Predictive Analytics: Solving the Construction Industry’s Gordian Knot

A 2015 study found a typical construction project typically has to address a host of challenges including, but not limited to, “delays, reworks, standing time, material waste, poor communication, conflict and being over budget, compounded by the global slowdown and the need to address sustainability issues” (Deutsch and Aia 2015, 262). Seven years after this study was published, the construction industry’s challenges are compounded with the ever-increasing complexities of the supply-chain system, labor and economic uncertainty, and the public health crisis—thus, the Gordian knot, or what may seem to be an unsolvable problem.

The Knot: Construction without Analytics

Architecture, engineering, construction, and operation (AECO) firms differ from the typical manufacturing industry in that their projects face high rates of change—i.e., location, materials, subcontractors, and management vary from project to project. Current issues facing AECO firms, such as labor and economic uncertainty, make finishing a project on time and on budget almost completely out of the picture. In fact, nine out of ten construction projects experience cost overrun and delay. In an industry where performance and projects are judged in terms of time and cost, these numbers are simply too high and indicate a widespread necessity to find a solution. 

Instead of being burdened with cost overruns, what if construction projects ran under budget, without delay; became harmonious to both clientele and corporation? 

Predicting the Future with Predictive Analytics 

What if you could predict the future? What if you could know weather patterns during construction projects, the probability of design changes, and other potential causes of cost overrun and delay?

Luckily you can, with something you already have: Data. Data is all around us—an infinite resource that has yet to be brought to its full potential, especially in the construction industry. Due to the vast amount of data and capabilities of modern machine learning, predictive analytics has the ability to cut the Gordian knot in construction management.

What is Predictive Analytics?

Predictive Analytics is just one aspect of construction analytics as a whole and is defined as a technology that learns from experience (data) to predict the future behavior of individuals in order to drive better decisions (Siegel 15). Note that here, individuals do not mean per person. Rather, individuals refer to individual circumstances, objects, or even outcomes. As opposed to having to make all decisions based on what a project manager suggests, which may be susceptible to inherent bias, predictive analytical solutions are empirical, meaning they are verifiable by observation and experience rather than theory or logic. 

Using data to drive outcomes is not a novel or new idea—many organizations have already embraced and integrated predictive technology into their business model. As a matter of fact, in today’s world, it is unlikely that you don’t already interact with predictive technology on a daily basis. 

Do you have Spotify? Waze? Do you shop at Target? All of these organizations have used predictive analytics to their advantage and have come to see the value in predicting unforeseeable insights. 

How It Works:

Predictive analytics functions by utilizing machine learning technology to transform risk into opportunity. While AECO firms undoubtedly should have some form of risk management involved in their business model, predictive analytics converts data from past projects, whether it be weather pattern delays, labor agreements, project size, materials needed, etc., to predict future trends in a project’s life cycle. For example, a construction project that utilized predictive analytics created a procurement tool that benchmarks a project’s final cost that converted contract negotiations taking an average of sixty days all the way down to two. 

According to Eric Siegel, founder of Predictive Analytics World, predictive analytics begins with a modest question (Siegel 153). For instance—what single factor in construction causes the most risk? From there, automated learning technology is able to take data—especially negative examples—and organize it to infer the unknown. The system used quantifies the empirical evidence gathered from data to measure a method’s relative strengths and weaknesses (Siegel 236). By eliminating time and money wasting ventures, AECO firms avoid complications that occurred in the past by predicting the likelihood that the same complications happen again in each individual scenario. All of this contributes to achieving operations maturity that results in future-ready performance. 

Effective Resource Management

The process of predictive analytics sounds daunting. However, one study permitted most of the causes of cost overrun in the construction industry are related to poor resource management. As resources become managed by a software system that will predict what you need and when, poor resource management translates into actionable data, allowing resources to be managed more effectively in order to complete a project on time and on budget. Furthermore, since companies likely own the data relating to resources being managed poorly, in the form of invoices and other historical data, predictive analytics functions as reverse engineering (Siegel 40). Fear not, there are plenty of ways AECO firms can get started using predictive analytics for a low cost with a relatively low time commitment—simply using resources they already have found in data.

Getting Started with Predictive Analytics

Uncertainty costs money, but predictive analytics can help with taking uncertainty out of construction projects. However, historical data has shown the construction industry’s lack of confidence in implementing modern technological resources into their projects due to the industry remaining document centric (Deutsch and Aia 2015, 261). Thankfully, there are a plethora of systems available to help bring the construction industry into modernity.

G2.com allows you to compare predictive analytics software to find what works best for each company and project. They list the key benefits of predictive analytics software to be:

• Accurately predict and forecast revenue numbers based on a wide range of variables
• Understand and account for customer churn and retention
• Predict employee churn based on historical factors for turnover
• Make more precise, data-driven decisions in all departments based on available data
• Determine both risks and opportunities that were otherwise hidden within company data

Predictive analytics provides companies with ample avenues for improved processes—allowing AECO firms to make smarter business decisions and avoid cost overruns and time delays that tend to plague the construction industry. Data-driven decisions provide a key solution: they guide business decisions based on factual predictions instead of inadequate guessing metrics which have led the construction industry into trouble. 

Get started implementing predictive analytics into your projects today. Watch your worries disappear as your wallet, and your watch will thank you. You can begin to access the benefits of predictive analytics using construction schedule analytics such as SmartPM, which captures and analyzes your construction schedule data and then helps you identify problems in your project before they arise.

By embracing data analytics, the construction industry may solve its age-old problem. 

Works Cited:

Deutsch, Randall and Aia, A.P Leed. “Leveraging Data Across the Building Lifecycle.” Procedia Engineering, vol. 118, 2015, pp. 260-267. https://doi.org/10.1016/j.proeng.2015.08.425.

Siegel, Eric. Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die. John Wiley & Sons, Inc., 2016.