New Zealand Engineering 1997 November

Confessions of a Rustler

feature by Ian Marsden

A race is not a stampede, but in business it is often hard to tell the difference.

A stampede is a race being run by a committee.

Remember the good old days, when an engineering project had enough time, money and knowledge to do the job "properly", in a sensible time frame? Unfortunately, the demand for stone pyramids has dried up now. Today there always seems to be a pressing need to fast track for a deadline; a shortage of funds and a peppering of holes in the scope of work where business decisions have not been made or technical problems are not resolved.

How does a project manager survive in this environment? No doubt all develop their own ecological niche, every business environment and personality being unique, aside from whisky and aspirin which I understand are universal constants. The following suggestions are my own and aren't necessarily the right way to do things, (whatever that might be).

Preparing for the stampede

Before you enter a race, you need to prepare months in advance to run it. Infrastructure is the first key to speed, like adding wings to your keel.

Make sure that you have a good set of standard, purpose made, general contract documents and technical specifications on tap, before you start. With these at your finger tips, it is possible to launch work in a day, that might normally take a week. As desperation mounts later, part finished drawings can be heavily stamped "preliminary" and stuffed down the fax machine along with the standard documents and a short covering note, and voila, you have a tender. By the time prices come back, the drawings are finished and work can start straight away, with the job still sharply (if not precisely) priced and properly under control. Without these standards, the same process results in expensive extras, blunders, disputes, incompatible hardware, delays and ulcers.

An accounting system is another prerequisite, and I am not talking about the organisation's own geriatric dinosaur. Standard accounts have two priorities: the first to satisfy legal requirements for taxation and shareholders; the second to reluctantly part with a few confusing figures to pleading managers to help run the business. Project accounting is usually only a long promised, distant possibility, that will happen one day when the IS staff find spare time.

Accounts that are available are usually at least six weeks behind your project and riddled with a confusing array of accruals, reversals, errors, and omissions of invoices and orders which haven't surfaced in the system yet. They are also useless if they can't be accurately assigned to a known point in time (to match against physical progress).

To control your spending properly, resign yourself to running your own system. The basis of my own system is the simple equation:

$ invoiced + $ ordered + $ estimate of remaining work =

$ projected to spend.

The difference between this and the approved budget, is what I have left over (ie: the remaining contingency sum). To maintain this equation it is necessary to accrue every order as it is written, then replace that with the actual final cost figure when the invoice arrives. A database-type program on a PC is one way to handle this; I run one assembled for me by a talented colleague, using Microsoft Access. The "estimate of remaining work" is reviewed fortnightly by looking at the actual work progress and figuring what it will cost to complete each remaining item, in light of the best current knowledge. Once properly set up it takes surprisingly little time to run a system, and, as I discovered to my cost in earlier years, a lot less time than perpetually, fruitlessly, trying to unscramble the organisation's own accounts.

Speaking of money, another thing you need to construct and maintain during your career is a "ballpark" database of costings and time requirements for typical work, based on previous jobs. Why, will become apparent later. From previous jobs, get a feel for what typical buildings cost to build per square metre, what typical pipework and drains cost, conveyors, platforms, pumps .... whatever are the most common building blocks of your line of business.

Staff are another project prerequisite; even if they can't be hired initially, the necessary consultants, draftspeople and contractors need to be lined up ready for an immediate start when the project finance rolls in. Using the same regular outside resources has the advantage that they too know your business, and already have some of that necessary infrastructure in place.

Starting the stampede

Like an America's Cup race, the technique to start a fast-track project is to be right at the line and going full speed when the gun goes off. The difference lies in not knowing when, if ever, it will go off.

The first pre-start problem for a project is "cash 22". To get money to make the project happen, it is first necessary to present a good proposal, properly programmed and budgeted, with the payback and so on all worked out. This research requires resources, which cost money. Money is available after the project is approved to proceed. Approval will follow when the proposal stacks up properly. Deja-vu? It is also required to investigate "all options", including all interrelated sub-options, the combinations and permutations thereof resulting in infinite work required, immediately, with no funding.

If you have ever been trapped on this merry-go-round, you will soon learn that the only way off is to cheat. This means finding free resources, virtual cash and a warp drive technique to breach the infinite workload.

A free resource is someone being paid by someone else. Borrow people in your own organisation who don't have to charge their time internally. Also, identify the external companies who stand to pick up profitable work later and get them to do some of the work for free. For example, if you are buying a cooker, get cooker manufacturers to offer typical process layouts and budget prices.

Not everything is free, hence the need for at least a small amount of cash upfront, which hopefully the organisation has enough forethought to budget. Unfortunately, it is a universal disease of large organisations that the purse holders are so remote from the action that they are incapable of managing it. Those at the coal face usually learn to play management juggling games with the resources at their disposal in order to get their job done, bending the rules to help the organisation in spite of itself. Enough said, consult your local juggler. This meeting never took place. Chew thoroughly and swallow this magazine after reading it.

Warp drive is the "ballpark" cost database mentioned earlier. There will never be time or resources to do proper designs and price them out in the pre-project roundup. Only the key large items of plant can be fully priced at this stage, use the database to rapidly fill in the rest of the picture. Do I hear the purists gasp in horror? Consider this: several weeks down the track when you finally do have definitive designs out to tender, what spread do you expect on the prices coming in? Typically ± 30 percent on the mean? Then how do you expect to do better, no matter how much work you put in? The main things to remember are: to be consistent with estimates, so that when selecting the favoured option, you are truly comparing apples with apples; and don't screw the estimates down too low to get approval, no one will thank you afterwards for overruns and cut corners.

Last but not least, the project itself needs to be sufficiently scoped to fill those previously mentioned holes, it is the unknown work that is most likely to blow the budget and the program.

Awaiting the starter's gun

Like the America's Cup, momentum is the key to a good start..... and there is no rule against cheating, provided that you cheat fairly, within the rules.

In practice this means getting started as far as you are able without actually committing more than token cash, before official sanction arrives. Tenders can often be issued for the larger items of "design and build" work with minimal work input, so that successful tenderers are selected, and the cost of their work defined, before starting. A word of caution, the contract and tender have a defined time base, which, if there is excessive delay in starting, can become toothless. Take care not to leave yourself stranded with a successful tenderer who has neither the manufacturing window nor the obligation, to perform on time.

Another great time waster is the consent process of local authorities. Wherever possible, the necessary cash should be scrounged to get permit applications in early, even if this means engaging consultants, and putting in more than one application to cover differing options.

The race strategy

Inevitably, someone will call your bluff and approve the project, too late to get it done in comfortable time. Now you can see that if you have prepared, you are in a far better position to meet that squeezed deadline.

The kneejerk reaction is to leap straight to the critical path program. This is a handy tool, essential for large projects, but a mixed blessing. It has limited use unless a consultant is hired to constantly maintain it, in laborious detail, and even then it rarely gets down to enough detail to truly avoid those irritating delays. For example, I recently lost a week on a critical job due to an unexpected delivery delay for two extra lengths of pipe. The best value I have found in these programs is at the strategic level, coordinating contractors whose work interrelates in the same space on site. But even here the program can be a hindrance, because it becomes a self fulfilling prophecy. Opportunities are lost when contractors watch the program instead of the work, and miss the chance to start sections of work early. If you are in a hurry, hide the damn program, put a competent overseer on the job and give him a portable phone.

Even with only a rudimentary program, for projects up to around $5,000,000, you should easily be able to identify the time critical work. But the important thing is not to become so focused on the critical path that everything else takes a back seat. A day saved at the start of the job is more valuable than one saved at the end, for at the end it is often too late to save any time.

Attack the job on as broad a front as possible with whatever resources you can afford. Rather than asking yourself "How long have I got to get this particular job done?", ask yourself "What is to prevent this from happening right now?"

Often you will find that with very little effort, large slices of non-critical work can be launched early, without encroaching on the resources for the critical items. Materials can be ordered early, pipe mains run, fringe demolition and alteration works done and so on. This may not be so hot for cash flow, but the point is that every job becomes critical if left long enough. Pushing work early also often reveals unforeseen weaknesses in the plan, de-cloaking like Klingon warbirds in your midst. Unexpected plant shutdowns needed, national material shortages, contractors tied up on other large local projects; all discovered early, can be accommodated.

Work eliminated early leaves resources and space clear for the final sprint at the finish. Remember, time is nature's way of preventing everything from happening at once, there is a limit to how many things can be done last.

In order to tackle a broad slice of work, you will soon discover another major bottleneck. There is one nitwit making all the decisions, who holds up everything ..... that person being yourself! The project manager can take any and many a role in the project, but ultimately there is a cutoff point where you must delegate, and this itself is the earliest and hardest decision to make.

Most managers hate to spend money on someone else's time, to do designs and make decisions that they would far rather make themselves. It is not necessary to steer the boat yourself in order to win the race, only be sure to pick a good crew.

Haste and speed - the difference

Crash programming is a favourite buzz word, and yes, you do need to speed up the items on the critical path in order to compress the program. But there are fish hooks in this. Contractors don't want to work weekends and overnight. Even humble tradespeople are human, and now that penal rates are largely gone, the enthusiasm for screwing up their personal lives to meet your imperatives is minimal. When they work long hours, and especially seven day weeks, the cash haemorrhages out but the productivity declines to a trickle. Adding extra people to work in a confined space also slows everyone down and increases accident risks.

The best way to use the seven day week is to split off specific "obstructive" tasks for weekends, and leave the normal working days clear for the rest of the herd to stampede around in. Finding one contractor to lay a floor topping on a weekend, or to alter a cable run that shuts all the power off, is a lot more realistic than trying to get everyone in to work on everything, continuously, all at once. People, like painters, will actually prefer this arrangement, rather than having to constantly peel off adhering electricians and plumbers from their freshly painted walls. Another word of caution, make sure your five day work crew truly are only working five days, and not being sent off by their boss to work over weekends for someone else! Otherwise you will be paying top dollar for worn out zombies.

Time is the ultimate enemy, the calendar against which you race has no extra days available in it, and all must be used to best advantage. A tender mailed on Friday will be looked at on Monday morning. Tenders mailed Monday to Thursday, will still be looked at next Monday morning, because it takes two days in the post, and because everyone at the receiving end gets tied up with the end of week rush. Miss your Friday deadline and you lose a week.

Develop routines which flow with the rhythm of life not against it, and trade haste for speed.