Decision 2020 - A Revenue Management Parable

It's election week, and there is an air of limbo in the country as we wait for the result and generally try to make sense of one of the strangest election cycles in memory. But when the no-doubt extensive obituaries of "Decision 2020" come to be written, the question most will have to attempt to answer is this: "What the **** happened with the polls?"

The inaccuracy of polls leading up to this week's election was nothing short of astonishing, particularly those taken in the all-important battleground states. Generous predictions for Joe Biden evaporated, with huge polling leads in states like Michigan and Wisconsin translating to the narrowest of wins, or narrow Biden leads in states like Ohio and Florida turning into comfortable Trump holds. Similarly, Senate seats that were "in play" according to the polls appear so far to have attracted vastly more money than votes.

The non-performance of reputable polls in key races has left even the most seasoned political analysts asking where polling goes from here. As we wonder what polling data will be useful for in future elections, we are reminded of another ubiquitous but potentially misleading form of market intelligence: pricing data. Multifamily revenue management decisions are frequently guided, or at least heavily influenced by competitor pricing data. But whether the data improves the quality of the decisions is far from clear.

When the data gets it wrong

Don't get me wrong; local market conditions are a vital input to pricing, and the movement of competitor prices is contextually important. But where we repeatedly see revenue managers going wrong is with their understanding of the data and the types of decisions that the data supports. Here are just a few examples of the challenges of using data to drive multifamily pricing decisions:

• Concessions: While listing data (i.e., pricing data from listings on ILSs) is abundant, it does not provide insight into concessions. With concessions now commonplace in many markets, it is easy to see how misleading this data can be.
• Comp Availability: A good set of comps needs to provide a few readily substitutable properties, i.e., close and roughly similar properties. There are many submarkets where there simply aren't enough properties to meet this threshold, or the data vendor does not have access to the ones that you need. Bulking out the comp set with dissimilar properties degrades the quality of the data.
• Unit Availability: As competitor unit availability fluctuates, the comp rents can vary wildly, creating an unrealistically volatile view of pricing.
• Everybody Lies: We all know that properties mislead data companies about their rates and availability, in much the same way that voters apparently lie to pollsters! This and other poor comp shopping practices also compromise the data.
• Comp Set Error: Picking the wrong comps or using inappropriate comp weightings in the revenue management system (RMS) can badly compromise results.

How to get it right

When we think about how we should use comp data, it's important to remember that one size does not fit all. Multifamily communities vary in the extent to which they should treat comp pricing as a factor. There are many cases where following prevailing market pricing trends does more harm than good (especially if the data isn't an accurate reflection of what is really happening in the market).

"Trust the system" is a bad strategy, for example, when the RMS is using wildly fluctuating data to drive pricing decisions. Here are a couple of things that we recommend to all revenue managers trying to put pricing data to make the best possible decisions:

• Get a detailed understanding of how your RMS handles competitor pricing and the relevant system parameter settings available to you. Different RMSs use comp data VERY differently, and they also have varying options and capabilities for dealing with the comp data issues discussed above.
• DON'T Set it and forget it. Markets change, as do property objectives, exposure levels, and so on. If comp pricing is a driver today, it may not be tomorrow! Assess and shop your comps regularly, and make sure that the RMS settings that you can change are set correctly for each property. 

Above all, as we have just learned with polling data, you must not trust it blindly. Understand the data, where it comes from, how reliable it is, and use that to determine which decisions you should and shouldn't be using it to make. This election cycle should make us forever skeptical of absurdly optimistic polling and its potentially nefarious use in fund-raising. Apply the same level of skepticism to competitor data, and your pricing decisions should improve too.

Photo by Element5 Digital on Unsplash