5 ways to improve IT project cost management
3. Include and manage contingency
IT project managers are getting better at identifying and managing risks and issues. One area that is not often addressed is the need to fund response plans, prevention and/or mitigation. A risk allowance must be formed to fund the management of risks where tangible costs will be incurred.
While IT projects will always estimate and fund testing, they rarely allow for the associated rework. Since it is known bugs will occur in developed software code, hence the test, it should also be recognised that rewrite of code and retesting will be needed. Rework is an unknown but will almost certainly be needed against many scope elements, not just testing.
Projects need time contingency. It is interesting how many schedules do not include time contingency. This is fundamental for costing as the overhead costs (admin, project management etc) will be required longer. Construction projects have always allowed for inclement weather, so too should IT projects allow for unknowns.
An allowance for complete unknowns should also be made, e.g. error based rework, ‘in scope’ scope changes, etc. If you were to extend your home, you might allow an extra 10% for unknowns: so too should IT projects. Given the typical cost overrun of IT projects at 43%, it is logical that some level of contingency is required to provide decision makers with certainty.
The concept of ‘the contingency budget has been used up’ is totally illogical. Regardless of budgets, to get from the current state of the project to completion will always require contingency in your forecast.
4. Educate senior managers
One of the more difficult aspects of cost estimating and management is informing senior managers and/or steering committee members who don’t have a background in project based cost management. Finance people in particular focus on budget versus actual, particularly year-to-date, whereas a different mindset is needed for project budgeting and cost performance reporting. Terminology is particularly confusing, especially where some words are used out of context.
We recommend adopting the PMBOK terminology, reinforcing concepts of forecast versus budget. While the words and acronyms may be new for some of them, constant use of the correct terminology will reduce confusion.
Earned value sometimes comes up in discussions with senior management but is often misunderstood. Application of earned value is beyond some project managers, either due to knowledge or through systems and processes not supporting it.
For a simpler approach to cost performance reporting than earned value, try Googling ‘Gate Value Method’.
Project managers will need to be assertive, especially where contingency is concerned and where cost forecasts are extending beyond what is funded or what is palatable. Project managers should carefully maintain budgets and ensure forecasts are based on their best estimates, not be bullied into changing them.
5. Feed history into future estimates
One of the best estimating tools is history. The construction industry has learnt this with repositories such as Rawlinsons capturing and feeding back history into future estimates. The IT industry is only now starting to realise this, with the best example being the database being kept by the International Software Benchmarking Standards Group. We can only hope that the industry expands this into testing, design and other areas.
Meanwhile, IT PMOs have a huge role to play, collecting and feeding back historical data. Through post implementation review data and lessons learned, the PMOs can provide organisations with excellent metrics and ‘rule of thumb’ data for future projects.
PMOs can also provide an independent oversight of estimating data and ensure that contingency is included or clearly spelt out as being excluded from cost data, so decision makers are better informed and can fulfil their governance role.
