Estimates are not reality

Estimates are not reality - no matter how much we might hope them to be.


You know the scenario: 
Roadworks are put in place. An end date is scheduled. The end date is then missed.


‘Why can't we plan these projects so that we deliver the work on time?’ is the common cry!


A number of us in the Midlands have seen this scenario with the roadworks on the M5 at Oldbury Viaduct (between Junctions 1 and 2). The project started in January 2017 and was scheduled to finish in October 2018 - so  22 months in total.


The October 2018 date had been pushed back to Spring 2019 and now Autumn 2019, an additional 12 months (or so) - or an increase of 54% on the original timescales.


The headline announcement in the news blamed this delay to the work on the heatwave we enjoyed last year. You have to dig into this a little bit to understand that. When the planning of the project was done, one of the significant activities required was to apply a waterproof seal to the viaduct. This was a large job and teams of people were scheduled to work 24 hours a day, 7 days a week. 


Now think about that for a moment. 


That means that anything which delayed or added to the work would introduce a delay to the project. Anything. 

A shortage of the materials used for waterproofing the viaduct = delay. 

A number of people working on the project leaving for (say) HS2 or other work = delay. Or as it so happened - the heatwave. 

Apparently, the surface temperature has to be below 30 degrees centigrade before the waterproof sealing can be applied. No one could reasonably have predicted the extended (and fabulous) summer we had last year but it slowed the work on this significantly, as work could only be done overnight.


To be fair, while the media have focussed on the heatwave, the Highways Agency have pointed to other factors. Two in particular: the weather and the size of the challenge.


The severity of the 2017/2018 winter caused additional delay "… with freezing temperatures and snow delaying our concrete repair operations and causing issues for workforce safety". (Source: Highways agency progress report).


Additionally, as is often the case in these sort of projects, work needed to start before the scale of the work could be fully understood. The initial estimate for the number of repairs needed was 1500, whereas 6000 were actually needed; a fourfold increase.


I trust from just looking at this you can see that estimating a project like this involves consideration of a large number of factors - some inside your control and a large number outside your control. 


Best practice Project Management says that you should break down your project into stages to allow you to estimate stage by stage. The more you know about work, the easier it is to estimate in detail. The trouble is, that the initial estimates for the project often become the definitive numbers for the project and the initial estimates become easy targets for people sitting on the sidelines.


When any project is run in the public domain, there comes increased pressure. News media outlets have an input, politicians have their say and local businesses and motoring organisations all put their pressure on to get the project done as quickly as possible. These are all stakeholders and the effective engagement with them is a key part of project and programme management.


Which brings me back to the estimating. Why would you estimate a significant part of the project at 24/7? Within projects, risk management tells us that an assumption (we can deliver this work around the clock) can lead to a threat to our project if the assumption is unsafe (we will overrun). 


In an environment where stakeholders are very visible and vocal, such as a project like this, any date put up will be subject to intense scrutiny. ‘What can we do to deliver the project faster?’ will be a common question. Working around the clock is one answer, however, that comes with its own risks which can hurt us significantly.  


I do think we need a grown-up approach to our estimates on projects. An internal one which allows estimators to come up with reasonable estimates regardless of external pressures and a publicly held one which acknowledges that estimates will never be a reality.