Over the past year, my series of articles hasconsidered how to improve enterprise projectmanagement capability. Soon I will be startinga new, shorter series of articles on matchingthe project manager’s leadership style andcompetence to the type of project forincreased project success. However, before finishing this series, I amgoing to discuss audits, health checks, reviewsand benchmarking. These are important toolsin learning and knowledge management,helping to identify weaknesses in currentcapability and thus possible areas ofimprovement. In this article I will be describing:l definitions of audits, health checks, reviews

and benchmarkingl the application of these tools to assessing

the capability of the project-basedorganization

l the application of these tools to performance assessment of the individualproject

l conducting a project management audit.

DefinitionsReviews, health checks, and audits areessentially variants of similar things.

ReviewsA review is a normal check of project progressor project performance conducted accordingto a predetermined schedule. Reviews may becalendar driven or event driven.

Calendar-driven reviews: These areconducted as part of the normal control cycle,(Turner, 1999). Current project performancedata is gathered, analyzed and converted intoreports:l for the project manager and project team to

control progressl for senior management, particularly the

sponsor, as discussed in an earlier article.

Event-driven reviews: These are normally heldat the achievement of a project milestone, orstage gate. Stage-gate reviews are now verycommon, held at completion of project stages.Figure 1 illustrates a gateway review process,from my book on managing Web projects(Turner, 2003). At the end of each stage of theproject a review is conducted to determine if thework of the previous stage has been properlyconducted and if it is still worthwhileproceeding with the project. In the early stagesof a project, stage-gate reviews will be go/no go

decisions, part of the benefitsmanagement process. The businesscase will be reviewed in the light ofinformation generated during thelast stage of the process todetermine if it is still worthwhileproceeding with the project. In laterstages, reviews may be go/go backdecisions. Has the work of the laststage of the project beenadequately done to enableprogression to the next stage, or isfurther work necessary?In my book (Turner, 1999), I suggestthat stage-gate reviews areessential for a project to progressthrough the early stages, and for ensuring thata project is financially viable. When you start aproject, the information available can be veryinaccurate. You think the project is profitable,and at the best case it is highly profitable, butat the worse case it might be unprofitable.However, based on the information available,you decide to proceed (ready to discover), andcommit a small amount of resource to thefeasibility study. Through the feasibility study,you generate more information, and reduce theuncertainty of your estimates. It is still notenough to proceed, but enough to commit alarger number of resources to the designprocess (ready to design). Through the designprocess you gather enough information toreduce uncertainty, enough to take the finalgo/no go decision, and proceed toimplementation, (ready to develop). I once worked with an organization where thenext stage gate (ready to deploy) on its productdevelopment projects was also a go/no godecision. They asked themselves if the marketwas still there to make it worthwhile releasingthe product. However, on most projects, onceyou proceed to implementation, you commityour self irreversibly to proceed to completion;so you want to know that the decision is madeusing accurate data, not available at initialconcept.One particular event-driven review is the post-completion review. Here you review overallprogress on the project, see how you did, anddo two things:l record performance data in your history

files to help estimate future projectsl see if there are any lessons on what you did

well and what you did badly to help thedevelopment of your overall projectmanagement capability.

In the articles on learning and knowledgemanagement, I discussed the problem ofdeferral of learning, putting it off until the end.Learning about how you did should be built intoall the stage-gate reviews.

Health checksHealth checks are ad hoc reviews, conductedby the project team, to do a one-off assessmentof the health of a project or of the projectcontext. In my books (Turner, 1999; Turner andSimister, 2000), I suggest you may conduct twotypes of health check:Project health check: This is a check of theperformance of a project: The team step backfrom the project for half a day, or a day, just tocheck that it has been properly set up, that allthe relevant controls are in place and are beingproperly used, so that the project is progressingto a likely successful outcome. It is very easy,once you are in the thick of a project, not to thinkabout whether it has been properly established.You are so busy working that you can’t thinkabout whether it is being done properly. And itseems impossible to take time to step back fromthe coalface for half a day, or a day, to considerhow you are doing. But it is better to spend halfa day to discover a mistake that might costweeks or months, than to rush headlong only todiscover the mistake when it is too late. Projecthealth checks are not the same things as stage-gate reviews. Stage-gate reviews are go/no-go/go-back decisions based on an assessmentof the business plan, using the data currentlyavailable to the project. Health checks shouldbe conducted about a quarter of the way intothe design and execution stages, to check thatthey have been properly established and thatthey are progressing to a likely successfulconclusion.

Building enterprise project management capability 13Audits, health checks, reviews and benchmarkingby Professor J. Rodney Turner

22 PROJECT MANAGER TODAY JULY 2006

Project management capability health checks:These are an assessment of the projectmanagement capability of the organization. Ihave termed these ‘ Projectivity health checks’.I will describe them in the next section, thoughthey have now largely been superseded bybenchmarking and the application of maturitymodels.

AuditsAudits are essentially the same as healthchecks, except that they are conducted byexternal, independent auditors. It is useful forthe project team to step back and considerhow they are doing, but because they are closeto the problem, they can still miss a seriousmistake. If a project is highly critical to anorganization, of high strategic value, and asignificant commitment of (financial)resources, the parent organization may want toensure it has been properly set up and so willinvite independent auditors to take a look.However, it is essential for the auditors to adopta supportive approach, ie, they are there tohelp the project team achieve a successfuloutcome, rather than an attitude that they arethere to check up on the project team. Thelatter approach can induce considerableresistance from the project team, and areluctance to help. If, however, the auditors aresupportive and work with the project team in aspirit of partnership, then the audit process canbe very valuable.Some post-completion audits are conducted tofind out why a failed project failed socatastrophically. Then the auditors’ role is tocheck up on the project team and they maymeet with resistance. But if the project teamhave failed badly, they may be feeling guiltyand so may offer little resistance. I haveconducted post-completion audits wheremembers of the project team try to blame eachother, or senior management, and havecollaborated with me in trying to blame eachother. Often they were unaware of their owncontribution to the failure, and in collaboratingcondemned themselves as well.

BenchmarkingBenchmarking is essentially different fromreviews, audits and health checks. Itsemphasis is on determining how the project isbeing managed in comparison to theorganization’s standards of best practice,whether it is financially viable, and to learnfrom successes and failures to improve theorganization’s standards. The emphasis ofbenchmarking is to compare the organization’sstandards and its project performance toindustry current best practice. Benchmarkingmay be quantitative or qualitative.Quantitative: In quantitative benchmarking, anorganization compares its project performanceagainst other projects from the industry. Itdoesn’t compare against individual projects,but against the mean and standard deviation ofa collection of projects. Qualitative: In qualitative benchmarking, theorganization compares its standardprocedures against industry standards. This isclosely linked to project management maturity.Maturity models, in effect, providequestionnaires for qualitative benchmarking.

Assessing the capability of theproject-based organizationI will be discussing two things in this section:1. Projectivity health checks2. Qualitative benchmarking or maturity.

Projectivity health checksThrough a projectivity health check, anorganization will attempt to assess whether itsproject management procedures meet bestpractice. Through a series of questions thatforce the organization to assess itsprocedures, it will assess whether they meetits needs for project effectiveness. In my books(Turner, 1999, Turner and Simister, 2000), I givepossible questionnaires. There is not space toreproduce them here but I can describe theircoverage and application.In an earlier article, I said an organizationshould be competent at three things:1. its application of the project life cycle,

including go/no-go/go-back decisions atstage-gate reviews

2. its application of the project managementlife cycle

3. its management of the project managementfunctions or knowledge areas.

So, to assess its project managementcapability, an organization should develop a setof questions (or use mine), which ask:Project life cyclel How well do we initiate projects?l How well do we make the decision to

proceed to feasibility?l How well do we conduct feasibility studies?l How well do we make the decision to

proceed to design?l How well do we plan and design projects?l How well do we conduct investment

appraisal and take the decision to proceedto execution?

l How well do we plan, execute and control projects?

l How well do we take the decision to proceed to closure?

l How well do we finish the work, commission the asset and achieve thebenefit?

l How well do we conduct post-completion reviews?

Project management life cyclel How well do we plan projects?l How well do we assign roles and

responsibilities for undertaking projects?l How well do we assign resources to

projects?l How well do we control progress on

projects?Tools and techniquesl Do we use appropriate tools and

techniques on our projects?l Do we tailor our procedures appropriately

to the size and type of projects?l How well do we manage scope?l How well do we manage project

organization?l How well do we manage cost?

PROJECT MANAGER TODAY JULY 2006 23

l How well do we manage time?l How well do we manage quality?l How well do we manage risk?l How well do we manage communication?l How well do we manage project

procurement?l How well do we manage human resources?l How well do we manage anything else that

is important?

What I suggest is that you ask people fromdifferent departments to complete thequestionnaires and look for areas where:l they agree that you are weakl they have differences of opinion about

whether you are weak or strong.

In conducting a health check you areassessing how well you think you are doingagainst what you think you should be doing.

BenchmarkingIn benchmarking you are trying to assess yourperformance against industry best practice.Huemann (2004) suggests that there are foursteps in benchmarking:1. Understand the detail of your own

processes2. Analyze the processes of others3. Compare your performance with that of the

others4. Close any gaps identified.

As I have said, project management maturitymodels provide a ready set of questions to helpyou both analyze your own processes andthose of others. However, many benchmarkingcommunities also exist. The Human Systemsnetwork operates benchmarking communitiesin Europe, North America, Australia and HongKong, and will shortly open a network in China. In comparing your performance, it is commonto prepare a spider-web diagram (Figures 2 and3). First you plot your own performance, Figure2, against parameters that you considerimportant on a scale of, say, 1 to 6. Then youplot the performance of others, Figure 3. Thegaps will be obvious. Figure 3 may also represent the standard youneed to achieve to reach the next level ofmaturity. One slight problem is that the maturitymodels require you to achieve a definedstandard against all the parameters,regardless of your own needs as anorganization. Against one or two parametersyou may not need to achieve the standarddefined, and against one or two others you mayfeel you need to achieve a higher standard. So,conducting your own benchmarking exerciseenables you to tailor the standards to your own

needs. Against that, the maturity modelsprovide a ready-made set of questions.

Assessing the performance of anindividual projectI will discuss two things in this section:1. conducting health checks on individual

projects2. qualitative benchmarking of individual

project performance.

Health checks on individual projectsTo conduct a health check on an individualproject, I suggest you develop a questionnaireto help you investigate key points.Questionnaires are suggested in my book,(Turner, 1999; Turner and Simister, 2000). Thereis not space to repeat them here, but they maycover issues like:l Are the project objectives understood and

accepted by all?l Are the success criteria and key

performance indicators understood andaccepted by all?

l Are the success factors understood and accepted by all?

l Have appropriate tools and techniques been adopted to help achieve the successfactors and success criteria?

l Has the project been well planned?l Are the tools and techniques being properly

applied?l Is the project being well controlled, with the

right progress data being gathered andanalyzed in the right way, and useful reportsproduced to control progress?

Again, the questionnaire should be completedby all project team members, and other people,including the users and other stakeholders. Youshould focus on:l where team members disagree about the

answersl where they all agree that the project is not

being well managed.

Quantitative benchmarkingIt is sometimes possible to benchmark projectperformance against industry databases. Forinstance, the European Construction Instituteand the Construction Industry Institutemaintain a benchmarking database. Memberscan enter project performance data forindividual projects and compare theirperformance to all the projects currently in thedatabase. This does not give members direct comparisonof their projects to those of their competitors,but they are able to see how they areperforming in comparison to industry norms.

Members are then able to improve theirperformance in areas where they are weak.

Conducting a project management auditAs I said earlier, an audit is conducted, onbehalf of the project owner, by external peopleto help ensure that a project has been set upand is being managed in a way to ensuresuccess. In my book I suggest a seven-stepprocess for conducting an audit:1. Conduct interviews2. Analyze data3. Sample management reports4. Compare against standards of best practice5. Repeat steps 1 to 4 as necessary6. Identify strengths and weaknesses7. Define opportunities for improvement.

Huemann (2004) suggests a seven-stepprocess, based on more recent experience:1. Analyze the situation2. Plan the audit3. Prepare for the audit4. Conduct the audit5. Generate the report6. Present the results to the project owner7. Terminate the audit.

She suggests that step four, the audit itself,consists of four sub-processes:4.1 Analyze documentation, including control

data and control reports4.2 Conduct interviews with the project team as

necessary4.3 Observe the project team in action4.4 Ask the project team to assess their own

performance, perhaps against a healthcheck.

Audits and emotionsAudits can be highly emotionally charged. Theproject team can feel embattled. It is importantfor the auditors to work with the project team,to help them see that they are there to help theproject team successfully achieve theirobjectives.

ReferencesHuemann, M., ‘ Improving quality in projects and programmes’,in Morris, P.W.G and Pinto, J.K., (eds), The Wiley Guide toManaging Projects, Wiley, New York, ISBN: 0-471-23302-1.Turner, J.R., 1999, The Handbook of Project BasedManagement, 2nd edition, McGraw-Hill, London, ISBN: 0-07-709161-2.Turner, J.R., 2004, Managing Web Projects: The Managementof Large Projects and Programmes for Web-space Delivery,Gower, Aldershot, ISBN: 0-566-08567-4.Turner, J.R. and Simister, S.J., (eds), 2000, The GowerHandbook of Project Management, 3rd edition, Gower,Aldershot, ISBN: 0-566-08138-5.

Rodney Turner is Professor of Project Management at the Lille Graduate School of Management, and chief executive of EuroProjex: the European Centre forProject Excellence, a network of trainers and consultants in project management. He is the author or editor of nine books. Past chairman of the APM, he hasalso helped to establish the Benelux Region of the European Construction Institute as foundation Operations Director. Rodney received PMI’s 2004 ResearchAchievement Award at the Global Congress in Prague in April 2004.E-mail: rodneyturner@europrojex.co.ukThis article was first published in Chinese, in Project Management Technology, published China Machine Press, Beijing.

24 PROJECT MANAGER TODAY JULY 2006