The Domino Effect and Traffic Management

(A combination of two work posts - about throughput and constraints)

1) the Domino effect of project failure

“Given the rule of thumb that one in ten projects will go badly wrong, and given a program of 20-25 projects, you will most likely have at least two that overrun seriously and will be canned. But the damage won't stop there. Those two projects will then cause the two below them in priority to be canned. Now there are four projects out of two dozen that won't happen.” The domino effect can be taken even further: “In certain instances those projects that have gone wrong are on the critical path for other projects - and they push those back as well.”

from Best Practices for Small Projects

Interesting.... and of course this doesn't apply only to small projects.

Wherever you have a constraint... time, resource, money... one project going over (or under) will impact others. In most cases a project being "under" time, resource, or cost represents an opportunity. And that opportunity will also have a ripple effect.

But where the project goes over on a constraint, then the impact is like throwing a stone in a pond... the ripples spread a very long way.

It is human nature to try and carry on with all the existing commitments as if there is no issue. But the result is that resources get spread more and more thinly, so more

2) which led me on to: Traffic Management

I've been trying to work out how to illustrate that there is a point where more projects just mean you reduce what you are delivering. Here is my first attempt.

Let's assume you have a 4 lane motorway - this represents your total capacity for work and change.

Now I have 3 convoys of vehicles. For ease of identification let's make each vehicle in a particular convoy the same colour - so I have a blue convoy, a red convoy, and a green convoy.

Because I have a 4 lane road, all 3 convoys can be en route at the same time. All my convoys can move at their maximum speed without interfering with each other. And in fact there is an empty lane so I could get one or more convoys to their destination even sooner. If a vehicle in a lane breaksdown, it does not interfere with any of the other convoys. And I could use the extra lane to make up some of the lost time. My arrival simply depends on the slowest speed in each convoy and the length of the convoy. As the last vehicle in a lane clears the start point, I can start another convoy on the road.

Now I add an extra convoy (white). This is still fine as each convoy can use a lane of their own. If there is a breakdown, there will be less chance to make up time as I will need to wait for one of the other convoys to clear the road.

Add a further convoy (black). Hmmm - now I start having to interleave vehicles on the 4 lanes as there are fewer lanes than convoys. This is going to slow up the last vehicle in each convoy getting to the end point. The planning is much more complex as I need to understand the speed of each vehicle in order to work out how best to schedule things into the lanes. There is limited room for manouvre and a breakdown is going to impact multiple convoys. It becomes much harder to tell you when any single convoy is going to reach the end point because it depends so much on what other traffic is in its lane.

If I keep adding in convoys to my road, there will come a point where the smallest slowing down in a single vehicle will trigger gridlock.

Of course for projects it is a more complicated situation in that the size of my "motorway" is not simply a number of lanes, the vehicles are all different sizes and may also require different types of road to run on.

But there is a still a point at which adding more projects in results in delivering less not more.