Have MLS membership levels hit bottom or is there still shake-out to come?

Jun 5, 2009  |  Michael Wurzer

Inspired by a recent post on Rain City Guide called The Third Bubble by Chuck Reilling, we ran some numbers across our 100 or so MLSs from across the country and came up with the following sides per agent trend graph:

Unlike Chuck’s graphs, this graph shows a significant flattening over the last two and a half years.  I’m interpreting that as indicating the amount of membership decline still in the pipeline may be pretty low.  Or is that just me being too optimistic?  What do you think?

7 Responses to “Have MLS membership levels hit bottom or is there still shake-out to come?”

  1. Jeff Allen says:

    Interesting! I bet you have a nice treasure trove of data to pull from.

    While the transaction sides may be levelling, can you run these numbers with dollar volume per agent? I would imagine that that metric is still trending down as more of what’s being sold is low-priced foreclosures and short sales. And ultimately, that’s a purer reflection of the money in agents’ pockets.

  2. Jeff, we did run the sales volume numbers, too, but ran into a few anomalies we haven’t quite worked out yet; we’ll post them as soon as we do.

  3. I sure hope you’re right Michael. Do you the data to be able to trend membership levels within the same 100 MLSs? That would be very interesting to look at. We’re going to be conducting a MLS Operations Benchmarking Survey on behalf of CMLS in a few weeks and we’re going to ask the membership question there. We’ll be sure to share it with you when we get the results.

    Also, if there are any questions you would like us to ask in that survey, feel free to throw them my way. The survey will be distributed to every MLS in the United States.

    Marilyn Wilson

  4. Very interesting stats; thanks! Would be very interested in seeing a chart of agent population for that 100 MLSs. We generally think that our Seattle/Bellevue market has lagged a lot of the country by about 2 years, and I would exoect our NWMLS agent population to drop more this year than last. I wonder what the 100-MLS pattern was of when the agent population started to drop vs when the price curve started to drop.

  5. Kristen Carr says:

    Great news, great post!!! I would guess a couple of things…

    The areas who saw the biggest growth in prices also saw the biggest subscriber growth (duh…right?). RMLS FL was at about 7000 agents when I moved to Florida 8 years ago, topped out just below 19,000 and is at about 11,500 right now. I’m not sure what the month-to-month stats are but I know in January and February they were gaining numbers rather than losing. While I worked there our shareholder associations (3 total) were projecting losses of 20% to 30% but, during those years (until January 2008) only saw a drop of about 15%. Last year was far worse but this year seems to be leveling out.

    Too bad there isn’t a way to see how many dropped agents were part-time. I believe retirement markets, like FL, NV, AZ, had a larger number of retirees jumping in to try to earn a fast buck!

    Finally, as someone who has worked in the support arena – dropping off the lowest denominator doesn’t always have to be a bad thing. The agents who do not work as frequently tend to need the most help. They fill the call queues, seats in classes and demand the most attention from their brokers. Some don’t have the drive, some just don’t have the skill. Dropping some of that weight can be good for the rest of the population. Just like wildfires – some of the old wood needs to be burned in order for growth.

  6. Matthew O'Brien says:

    This graph represents two variables, sides and agents, and you’ve drawn a conclusion about one of the variables. I wouldn’t recommend that approach. A graph of the ratios of two independent* variables only tells you things about the ratios. The sides:agents ratios 8:1, 160:20, and 4 billion:500 million (taken in any order) could be the raw data for years 2007 through 2009. Realistically, none of those are near the actual figures, but it illustrates the point.

    You may be right that membership levels are ready to level off, but that conclusion isn’t supported by the data you’ve provided in your post. Raw side and/or agent count data would support more conclusions. If you want to protect the raw numbers, you could construct a graph normalized to 2003’s data.

    * I use the term “independent” loosely here, not in the formal statistical sense.

  7. You guys all make some great points. My approach is flawed in data and theory. Back to the drawing board!