Why You Need Industry Data in Strategy Planning but Shouldn’t Trust It

Data is a critical commodity in business, and there is lots of it to be had in commercial real estate. But the abundance of information can lead to some spurious conclusions in strategic planning.

You regularly see studies looking at vacancy, occupancy, rents, prices, cap rates, and more in CRE. Companies large and small with significant amounts of data to be had run reports and analyses. Executives are supposed to use the information to make smarter and more effective decisions.

The vendors making analyses of their data available are likely working with the best of intentions. But there are inherent problems. In recent interviews about conditions in certain sectors—office is a great example—many professionals have lately been telling GlobeSt.com that they’re questioning some of the numbers they’re hearing. For example, a good number are certain that the occupancy rates they hear of people actually in offices to work seem significantly higher than what they’ve experienced.

Data is great, but is subject to that statistical bad boy, sampling. The mathematics of statistics depend on getting large enough random samples from a body of examples. Even then, the results may be off, but that error can typically be estimated to fall within a certain range.

The problems arise when the samples aren’t truly random. Then they can get squirrely, like political polls have been during recent UR presidential elections. The numbers suggest candidate A will win, and all of a sudden B takes the vote.

When a sample isn’t truly random, like when one group of voters is more likely to refuse to be surveyed, or will possibly give false answers, the responses of the survey aren’t going to represent what’s really going on.

The same thing happens when pulling in data from building operations, transactions, and so on. Maybe a company has data that’s largely on the biggest CRE operators in a sector. They’re important, but they probably don’t own a majority of all the properties. Suddenly, there’s a large group of smaller owners and operators who represent the bulk of the industry who aren’t being included. It’s called a self-selected sample.

Or the type of data may not be accurate. Looking at rents in multifamily, for instance, takes on a different sense if the figures are gathered from advertised rates but the collection of data never had access to the rates in the singed leases.

There is no such thing as perfect data in business. You’re restricted to what someone could gain access to, and the amount of time and money they had to collect information and analyze it. And yet, working without data is walking about with your eyes and ears covered.

The solution is to make allowances for how far off things could be. Do scenario planning, seeing what results might be if one factor is under or over what it says, then seeing what happens if the next assumption varies. It is time consumer and far from a perfect approach, but it can aid in better risk management than tossing the advantages modern data collection and analysis can provide.