In Learning

Building Bridges

Bridge building projects typically deliver on-time, but software development often doesn’t deliver on-time and on-budget. This little insight garnered from the Standish Group CHAOS report formed the early stages of the Emergn Value, Flow, Quality Research team’s work into investigating on-time delivery. There are three main reasons put forward by the paper on why this might be the case:

  • The design specification of bridges is fixed.
  • Little flexibility is offered in specifications.
  • Bridges have been built for 3000 years (with public investigations into their failure, compared to a I.T. industry who typically covers up, ignores and even rationalises failure).

As a result, it argues that I.T. keeps making the same mistakes over and over again.

But bridges aren’t software – one is a physical product with a relatively clear definition of success, and the other is an often intangible product with a very unclear definition of success. So time to find an industry that might be more like software development.

‘This is not the industry you are looking for’

Maybe the film industry? A collaborative creative intellectual industry based on knowledge workers, it could benefit greatly from feedback, starts with initial requirements (often specified on paper), has large scaled projects and I heard a rumour that some of the biggest films of all time, Star Wars and Titanic were delivered late and over budget. Excellent, maybe we will discover another industry with a similar performance to I.T. However, what we found was not what we expected…

Oh No! Most films are delivered on-time and on-budget.

Now the transparency of the film industry is almost as bad as the I.T. industry, there is lots of financial jiggerypokery, often designed to cover up mistakes, and reduce tax implications. However, we did discover something very different to software. Project Insurance!!!

The film industry often employ the services of companies like Film Finances to provide insurance to deal with cost overruns. Since the 1950’s, to cover the risk of projects overrunning, the insurance industry has stepped up with services designed to protect against those risks.

So why doesn’t this exist in the I.T. industry. Why can’t big business and government buy insurance protection?

We still don’t know the answer other than hazard a few guesses.

  • Maybe software is more complex?
  • Maybe films have more repeatability – they key costs and time constraints are understood.
  • Maybe it’s easier to measure the progress of film production and its relationships to cost.
  • Maybe the benefits are clearer, though insurance benefits from the control of costs, not typically the final revenues.

The strange thing, each of those things would reduce the need for insurance. Only one other thought came to us, maybe insurance companies could not make a profit, IT just fails too often. Like many, we feel it is foolish to make precise schedule and budget predictions in I.T. We encourage people to embrace uncertainty and gain many of the benefits of focusing on faster delivery by lowering lead and cycle times. Yet, the film industry is doing something different. It may be damaging its outcomes as a result of the insurance protection, but it is interesting to see how they deal with the risk.

Sign up to receive our newsletter and new posts

Recent Posts
Showing 3 comments
  • Philip Black

    It might not be that they fail too often, but I wonder if it is the fact that blockbuster films have the ability to make millions and billions at the box office.  The upside is massive for a success.  Therefore the bet (insurance) is worth the risk.  The tangible benefit from Enterprise or Government IT projects doesn’t seem to be quite that tangible.

  • Graham Oakes

    Good question.

    There are some interesting differences between IT & films.  e.g. film crews have well-defined roles, whereas IT relies more on “teams” where the roles are more fluid (esp in self-organising teams).  Stuff Dave Snowden is saying lately about crews versus teams is interesting in this regard…It’s worth looking at how the film industry makes completion bonds (i.e. insurance) work.  e.g. the insurer has the right to examine any detail of the shoot at any time, and to step in with their own crew if they don’t like the way things are going.  (Having well-defined, industry-accepted roles helps make this possible.)  And, as Philip suggests, if the insurer does need to pay out, then they can recoup their costs from the film’s takings.

    Laurent Bossavit raised this question a few years ago and I looked into it in some detail as I was writing my book on project assurance.  Chris Matts has aso looked at it.  I think there could be a viable business model for insuring IT projects, but you’d need to tune the model very carefully & reject a lot of projects as being fundamentally uninsurable.

    (But knowing that a project is uninsurable would be extremely valuable information to many organisations trying to initiate the project.  Just like knowing the cost of insuring your teenage son to drive a sports car tells you a lot about whether you should actually let him drive a sports car.)

    • Paul Dolman-Darrall

      Thanks Graham

      That is really useful information, Laurent sent me his blog from 2003.

Leave a Comment

The BADD Formula - A Business Case for Agile