Summary: Many non-technical organisations are scared of IT projects. High failure rates. Technical jargon. Expensive consultancy. Rather than face this, they avoid IT projects completely. This hurts all of us - the organisation forgoes the value that a project might deliver, and developers forgo valuable work. Completion bonds (insurance policies that a project will complete on budget) are common in the film industry. I’d like to explore whether completion bonds might be a way to get over this “trust hurdle”, and how agile approaches might align to such bonds.
Rationale
Independent film producers often need to arrange “completion bonds” in order to obtain financing for their films. These are effectively insurance policies that projects will deliver on time and on budget. The financier pays a premium to an independent third party (the guarantor), who then monitors the project and ensures it stays on track. And if it doesn’t, then the guarantor repays the original financing.
The guarantor takes a very active role to manage risks affecting the project. This starts with the “specification”. Before giving any guarantee, the guarantor conducts a thorough review of the script, production schedules, budget, and so on. They meet with key members of the production team to confirm that appropriate direction and controls are in place, that the plans are feasible, and that the team has the necessary skills and experience to deliver them.
The guarantor then actively monitors status throughout the production. They receive daily progress reports showing what scenes have been shot, how many people have been on set, what materials have been used, and so on. They review details of expenditure against budget. They retain the right to visit the set at any time, unannounced, and to meet with anyone involved with the production.
If the production runs into difficulties, the guarantor may step in to take control of the troubling aspects. They may, for example, impose additional financial controls or seek changes to the script or personnel. In extremis, they may bring in their own producer to take control of the whole project. In the UK, the guarantor steps in at some level on about one in five films.
For the financier, a completion guarantee helps bound their investment risk and provides access to independent expertise. A completion guarantee may cost between 2% and 6% of the total production budget, but that’s a price many are prepared to pay for effective risk management.
What’s this got to do with agile development?
I think many organisations, especially small to medium businesses and public sector bodies, are failing to initiate valuable projects. They read statistics such as the Chaos Reports and get scared of the risks of project failure. They talk to IT companies and get confused by the technical jargon. They don’t trust the consultancy industry to help them navigate this maze. So they don’t initiate projects that might actually deliver a lot of value. Or if they do initiate projects, they bound them with such tight “risk mitigation” practices – heavyweight procurement processes, long contractual negotiations, burdensome project oversight structures – that an atmosphere of mistrust and bureaucracy is built in from the outset. This, perversely, increases the risk of failure even further, creating the conditions for an escalating “cycle of mistrust”.
This mistrust hurts all of us. Organisations are forgoing potential benefits. Developers are forgoing potential work. Or if we do get work, it’s set up for failure from the outset.
It’s easy to say “we should all trust each other more and this cycle would be broken”. It’s a lot harder to actually build trust. Agile approaches already do a lot to build trust. Working together day-to-day, regular visibility of progress, openness – all help build trust. But many of the organisations I work with have trouble getting over the initial trust hurdle to try out agile in the first place.
I think completion bonds could be a way to reduce the size of this trust hurdle. I also think agile development may create many of the conditions necessary for insurance companies to offer completion bonds – regular visibility of progress and constant checking that the client and development team remain aligned, for example. This session will explore what we might need to do to make it possible to offer completion bonds on agile projects.
The workshop is really about exploring a couple of questions: a) What would we need to do to make it possible to offer completion bonds on agile projects? b) Would such bonds help some non-technical organisations to be more confident to initiate projects?
I’d therefore like to use a facilitated brainstorming approach based on the Consensus Workshop and Focused Conversation methods (http://www.ica-uk.org.uk/facilitation/gfm.htm#consensus), which I’ve used with a number of clients. My proposed agenda for the workshop is therefore:
15 mins - Introduction to Completion Bonds
60 mins - Consensus Workshop on question (a)
15 mins - Wrap up and Focused Conversation on question (b)