The Standish Group in their Chaos Report (www.standishgroup.com) suggest that projects over $10 million have almost a zero success rate. That may be overly pessimistic, however large projects do have unique problems that require serious senior management attention.
Over that last seven years I have been closely involved with a number of such projects which were successful in terms of customer or owner satisfaction criteria. They did not all meet original time and cost expectations, but it is very difficult to initially correctly estimate such projects both due to their size and complexity, and due to when these original expectations were set.
In reality there are three phases for estimating time and cost as shown in the following table.
Phase Type of estimates
Conceptual: what are our broad needs? Wide rangeFunctional: what do we want to do? Budgetary with large contingencyDesign: how will we do it? Firm with normal contingencyUnfortunately, figures provided at the early conceptual stages get set in concrete. Organisational norms often do not allow for any contingency in estimates -- either assuming that estimating is a precise science or that "real" managers don't need such crutches. Realistic and experienced mangers provide for contingency but manage it carefully.
For the purpose of this article, large and complex project are characterised by all or several of the following:
¥ $10 million plus cost.
¥ timescales from inception to delivery greater than two years.
¥ large teams of more than 50.
¥ numerous stakeholders.
¥ external suppliers who are critical to success.
¥ complex integration of a number of different components.
Heads in the Sand. Some senior managers are surprised when large projects don't work as planned. The literature and war stories abound in tales of failure, particularly in IT projects, but managers continually assume that "it won't happen to us". Large projects are even more complex than running a small organisation since they exist in the context of a larger organisation that sometimes gets in their way. Also they are about unique outcomes leading to higher levels of risk.
The following six key lessons have been extracted from my personal experience or knowledge of four such projects, three in the $15-30 million range and one over $100 million. Each project had its own unique characteristics but the lessons are still the same.
1. Get the right team in place at the startThis covers all the skills required including project management and sponsorship. It is preferable to over-resource during start-up when planning is heavy since you will lose key staff along the way when it is very difficult to replace them.
By carefully assessing the experience required of the project manager based on size and risk of the project, you can avoid the frequent changes of project managers experienced in many projects. Look for a project manager who anticipates problems rather than reacting to them.
It's unlikely that a single person can provide all the skills required and a coalition of skills can be assembled if people are prepared to cooperate. There will still be an overall project manager but the sponsor, project manager, project administrator resources and team leaders can form a powerful collection of complementary skills.
Be cautious of the "heroic project manager" who believes they alone can rescue the project. They are often opinionated and inflexible and do not get the best from the team since theirs is the only way to achieve results. They may have a short-term positive effect but longer term will most likely destroy goodwill with stakeholders and motivation by staff. On the other hand consider carefully what they say since they may be raising issues which the current management team is ignoring.
The project's identity and culture is important in retaining commitment by the team since staff will work on the project over an extended period. Senior management have a key role in being available to foster and support this.
2. Don't underestimate the overheads
Additional overheads will arise from the extra management required to plan and control the broad range of activities, to manage the interest of diverse stakeholders, and to deal with the inevitable "politics" which larger projects attract.
It will also come from the need for evolutionary development through delivering in smaller chunks. This evolutionary strategy is necessary to reduce the risk, meet aggressive timetables with a partial delivery, and to provide stakeholders with comfort about progress to keep them interested. The duration and resources to set up and plan for a large project are considerable and you should allow for these overheads in setting initial expectations.
3. A positive relationship with suppliers is critical All the research on this topic stresses the need for trust but how many relationships with suppliers do you know which exhibit this characteristic? Ultimately it largely depends upon individuals you deal with so multi-level contacts with suppliers can help mitigate the constant change of personnel. Too often customers look for a competent supplier project manager but they are hard to find and often you end up with someone with less than the required experience level.
The supplier must also engender confidence by providing visibility of sound project management processes they are actually using to manage their work, and not the ones in their methodology/proposal that only get used when all else fails. Even with large multinationals claiming expertise in project management, real experience is patchy.
In contracts with suppliers, organisations could insist on provision of a specified level of project management expertise and the right to approve the project manager put forward, but this may also diminish the supplier's level of responsibility. The contract should specify in detail the type of planning and reporting expected from the supplier to avoid debate later on. Payment based on results is an important principle in retaining interest and commitment from supplier management.
4. The business must own the project
This means the overall project manager works for the business irrespective of where they come from in the organisation. There may be other project managers for components of the project (for example, technical and supplier). The business must commit sufficient time and resources to solving problems and ensuring that the changes implied by the implementation are closely managed. Also they will clear obstacles to allow the team to focus on their work and ensure the project is not caught up in other organisation initiatives and misses its own deadlines.
John Roberts of the GartnerGroup in an article in The Australian notes that effective business sponsors keeps the focus on "why are we doing this", ensuring the appropriate business and technical resources are kept on the project, managing resistance to change and being accountable for delivering the business benefits. Easier to say than do, but certainly critical to success. If adequately implemented it will greatly help to reduce the old problems of scope change/volatility.
5. A stable and sound project organisation is importantSmall teams can be managed informally but larger projects need more formal structures which include:
¥ clear lines of responsibility and authority ¥ defined communications channels to ensure information is effectively conveyed to all relevant parties¥ formal management processes (for example, disciplined detailed planning, change control, issues management and so on)These structures will inevitably change during the project but not too often -- crisis management is a strong indicator of lack of control and a precursor of doom.
Harold Kerzner in his book In Search of Excellence in Project Management -- Successful Practices in High Performing Organisations, John Wiley & Sons, 1998 talks of "informal project management" but notes that for this to exist we still need formal processes in place which we selectively and consciously bypass plus a culture of cooperation, trust, teamwork and effective communication. Such a culture will require considerable work to achieve but will be supportive of large projects.
6. Effective and sufficient controls to understand the real status There are two levels of controls required -- project and corporate level. Corporate level controls will ensure there are appropriate expectations about how projects are managed and resourced, high level reporting of status against time and cost, and how the business benefits are being realised, and that the necessary project controls are in place and used.
Management should consider what information they need to monitor the health of the project and to trigger actions if needed. It's a fine balance judging when to respond but it is common that management fails to act in a timely manner.
The Denver International Airport debacle is a salutary lesson. They lost over $1 million per day for over a year due to late delivery of a baggage handling system which had been signalled to management as having problems for a long time in advance. Some organisations mandate "gating" reviews for all or large projects to mitigate such risks.
Project level controls will be about detailed tracking against schedule and cost, plus monitoring risk and quality. Reporting of status needs to be holistic including risk and quality, not only schedule and costs. Metrics which complement written reports are a necessary reality check. A method for forecasting future schedule and costs needs to be agreed and used for senior management to have any comfort about accuracy of status reports.
In the US, Earned Value Accounting is popular since it links schedule and costs together and provides a more rigorous basis for reasonably accurate forecasting of progress.
Tom DeMarco in an address to the Project Leadership Conference in Chicago in June this year proposed "serious" risk management, which is in contrast to "can do" management that relies on a high degree of good luck to achieve the schedule. A realistic assessment of risk is required and its impact on the probability of achieving the published schedules. Many scheduled delivery dates at best have only 50 per cent or less chance of being achieved.
Courage Under Fire. Since all the above techniques are well known why are they not more commonly applied? In fact, many people believe they are doing all these things, but aren't, or think they do not have time to do it "properly".
Kerzner in his book says, "we never have time to do it right the first time, but always have time to do it over again when it fails". Tokenism abounds in projects -- if we do a bit of everything then it will somehow work when what is required is rigour in key processes.
The key is not to "lose your nerve". This does not imply keep persevering when it is hopeless to do so, but expect problems to occur especially in the early stages. False starts are common. Look for the real underlying reasons behind problems and not the symptoms, and act with courage.
If you don't succeed be flexible to try alternative solutions, however make sure you get independent advice on the available options to avoid the "sunk cost syndrome" which results in more money being spent on an already doomed project.
Harold Ainsworth has a background in large and complex business and IT systems projects and currently consults and teaches in project management. He is a principal of Project Managers Network Pty Ltd (www.pmnetwork.com.au) and can be reached at firstname.lastname@example.org
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.