year. And this is only cost savings. If you factor in the sale of twice as much software you should get a good multiplier.

I always advise never invest any money in Scrum unless you plan to get over 1000 percent return the first year. If you plan on it you will get it.

Yusuf Arslan said...

The most important added value (of implementing Scrum) in my situation is a common terminology that everybody can use. A simple word such as "velocity" with a clear and simple definition makes it possible to discuss without confusion. After our Scrum training (by Jeff) everybody is forced to speak with the same common language and think with a common sense.

While it is still a challenge to remove impediments to really get hyper productive, we have shortened the delivery time with a half.

It is really true that serious problems can be identified very much sooner. (It is still difficult to find solutions.)

Jeff's advice to aim for 1000 percent ROI is really valuable.

Joe Little said...

Hi Yusuf,

Yes, I agree completely that Scrum (and/or XP and others) bring additional values than the one I mentioned. Common terms is an important one.

Jeff's comment in part reminds me of Henry Ford's famous quote: "Whether you think you can or you can't, you're right." We are not saying that 6-8 people can change everything, but we are saying that we know (virtually all the time) 6-8 people can get some important changes to happen around here. Wherever their "here" is.