What is so Great About Cost & Time Certainty on Projects?

In Greek mythology, the Sirens lured sailors with their enchanting music and voices … to their deaths on the rocky coast of their island. 

And the drive for up-front certainty on capex projects is like a siren - it lures us into dangerous waters where our projects run aground. 

Because certainty, like the mythical Siren, is unachievable, and trying to achieve it is folly. 

We are happy to accept the app on our phone telling us there is a 76% chance of rain tomorrow, because we know weather forecasting is an uncertain science. Even though the forecasters have supercomputers, are only modelling a few days into the future, have lots of PhD's, and the physics of climate is relatively well known, we accept uncertainty.

But on a project we are trying to predict at least 12 months into the future, the best we have is a finger in the air and a spreadsheet, and we haven't yet found the mathematical formulae describing the behaviour of clients and contractors.

But we (and especially project clients), want a 100% certain prediction.  I read two separate prices of research this week into UK infrastructure projects, and both highlighted this 'cost and time certainty' as a top client requirement.

I'm sorry to be the bearer of bad news, but the future is unpredictable. Always has been, and very probably, always will be.

And the inability of project clients to embrace this fact is probably the main cause of the problems that they are suffering from today on their projects. Trying to force certainty onto projects, is what is causing projects to have such a woefully poor success rates. 

It's a vicious cycle, and creates most of the waste and inefficiency in today's projects - people trying to do the impossible, as if the project itself wasn’t hard enough. We waste time, money and our best resources, by making a detailed 10,000 task prediction, and then during execution we force people to fit our prediction. Task estimates have safety-on-contingency-on-safety, and if someone has the audacity to tell the truth ("It was just a guess, and it didn’t work out how I guessed it"), it can be career suicide!

Embracing uncertainty however, is not the same as accepting chaos and not managing - far from it. In fact, accepting uncertainty and using methods to procure and manage projects that are designed for it, deliver much better results.

We can be boil our choices down to two options.

Option A:

Try and force the world to be 'certain'. Push risk down the project supply chain, and plan and manage the project based on fixed price contracts and fixed time deadlines for each work package. 

Aimed for results….

• More reliable and predictable projects
• Best value for money

Achieved Results…

• Very poor predictability (very few projects achieve their original time/cost/scope objectives)
• High costs and long durations
• High levels of waste and stress in the process - people are very busy and trying hard
• Predictability comes at a cost (higher budget, longer schedule)

Option B:

Embrace uncertainty. Protect the project objectives with managed scope, cost and time 'buffers', which provide protection for the whole supply chain. Use collaborative contracts and CCPM (Critical Chain Project Management) to embed this idea into a practical operating process.

Aimed for results….

• More reliable and predictable projects

Achieved Results…

• Reliable and predictable projects (90%+ on-time and on-budget)
• Lower costs, and faster projects
• Higher efficiency & productivity
• More profitable supply chain, despite the lower costs to clients
• Less stress

But to take Option B, you will probably need to throw away your library of model contract forms, your planning & project control systems, and the notes from hours of training courses.

The reason why Option A (ie what we have been doing for decades), will never work is down to the science of systems and complexity theory. 

Without going into too much detail, projects are examples of things called "Complex Adaptive Systems" (CAS). And CAS's don’t work well if you try and manage them using the methods suited for simpler deterministic systems.

Option A is not how to best manage CAS's. Training and incentivising people in our currently accepted 'best practice', just won't work because it is the wrong tool for the job.

Option B is an example of an approach that is more suited to managing a CAS, which is why it works. Where it has been used on projects it delivers much better results, though sadly there are, as yet, few cases in capex projects - we are still in Option A mode.

Here's a recent example from manufacturing to illustrate that this is not just some weird academic theory.

Michelin is the second largest tyre maker in the world, with over €20 billion in annual sales. They make about 200 million tyres a year, and have been in operation for over a century. So you would expect that by now they would have got quite good at running their business.

But in 2016 they changed how they controlled their manufacturing plants and distribution system. 

Previously they ran their supply chain in the way that almost every other manufacturer did. They had an MRP system built into their ERP system, which was used to trigger manufacturing orders and distribution movements. Since this didn’t deliver the results they wished for, they added further complexity to MRP, with forecasts and manual overrides. 

In project management we have our Iron Triangle - "Cost, Time, Quality - choose and two!". The supply chain managers' equivalent is the trade-off between service levels, inventory and lead-time. And this is what Michelin had historically done - in order to achieve great service levels and availability, they held high stock levels.

In early 2016, one factory turned off the MRP system, and used a new method called DDMRP. DDMRP is based on the science of CAS's, and like my Option B above, it embraces uncertainty and manages it using strategic 'buffers'.

Within a month, this factory noticed dramatic improvements. Service levels were hitting new highs, whilst inventory levels were falling.

A year later, they are using DDMRP in most of their factories, each repeating the pilot study success, for example one site improved service levels from 89% to 99%, whilst reducing inventory by 10%. And reducing lead times are helping them increase sales at the same time! 

The Michelin case has many parallels for construction - supply chain managers have been struggling for decades with poor performance and predictability, methods like lean were producing inconsistent results, and they put their faith in the global software giants to know what was best. Also Michelin was not an immature organisation - they had invested significantly in their supply chain process, and in terms of comparative benchmarks, were doing OK. They just didn’t realise that they could be doing so much better.

Until DDMRP came along, most were trying to do the same as we are in construction. Making the forecasts better. Then trying to force certainty onto an inherently uncertain world.

It is only in the past few years that innovations like DDMRP have shown supply chain managers, that there are techniques that can be used to produce significantly better results, with a process that is easily learned. 

We just have to close our ears to the alluring call from the sirens of certainty, and have the confidence to realise that what we thought was helping us, was actually having the opposite effect. 

Ian is an independent coach and advisor on improving capex projects.

If you are interested in how you can manage projects and project portfolios without insisting on "100% certain commitments", read Ian's report "The 2 Changes that will Transform the Performance of your Capex Projects".