The highest 5% of trades on a construction site out-performs the lowest by more than 4.5 times.
Within a typical construction site, 35% of the work is done by the top quartile and 11% is done by the bottom. The variance in pay for these guys is minimal, almost the same. When you stop to think of the economics of this arrangement it starts to go some way to explaining how projects can be so successful as well as end so dramatically.
The image above represents a population of trades, each bar is the average productivity of an individual over 60 days, the y axis is MPS (Mafic Productivity Score) which reflects the time-on-tools. The difference between the highest and the lowest performers is quite clear.
The same data can be represented in a slightly different way. The scatter chart below represents the time on tools and active time for a complete workforce. Each dot is a day from one person, the x axis is the rate at which they are using tools and the y axis is their active time for that day. The productivity or MPS is represented by the colour, darker green representing the higher productivity. Productivity increases to the right and up. Efficient operatives can also be identified, higher tool use and lowest active time whilst less efficient operatives can be seen on backside of the scatter, low tool use and high active time. What is very clear, as in the figure is the large range of outputs seen on a site and hence how difficult forecasting and planning can be when you have a variable such as this to contend with.
Is construction really ‘low-skill’?
Quite often construction is considered a ‘low-skill’ career and people often fall into it when they run out of options. To get a job on most construction sites in London you need a CSCS card, complete an induction and then turn up, often you don’t even need to arrive with PPE. Once on site, a piece of work is assigned and cost, schedule, quality and safety firmly sit on the shoulders of the new individual, regardless of their experience or capability. Repeat this for everyone on site and it is easy to see how you can get so much variance in cost and schedule even with the best planned, organised project. Often you will hear trades talking about being “worth their weight in gold”. When you consider the cost per unit output of the best and worst people on a site, I don’t feel this is so much of a metaphor.
How we know this?
At Mafic we measure productivity by combining active time (doing something with your hands) along with the time someone spends using a tool directly. We call this MPS or Mafic Productivity Score. This avoids the troubles of CPI where the number is often tainted by budget and allows a true apples to apples comparison across different types of work, trades and also projects. All of this information is derived from the movement of the head. Every construction activity has a unique movement signature which we have captured and trained an ML model for. This allows us to count the metres of weld laid down, number of times a nail is hit or the number of screws installed in a wall. What we have been able to do with the ML is to paint a picture of exactly what is happening on a construction site and hence removing all the ambiguity often seen in the data collection and productivity analyses. The data above is aggregated from approximately 60 000 hours collected from construction sites in London, powerful stuff!
5 ways can we improve the output of the workforce without sending the workforce [back] to school
1) Hiring and Retention
To start with, getting the right people in through the door or keeping the right people is the first opportunity to reduce labour costs and improve schedule. Building a relationship with your Gangers, resource providers as well as individuals (current or future) is paramount. This should start with invoicing and payment. Paying on-time or if you can’t, being transparent is the best opportunity to build that loyalty. I recommend put a plan together and assign a budget if you can, to build and maintain these relationships. Even if it’s the lightest plan you can think of, having this written down along with monthly objectives can tell you quickly if you are achieving your goals.
2) Motivation and incentivising
There are a couple of ways you can address this. It can be a natural extension of your retention policy, introduce a recognition system where vouchers, clothing (good clothing) is awarded publicly every couple of weeks to individuals who have performed well or demonstrated a positive behaviour. Alternatively you can reward people for other positive behaviours using data you may already have. Individuals doing half days is expensive and hurts your schedule. If someone leaves site early, less work is done which drives up your preliminary costs and adds more days to your schedule. We have also found people intending to do half days generally perform at a lower level, we are not yet why but it could be similar to that Friday feeling, wind-down syndrome.
3) Enablement
On average, more than 2.5 hours a day is spent ‘off work-face’. Eliminating the reasons individuals have for leaving the workface is one of the biggest opportunities to improve productivity and then schedule. From the data, we have found there are individuals who have very high tool counts however they find themselves towards the lower end of productivity curve because their workface time is low. When they are on workface they are stella but their average time spent here is poor. The reasons they leave the workface are not having all the materials, not having the right tools, seeking technical help from a supervisor. This is all stuff which can be remedied by the management – immediately. Kit all the workfaces with exactly the right materials, ensure the individuals have the right tools and perhaps reduce the range of activities someone is doing, use a ‘one-man-one-task’ approach for individuals who struggle to stay focused. Ensure your supervisors have adequate coverage for all. You don’t need sophisticated tools to do this and the upside could be up to an extra hour on workface per man every day, 22% reduction in labour cost! (you probably only have 4.5 hours of on workface time).
4) Site setup
There are a number of things you can do straight away to improve inherent site performance. We have seen getting the breaks right in a shift pattern can save up to 8% labour cost by reducing fatigue (QUICK TIP: make sure you have at least 2 breaks and don’t run a session between breaks of more than 3 hours). Unless you have disproportionately high preliminary costs ensure your working day is not more than 9 hours + 1 hour break.
5) Design
If you are running a team of more than say 30, the chances are you have not hand-picked them and so the range of capabilities is likely similar to the images I presented at the start of this article, i.e LARGE. You should be looking for an assembly process on site which is not overly complex and hence any one individual regardless of where they sit on the performance curve can crack on with the work and be successful. You are not asking them to negotiate a complex process which includes 20 different sequential activities, each with high tolerances, multiple tools all with a high chance of failure. Ideally this is as simple as possible which will allow a range of capabilities to be equally successful. For example, colour code all your head track for different widths so it can be immediately seen what the spec is without getting their tape measure out. Kit the apartment or space properly, don’t give people materials they don’t need. Its like those excess pieces of Lego at the end of a build, what to do with them?
Where do we go from here?
What I have hopefully demonstrated here is the size of the opportunity. As an industry, up until now we have been completely blind to these opportunities since we simply have not had the tools to measure them at the detail we need. At Mafic we help construction companies better understand exactly what is happening on their site and drive improvement from all angles. Using this data I want to generate a public source of information which can be used for immediate improvement. There is so much you can do before you get stuck into the data, the only problem being without Mafic technology you have to watch the bottom line to see benefits flush through instead of the Mafic daily dashboards. Focusing on your workforce makes a MASSIVE difference to your cost and schedule performance and also helps you build a more engaged team which your people and customers WILL choose over your competitors.
