A brisk whisk of risk

I've been in around change for a while. If you're reading this, I hazard you may have been too. 

I'd like to ask three questions about risk.

1. How much value or benefit do your current risk management activities add to your project or programme?
2. How many of your programme's issues were initially identified in the risk log?
3. Have you ever worked on a change initiative particularly beset by adversity (our nice word vicissitudinous springs to mind)

I'm guessing that in at least some (if not in most or even all cases) the answers were somewhat in line with "no idea, none, and yes of course".

So, putting my imaginary project sponsor hat on "What the hell are you spending my money on?".

Top tip #2 - Resource modelling in MS Project

One of the more important (and somewhat intractable) questions in project management is "How long could it take...?". Intractable because in the absence of a data driven answer, one or other existing 'compelling' dates is summoned out of air and a general preference articulated to complete the work in question by that date. Around 24 hours later (sometimes considerably less) that preference will assume the weight and authority of a board sponsored deadline.

If you're to dodge this bullet, there is a requirement for some preparation. First, you'll need to anticipate the question and the forum in which it will arise. Ideally, you control this. Next you'll want some sort of high level process or plan. The following is the sort of thing you should be able knock up in about 2 seconds and the only real requirement is for a fag packet to record it on. If you don't have the information needed to define the work to this level of detail - stay late until you do. I've included an optimistic forecast (in green), a pessimistic forecast (in red) and a median forecast in (orange).

Now in fairness, if you've got a requirement for a very large multi-hundred line project plan - that's not the sort of thing you can assemble in about the time it takes to drink a cup of tea. However, don't despair - you will nevertheless be able to apply the technique of resource modelling which I'll explain in a short-while. 

Where this technique really comes into its own is for activities which comprise a cyclical process which you'll negotiate dozens, hundred or even thousands of times. This could be migrating applications, migrating services between data centers or a large volume procurement type activity. Or anything for which there is a defined process. Typical project stuff, not quite all the same and not all different - but the sort of bread and butter activity that, as a professional change practitioner, you'll be comfortable with. And what you're really asking MS Project to do is use a set amount of resource optimally - something which would be extremely challenging to do via other means

Our process above has four steps roughly defined with some outline forecasting on work and duration. It translates to 4 lines in a project plan and it hopefully isn't to difficult to see how we got from the illustration above to the the outline plan below. I've set out the three different scenarios (optimistic, median and pessimistic) in green orange and red respectively.

I've made sure I've included both the work and the duration and you can see in the excerpt below I've assigned some resources (called optimistic, median and pessimistic).

Now, for reasons which are are better explained here you'll want to change the MS Project default from fixed units to fixed duration for this sort of modelling. But there is one little gotcha. When manipulating a project plan with the setting of 'fixed units', in theory MS Project shouldn't change the work or the duration when allocating resources. But actually - it does change the work on the first allocation of resources. When this happens you'll just want to pull it back to the original values. For the purpose of this activity - you'll only have to do this once for the 12 tasks shown below.

So - you might be thinking this seems like quite a bit of faff of no obvious gain. Bear with me and take a look at the next excerpt.

What I've done is created 10 process cycles under each forecast (optimistic, median and pessimistic). If I wanted to I could create 20 or 200 quite easily (and very quickly just cutting an pasting). MS Project very handily uses 'relative' (similar to Excel) predecessors and you don't need to jiggle these - it's cut and paste to your heart's content and in practical terms your limitation first and foremost is going to be your computer's capacity to level resources on project plans of many hundreds of lines .

Now, you have fair degree of flexibility in what you do next. You can plan for 2 resources or say 8 resources. (Just bump up the Max Units of the resource to 200% or 800% for instance). You can (if you must) put in constraints to that each process cycle must finish sequentially but as far as possible let MS Project do it's job of using the resources to the best possible degree. 

And, what I usually present to stakeholders is something that looks a bit like this. This is simply the timeline in MS Project with the appropriate summary tasks added to it. 

If I'm asked - I'll also supply the very first bit of network diagramming too. 

Note that one of the things I've done is tied in the time-lines with resourcing right from the get go. That's crucial as it starts to set in place the expectation (and dependency) that if I don't get the resources - I won't hit the dates.

What I'll do in a future post is put forward some fairly re-usable approaches using Excel and SharePoint for how you can prop up a workflow and have it generate rudimentary dashboards in real time to record and track the actual work being done.

Put project controls where you need them...

Although a quantitative kind of project manager generally, my default isn't to measure everything to oblivion and beyond. There's reasons for that of course but there's probably something to be said for '...if you can't measure it, you can't manage it...' and the less common supplement that '...if you can't manage it, you can't deliver it...'. But that's somewhat specious isn't it? Because it infers that measuring something is managing it. And managing it is delivering it. And, neither of these statements is true.

That all said - one of the first things I do on a new assignment is to check two things. The existence of Excel on my laptop and SharePoint on the network. Give me those two things and I can create custom lists and workflows which support well documented processes and dashboards in Excel which dynamically update and track the work minute by minute. People like that sort of thing - and I understand why.

Is this governance? Well, no - not by my yardstick. Is it control? Well maybe, maybe not. You might be either controlling what doesn't need doing or not doing all the things that need controlling. In practice you're probably doing both of those things to greater or lesser extent. Certainly more than you'd want do.

So what does the thinking PM do about this? We'll (of course) it depends on the size and shape of the red box in the below illustration.

So - with a little three dimensional visualization and we can see that we've got a relatively tight margin on cost but, a bit more leeway on time. Quality-wise - well that looks like we can just throw it over the fence and it'll be just fine.

You could quantify all this rather than just obsess over post graduate standard colouring in and this will probably bring some benefits.

But even more useful you could instate some fairly bespoked controls which would tightly control cost, control time satisfactorily and ensure that no more money was spent on quality control than was appropriate.

...but there's more.

If I were the big bad boss of change in an organisation, or the Chief Project Portfolio and Project Officer (C3PO) I'd set myself some loose objectives along the following lines.

1. Reduce the overall cost of project failure by failing less and failing quicker
2. Put in place KPIs / controls to support and track progress against the first objective
3. Develop quantitative approaches to inform project initiation and approval activities. i.e. stopping or changing potential candidates for failure before they start.
4. Reduce the amount of change incident on projects and, where inevitable, re-evaluate that project. We'll come back to this point much later on.

So how does our bit of three dimensional doodling above help with this? Let's take a moment to look a little closer.

There are the typical sam three variables which can be teased out and analysed.

Cost has a headline figure (well say £100) and we have some acceptable margin of variance (say £5).

Time-wise, it's a similar undertaking. We can say that we have a duration of 50 hours with an acceptable variance of 5 hours. It's not appropriate to overlook incorporating the level of effort in some shape or form - but I'm not going to worry to much about that just now.

Quality is just a little less straight-forward. And, I'd probably look to have some sort of qualitative assessment based on a scoring matrix. So we could have a range of (say) 1-5 to indicate a range of required quality from low to very high. There are undoubtedly numerous other approaches.

Sounding laborious and complicated? Not really.

We stick some numbers in a table that make sense. In point of fact your duration is more likely to be in weeks or months. And, your cost is likely to be in £000's. We can use some very straight-forward Excellery to create a very simple calculator. The specifics of which are very much down to you. It is the case however that as Duration, Cost and Quality figures goes up - so should your score and as the Variance goes up, your score should go down which is why there's a little fiddling about with the figures.

...but there's more.

The figures we could collate when reflected across historical information may yield up a threshold beyond which success becomes a very uncertain outcome. That's handy - because we could stop or re-configure those projects before they start - or at least make sure that they sit under the stewardship of the A-team, not (for instance) the B-team. Which bring us to an interesting point - the controls and scope of ambition ascribed to any project should clearly relate to the trust and experience of the project team.

Remember point 4 above which we'd said we'd come back to? If you move the goal-posts within a given project, all of the good work above will be for naught. The controls you had may not be the controls you need and you may suddenly find yourself sitting on a programme that far exceeds your appetite for risk. So, how can we defend against this?

Well, one way is to use the existing analysis above. You can see my use of the adjective 'moderate' and by inference there are other adjectives - maybe 'low', 'high', and 'abandon hope'. So, for me, I'd create a wrapper around the existing quality, cost and time constraints which constrain the project to remain within set boundaries. If you start out as a 'moderate' project, you finish as a moderate project. If this changes - you need to come back to project assurance to make sure you're still operating within acceptable boundaries. 

Because we've gone to the trouble of quantifying things - we're not limited to gut feelings in a project board meeting. We categorically know that if we have change (up or down incidentally) that is in excess of predetermined boundaries  - we must seek leave to proceed. This has the advantage that it will probably reduce project change. And very lastly, I think with a bit of work this could quite easily be incorporated in your organisation's default MS Project template - so once a project was baselined, it would immediately alert project planners if you breached the allowable boundaries.

On making up words and other PM proclivities

When I hear Sir David Higgins speak, I prick up my ears, gag my son and turn up the radio. As a result, I'm an HS2 believer where before I was a naysayer. I also think his option to use the word use of the word 'deliverability' in a recent interview was, if not revolutionary, a useful addition to the PM's lexicon.


This is, in my opinion, an acceptable bit of etymological license. It's got its feet on the ground, it's readily intelligible and hasn't been "Brittassed" or "Brented" to death. More likely to ingratiate than alienate.


Synergy, synergise, synergistic, synergies on the other had are the worst sort of PM excreta.


While the word synergy is a word, the other variations are not. Yes, yes neither (technically) is deliverability but who cares? You mention deliverability - you might put the ball in the back of the net. You say synergistic - you're more likely to lose the locker room.


Which is a bit of a shame, because hidden behind the grotesquerie is reasonable ambition which might even help with deliverability. So, I have set myself the task of finding a suitable substitute.  My principal source of reference was of course the work of Mssrs. Noblet and Yabsley


I can assure you, there will be no need for the stated 10-15 scalpel or razor blades, the biohazard or heavy duty plastic bag for animal carcasses or even (heaven forefend) a fistulated cow (I kid you not). However (and now is the time to pin back your lugs) Mssrs. Noblet and Yabsley's paper "The Good and the Bad: Symbiotic Organisms From Selected Hosts" does summarise nicely the following types of naturally occurring symbiotic relationship.

• Mutualism - a symbiotic relationship in which both host and symbiote benefit.
• Commensalism occurs when one organism (commensal) receives benefit from from the association, while the other (host) is not affected (neither benefited nor harmed)
• Parasitism - when one organism (parasite) receives benefit at the expense of the other(host).

Oddly there is a relationship in project management that is distinct and not found in nature called a commercial dispute in which neither party benefit and both suffer expense. Sometimes that's called egotism.

Good things come in threes...

There are certain topics that I've historically given a wide berth to. Project governance is one of them. It's too easy a target, too contentious and every one has an opinion on the matter. Here's mine.

One of the things that's constrained having any sort of consistent dialogue / vehicle / entry point (delete as appropriate) via which governance is (or ever could be) dealt with in projects (and programmes) is the lack of a unified consensus.

Well, you ask, what does that mean? Good question. If you asked 10 project managers what do they understand to be implicit in the term project governance you will (I hazard) elicit a variance in the responses that is extremely difficult to unify. You could even swim into less certain waters still by asking "What are the hallmarks of a well governed project by comparison to a poorly governed project". Could anyone be confident of any kind of spontaneous consensus?

The splenetic Steve Jenner (who will I hope overlook any referencing of his words in deference to my describing him as splenetic) offered that "...governance is the job of defining what decisions are made, by whom, when and to what criteria...". I remember feeling a sense of overwhelming gratitude to Steve for having (I felt) pinned the tail onto the donkey quite so precisely, succinctly and without apparent contentiousness.

But then, you've also got corporate governance, IT governance, data governance and quite a bit more besides.

...and that's just the different areas of governance. There are different approaches to governance such as collaborative or multi-stakeholder. And then of course there's applying both the required area of governance (say Project Governance) to (say) an IT Project in (for instance) a multi-organisational project. Oh, and just in case that wasn't enough - how about meta-governance - the governing of governance. (I do yearn for a bit of this now and again mind you).

Incidentally - just in case you were about to yawn and look for your entertainment elsewhere since this wasn't a terribly interesting topic, consider for a minute a project in which the wrong decisions are made, by the wrong people, at the wrong time and to the wrong criteria. Does this sound a bit more engaging now?

Like any project manager faced with unmanageable levels of complexity I revert to the mental state of a 3 year old, effect a brief but meaningful inner tantrum and cling to the mantra - surely there is a simple way of ensuring that tomorrow is better than today.

...and of course, there is.

Simple step one to improving governance today

Documented process is an enabler (dependency?) for improving governance. So, in the very simple situation below we have an input, a decision and two possible outcomes. We have a decision point, some inputs (criteria) and presumably a forum or framework in which that decision will be made. Sounds a bit like we're half-way to something Steve Jenner might just suffer to include under the heading of governance.

Simple step two to improving governance today

Create a register of all the formal documentation to be produced during the course of your project / programme. Include information about who is going to produce it, who is going to review it, who is going to sign if off and who is going to own it going forward. And, any schedule information for when this documentation will be created. Publish this register to appropriate stakeholders (and in many instances this will simply be all of them).

This resource, at a glance, allows a range of stakeholders to discern what documentation is to be produced (and what documentation isn't), by whom, when and who is (and isn't) involved in review and sign-off.

Simple step three to improving governance today

Create and publish a product breakdown structure (PBS). Derive from this appropriately detailed product descriptions and keep these someplace useful (say a product dictionary - a.k.a. an Excel worksheet. This is very much the same carry on for step two (which applies specifically to documentation) but in this case we're looking at specific project deliverables (which may or may not be documents).

This resource, at a glance, allows a range of stakeholders to discern what products are to be produced (and what products aren't), by whom, when and who is (and isn't) involved in review and sign-off.

And that will do won't it? No expensive management consultants. No particular undue stresses and strains. Not even a requirement for the oversight of assurance satsumas mandarins - just a bit of practical elbow grease.





Problems with unrealistic objectives

A while back I wrote a post entitled "Fail less with project management" I remain broadly happy with the sentiments I expressed. Perhaps the most succinct summary I can relay is via the metaphor of an aircraft flight in which a plane, encountering stronger headwinds than expected, will burn a bit more fuel and take a bit more time. It will however ultimately touchdown safely at its intended destination. A success in aviation terms. Not quite such a clear cut case in project delivery terms.

I only found out recently that the Association for Project Management has a vision of "...a world in which all projects succeed".

Does this mean that collectively we cease to embark upon projects that are deemed risky? Does this mean that project managers who do not succeed in delivering projects have failed?

And now for a bit of multi-choice.

There is often a tension between two competing imperatives for the project manager. The delivery of a successful project and the successful delivery of the client's strategy. As an experienced and seasoned project manager is it your job to;

1. Deliver a successful project whether or not it is the strategy of your client
2. Deliver the strategy of your client whether or not it results in project success
3. Neither option

For me and my blog, the answer is option 2. In a client facing scenario, what I'm actually going to say is that as a professional change practitioner my job (and that of the team) is to reduce risk, add value and accelerate delivery. 

For those of you who went  for option 1 - best of luck - I recommend you keep the phone number of your professional indemnity provider handy.

For those of you who went for option 3 - with some sort of managed approach to influencing the sponsor away from his or her chosen strategy (assuming you have good reason for doing so) I would offer the following. First, the most effective way to do this may well be to fail quickly and conspicuously in any attempt to deliver it. Secondly and more importantly, as an "experienced and seasoned project manager" that isn't your job. Is it? (If you can't live with this, fear not. A successful career in project assurance may be beckoning).

But I said 'problems' with unrealistic objectives. Not 'problem'. Personally (though it may yet limit my career) I don't have any problems in telling the client that we've burnt a bit more fuel and we're going to take a bit more time because the headwinds we encountered were a little greater than we forecast 6 (12, 18, 24+?) months ago. One of the reasons I don't mind is that my job is to add value, reduce risk and accelerate delivery first and to deliver successful projects second.

If you don't adopt this sort of credo, you might find yourself overusing words like "contextualised success", "positioning" or possibly even "the war situation has developed not necessarily to Japan’s advantage".
 
So, if your client has opted for an approach to change that you're not altogether convinced about, I'd recommend helping the client develop the necessary information assets as quickly and cheaply as possible so that you can back-up, rethink, re-plan and take a second swing at it.

Other views undoubtedly exist.

This is not a post about critical chain project management

Ever had an itch you can't scratch? I don't have one of those.

What I do have is a lingering doubt. A sense that something or someone could answer a lot of questions if I just knew the right questions to ask and the right forum in which to ask them.

Some of these doubts and uncertainties have been discussed previously in this blog. But for clarity I'm going to cite the half dozen things that (I think) are just not good enough in project management right now.

1. Planning. Consistently falls short of what is required. Work is not identified, not adequately defined, not forecast accurately and doesn't have the corresponding supporting information to enable the brokering of appropriate resources. Successors and predecessors aren't pinned down accurately, F-S, F-F, S-F relationships aren't identified with the precision required
2. Resource brokering. If (as is the case within technology infrastructure projects) your project's delivery is dependent on skilled and scarce human resources  then resource brokering should be fit for purpose. Hands up all those people who work on projects with great resource brokering.
3. Success. Don't get me started (mainly because there's an angry man here who already has)
4. Perception. Does anyone out there hold projects, project teams, project managers in high esteem?
5. Benefits / return on investment. Goodness me but there are far to many people charging out for delivering change when surely the game should be delivering benefits.
6. Profits. Executing change effectively and efficiently will enable you to deliver your strategy with less risk and more value than your competitors. It follows that your costs will be lest, your agility greater and your profits will reflect this. 

Notwithstanding all this, there's Prince 2 and PMP accreditations, quite a bit of software intended to solve some of the knottier project management headaches, a recognition that practitioners of change are specialists and even that some expenditure on those specialists is a worthwhile investment. And, health checks, specialist consultancies, more historical data than you can shake a stick at, more than one MSc and a partridge in a pear tree. 

Perhaps most importantly (and I can speak a bit from experience here) there's quite a few bright folks in the industry and there is no limit to their determination and tenacity in trying to bring projects in on time (whatever on time means - 'cos I'm starting to get a bit jaded around that - but more on the "Aggressive but Possible" (ABM) vs. "Highly Probable" (HP) paradox another time. Discussions of 'on-time' aside, there's still the questions of scope, quality, benefits and cost.

Could it be there's something fundamentally wrong with the underlying mechanisms of project management or their application? I could go the route here of an extended negative hypothesis routine in which we discounted one by one all the other candidate explanations. I'm reasonably certain that a blog post is neither the time or the place. So let's cut to the main course.

Planning assumption #1 then; there's something wrong with project management approaches / mechanisms. Or at the very least something wrong with the way they're applied. 

If this planning assumption is born out (and there's quite a weight of evidence to suggest that it might be) then ideally we'd need something waiting in the wings as a substitute. While the current status quo is (ripe?) for improvement, it wouldn't be impossible to make it worse.

And that substitute just might be Critical Chain Project Management. I'm not going to tell you what I think. There would be an implication perhaps that you too should be thinking that. I think we all need to come to our own independent conclusions about Critical Chain Project Management but I will post on it again in the future and if it elicits only critical scrutiny of existing methods and approaches then that alone will be a worthy objective.