Embrace progressive data strategies to drive improvement to top and bottom lines.
Executive Summary: This is Part 1 of a three-part series about how the construction industry is embracing progressive data strategies to drive improvement to top and bottom lines. Recent “FMI Quarterly” articles focused on the “why” and “what” around data management (see “Deconstructing Data” by Madison, Tison and Paré). Here we will focus more on the “how” of developing and implementing a data strategy for your firm. For this subject, FMI has teamed up with Tilson Consulting, a leading IT advisor to the construction industry.
Get work. Do work. Keep score.” Those six words, as coined by FMI’s founder “Doc” Fails 60 years ago to describe the business that contractors are in, have reverberated the walls of our industry for decades.
If asked, most of our clients would tell us that the not-so-sexy “keeping score” aspect of the business excites them the least. Until now, “keep score” has invoked images of accountants sitting in the windowless caverns of the office punching away at their 10-key pads.
“Keeping score” has been manual in nature and disconnected from the operational piece of the business. Most people enter this business to build tangible structures, not spreadsheets. But in the margin-compressed and chock-full-o-risk world that you live in, being smart about improving fees and protecting your balance sheet will require not only accurate score-keeping, but also the ability to better predict and understand what is driving the scores. Just as your project and field managers should know the accurate “score on the project scoreboard,” owners and senior management should know the “score of the enterprise scoreboards.” We meet many owners and senior managers of construction firms who do not know the score… until the game is over and it is too late to do much about it.
Don’t take that as a cheap shot. The fact of the matter is that most of our clients are disciplined people. They have been successful in developing strategies around getting work and executing work. However, to this point, there has been very little discipline in our industry around utilizing data and trends to our advantage — a concept we call data strategy. “Big data” is the buzzword used in today’s business world, but we are first going to focus on the power of “little data.”
Harnessing Little Data
“Big data” feels an awful lot like BIM did 10 years ago: a term (mostly) driven by software vendors to describe an important evolution in the technology that they sold. It captured the zeitgeist of the moment by suggesting, in somewhat aspirational terms, a concept of better integration, better visualization and faster decision support for the A/E/C markets. The other side of that coin, however, is that many companies were frustrated by the fact that buying software and training the latest graduates did not really translate into a successful BIM strategy.
“Big data” is also largely aspirational and, as such, we are already seeing some frustration with clients about this term. The buzzword has become so pervasive in the popular and business press that hardly a day goes by without seeing it as a focal point of breathless, somewhat unrealistic portrayals of the power of data. In the same way that we recommended that clients address BIM as a “table stakes” issue, we suggest reducing the oh-my-we-haven’t-one-anything-about-Big Data-yet anxiety with a few simple steps.
What’s the Fuss?
Let’s start with a few conceptual ideas that will hopefully give you some better perspective on what “big data” is and what it is not. Big data is often more about unconventional insights based on new and previously unavailable data. You probably have some interesting and useful data already in-house. For example, email and/or Web surfing logs (being careful to manage employee confidentiality) provide fascinating and useful data about what your workforce is working on and with whom they are communicating.
Ultimately, it is mostly about math. Don’t forget that, most of the time, big data is the result of simple statistical regression, skills that likely need to be bought or developed within. Once you find some data, ask someone with some math skills to help you think about whether it means anything.
But, even with good math, an important attribute of big data is how it is visualized and communicated. This is where statistics meets art, igniting both sides of the brain and allowing for individuals who typically stay on one side of the brain to communicate better with those who are more comfortable on the other side of brain. As Edward Tufte (recommended reading is his book “The Visual Display of Quantitative Information”) asks, “Can the same image prompt different stories and memories in different people?” The obvious answer, is “yes, it can.” Good Data Visualization enhances the effectiveness of any data insights. If people can’t understand the results, the work is a waste of time and money. What is the picture that this data could paint? What colors might call out what it means?
And, as we all think back to our first exposure to statistics, we shouldn’t forget that millions or billions of records are not necessarily required to provide value, particularly in the business side of the A/E/C space. Sometimes a few hundred or few thousand records can produce a valid result. It is all about statistical significance and the relative importance of data that represents results with a high level of confidence. This is an excellent illustration in simple terms of what we’re talking about: http://www.wikihow.com/Assess-Statistical-Significance
Lastly, we have a concept called “little data,” which we define as the cleanup process around data. Creating database structures, establishing QA methods and building good data capture processes. Polluted data, even with good math behind it, is of no use to anyone. Good little data is surely better than bad big data, and it can even be useful for your customers. A simple example might be showing a customer that it would be better off with a monthly rate on a piece of equipment over the expected life of the project rather than a weekly rate. Or an email analysis showing that your people communicate 10x more with small, less profitable customers than they do with big, important customers. Clean data, when structured well and feeding a visualization, works.
Starting the Conversation
When you start the discussion of data strategy at the executive level, have some dialogue with your leadership team on the following subjects:
- How comfortable are we on how we are betting on the future of the organization with our current strategy and insights? Are there additional data points that would give us clarity and confidence in our decision-making process?
- How are we dealing with disruptive forces (internal and external) on our business? Are we merely being reactive in trying to adjust, or do we feel like we are ahead of the curve for the most part?
- What’s really important to us? Could it be our people, our equipment, our customers, risk to the balance sheet, economic trends, operational excellence, etc.? How well do we actively monitor the health of these important aspects of our business? (Think about the benefits of proactive “health care,” not reactive and expensive “sick care,” a health care industry parallel.)
- What types of insights, with the appropriate lead times, would make us more comfortable asking the above questions? Do we think the data exists out there (albeit scattered and in different forms) that could help us be more comfortable with how we make decisions?
Where to Begin
This article aims to provide you with conceptual understanding that enables you to start thinking about a data strategy for your organization, while alleviating some stress around having to keep up with “big data.” But as with any initiative, if you want to move the chains, you will need the right leadership team and tools.
In subsequent articles, we will speak more specifically about how to evaluate systems and tools necessary to achieve your desired state, but first we want you to start thinking about foundational organizational elements and steps needed in order to begin implementing.
Identify your “champion.”
Based on FMI’s and Tilson’s experience, the hardest part of executing on your data strategy will be leadership. Clients often ask, “What is the best department to spearhead this effort; should it be business development, operations or back office?” Our answer is that it is more about finding the right individual, not department. This person would be:
- Someone with good all-around business understanding and internal influence
- Someone with the energy and reputation for follow-through
- Someone who understands your data — where it is, how to get it, and whether it requires cleaning and structuring — and navigates the internal politics
Remove the politics.
Once you identify your champion, make it clear to that individual that he or she does not work for a singular department when representing your firm’s data strategy, rather for the greater good of the organization. Unless the role is delineated (even if it means putting that person in two places on the organizational chart), the temptation may be to act more selfishly towards his or her department. One of the biggest impediments to success is when other departments begin to sense that the effort is one way — i.e., the others provide time and information and get little in return.
Treat it like a project.
Many of the reasons why contractors can build beautiful structures but have a hard time implementing internal change is how it is managed. Without a scope, schedule and budget to start with (even if done in small incremental phases), how can these efforts be successful?
Keep a list of those “million-dollar questions” that data can help you solve.
Although you may not have access to the data or it’s not structured appropriately, keeping a prioritized list of questions will help the organization stay crisp on the “why” behind the effort. Finding the person or team is likely the most stressful part. We continue to use the BIM metaphor here because it is the closest phenomenon, which we can think of, that will change the industry as big data will change it. That said, how many good candidates did you have to manage your BIM strategy and implementation 10 years ago? Was it easy to find that person? Was it (probably) a person who led the effort but brought others along?
Between now and publication of subsequent articles around specific tools, systems and case studies, be thinking about the above questions, along with the leadership and organizational elements needed to be in place to make it successful. It’s very easy to jump to decisions and discussions about the systems elements (beware of any “silver bullet” sales pitches), but there is a reason why systems implementations receive a black eye in our industry. Many of the reasons stem from root causes that we suggest you avert by thinking first about the bullets above.
Of all of the disruptive forces that will shape our industry, we believe that this will be one of the forerunners. While it will take time to play out, being an early adopter of good data practices —not necessarily big expensive systems — will only help you stand out from the others.
Mike Dow is vice president of consulting with Tilson IT Consulting. He can be reached at 720.370.5381 or via email at email@example.com. Jeremy Brown is a consultant with FMI Corporation. He can be reached at 303.398.7205 or via email at firstname.lastname@example.org.