Back in February we wrote a post about conceptual estimates and the challenge of effectively using historical data to create them. In response, we heard from Assistant Professor John Hannon of the University of Southern Mississippi.

John sent us a recent study he published with the Association for the Advancement of Cost Engineering International, entitled, "Historical Data Practices of Construction Contractors." In it, he analyzes 130 responses to his survey on how contractors manage and use historical project data.

While our original post was focused on how building owners create conceptual estimates, we found John's survey of contractors highly relevant. In part that's because the processes are similar, but also because so much of building owners' planning is based on contractors' costs.

I'll share some of John's findings here:

What sources do contractors use to capture and analyze historical cost data?

• Commercial database   -   5%
• Database self-created and maintained  -   33%
• Estimates saved for reference   -   23%
• Accounting & job cost module  -   35%
• Other  -   4%

On the one hand, I like that job costing software represents the largest source of cost data - that's what it's for. In addition to keeping the books in order, accounting and job costing applications provide a wealth of insight into how the business operates and how it can improve. The challenge is digging in and analyzing that data. Someone has to be proactive enough to take on that task.

In a subsequent question, it was revealed that 57% of respondents are using spreadsheets as the foundation for their data collection. That's not surprising and supports what we discussed back in October about preconstruction management.

What type of data is collected?

• Costs 96%
• Productivity 56%
• Resource usage 28%
• Other 4%

Clearly, most respondents are focused on collecting and analyzing cost data. John points out that while cost data is important, it becomes less relevant as material and labor costs change over time. Productivity analysis may be more relevant, especially when applied to up-to-date material and labor rates.

How quickly is this data available?

• Immediately 25%
• Within a week 35%
• Within two weeks 9%
• Within a month 17%
• After more than a month 9%

This last set of data is less encouraging. With a properly implemented IT infrastructure (i.e. accounting, job costing software) users should be able to run real-time reports on all historical and in-progress projects. Of course, there is the human factor of having to input data and close the books. Regardless, all historical projects should be available for reporting and analysis immediately. There's certainly room for improvement here…

What are the impediments to collecting data?

When respondents were asked to identify the major impediments to capturing historical data, "problems capturing data in the field" was far and away the biggest issue. When investigated further, this problem also tied back to software implementation; the vast majority of respondents indicated that their software systems did not integrate effectively or that the level of detail in data collection was insufficient. In keeping with my mantra, I'll assert that these are mostly people problems, not software problems.

You can download John's full report, which goes into far greater detail than this post. I'd be curious to hear folks' reactions to the report.