It’s 106 degrees outside as I’m writing this, cold beverage in hand, just after the conference has wrapped up here in Scottsdale.
I, along with 250 of my colleagues and peers from across the industry, just spent three days listening, sharing, and networking with each other, finding out how each of us is tackling the task of collecting, cleansing, transforming, storing, and analyzing company data. Attendees were from across the construction industry, from 100 to 2,500+ person firms, general contractors to heavy/civil companies. CFOs, COOs, and CIOs mingled with Directors of Innovation, Project Engineers, VDC and Safety Directors, PreCon Managers and Data Analysts. The common themes I heard from people attending were to benchmark themselves against their peers and learn which KPI’s other companies found most beneficial for their organizations.
That so many firms were attending, I took it as very encouraging for our industry. In an industry that is often behind in terms of technology adoption, the recognition by so many that data is an important company asset, and that management of that asset can provide significant long-term value, is a positive sign for our industry.
The following are a few observations from the conference. I had the opportunity to talk to roughly a third of the attendees over the three days, and very few of those had a clearly defined data strategy in place. The common theme seemed to be that data analytics started within IT, often with the development of a data warehouse to pull in data from one or more disparate systems, from which some initial dashboards were built. For some, a dashboard request originated either from accounting or operations due to limitations of ERP analytics tools, although only a few indicated this was the starting point.
Many attendees seemed to be struggling with where to start with their data analytics approach. They generated data they believed would be useful for the business, but they didn’t necessarily achieve broad adoption. The successful ones recognized that their data analytics program needed to be a business initiative rather than an IT one. They identified that, in almost all cases, a lack of common understanding of the meaning of certain data fields and lack of trust in the data were initial stumbling blocks. Until they developed a data governance approach to improve data quality and created some training and common terminology for certain data elements, obtaining buy-in was a challenge. By identifying data custodians/owners within the organization and establishing a partnership between business and IT, those that employed this approach saw a significant increase in data quality and analytics engagement.
Many of the attendees were from PreCon/VDC, looking for examples of what other firms were doing with the mass of data collected at this stage of the construction lifecycle. While a couple of the presentations I attended touched on this area, it was not one where I heard a lot of innovation. The lack of data consistency between the different applications used at this stage required a significant amount of data cleansing and the implementation of governance programs before analytics could yield any substantive value.
Safety metrics was another area where there was a lot of interest and activity. There were several examples of how contractors had been able to leverage safety data to derive meaningful insights. In one example, a contractor was able to identify that certain safety issues always occurred at the same phase of a project, prompting their project managers, superintendents, and foreman to prioritize safety training and reminders during that specific point in a project. Another contractor discovered that the majority of injuries were a certain type of hand injury, leading them to collaborate with a glove manufacturer to develop a specialized glove that provides extra protection in the area where the injury typically occurs (quite impressive if you have enough leverage to do that)! However, we also heard about some of the challenges associated with collecting safety data, including an important lesson learned about performing quality assurance and quality control on one’s own data. One contractor, upon reviewing their safety records, discovered that one of their safety inspectors used the exact same photograph for every single inspection!
In terms of technology used, by far the most common analytics tool was Power BI. While some mentioned Tableau, Domo and Qlik, Power BI was the clear preference, as was the Azure platform for data storage, whether that was in a data warehouse, data lake, or a data lakehouse. The discussion on data warehouse versus data lake was one of the most popular sessions at the conference. As expected, considering how many of the data analytics initiatives began, the initial focus was on a visualization platform rather than the overall architecture. Some attendees were using middleware and data piping tools to ingest data into a data storage platform, (mostly some variety of Microsoft SQL, Synapse or similar), however, not all. Very few had considered the benefits of a master data management platform, data quality tools, or other critical components such as a data catalog, data map or data glossary to capture metadata about their organization’s data.
While cancelled flights caused some challenges for the conference organizers, resulting in several scheduled speakers being unable to attend, overall, the conference provided a good level of insight into what was possible and what has been successful for those who are further along in their data analytics journey. It confirmed that data analytics for construction is a key investment for contractors. However, to be successful, it is necessary to have a plan or strategy that incorporates both IT and the business. While technology is a useful tool, people and processes need to be the driving forces behind any data initiative.