Michael Wurzer My last two posts (here and here) have been about the draft Listing Exchange and Access Policy (LEAP) being proposed to the NAR by CMLS. This post is a continuation of that discussion and makes the point that in its current draft form, LEAP doesn’t yet distinguish between display and other use cases for the data, including creation of derivative works like valuations, new types of media, or other products.
First, I’m unclear whether the LEAP draft proposal is complete or not, and I certainly don’t want to be critical of something still being worked on. So, if you’re one of the people working on LEAP, please know that I’m writing with the goal of improving the draft. With that in mind, in a section called “Uses”, LEAP provides:
Associations of REALTORS® and their multiple listing services shall enable MLS
Participants to display aggregated MLS listing content by electronic means in
accordance with this policy. MLSs may not impose any limitations, restrictions or
conditions on the use or display of the MLS listing content other than as specified in this
policy.
Taking this language at face value, MLSs may not impose any limitations or restrictions on Participants’ use of the MLS compilation other than with regard to display purposes as outlined in the document. There literally are no provisions I can find dealing with any uses other than display, which then leaves the catch-all prohibition against MLSs adding other terms to govern and that means there can be no restrictions on uses other than display. I’m hopeful this is because it’s still in draft and, as the draft already has a section called “Display”, I’m hoping they’ll add another section called “Other Uses” as well outlining specific terms for other uses.
Detailing how the MLS content can be used and the license terms for such uses really is critically important. Here’s my favorite example of why such provisions are important. It’s a press release from Zillow announcing a material improvement in their Zestimate derived from machine learning using photos. Where did you think the majority of those photos came from? How much do you think Zillow paid for the right to create derivative works from those photos, even before it became a Participant entitled to IDX and VOW feeds? Now that the Zestimate is a critical part of Zillow’s iBuyer program, how valuable do you think the improvement was?
One counter to this point might be, hey, now that Zillow is a Participant, they should get the MLS content for free just like every other Participant. What I’m suggesting is that no one should get the data for free and, most importantly, the value of the data for creating derivative works is entirely different than the value for display purposes. I’m sure we’ve all heard how data (and media are data) is the new oil, and that’s because it powers new products created by machine learning. Without data (especially trained, structured, and contextual data), machine learning is not possible.
I also can hear Participants saying, hey, this is OUR data, we should get it back for free. The problem with this is that there’s a huge difference in value between one broker’s listings and the collected listings of all brokers (i.e., the MLS). So, no, licensing the MLS data is NOT the same as licensing a single broker’s data. Not the same thing at all. This also is where the argument about who owns “the MLS data” comes full circle and consumes those who argue it should be provided back to Participants for free or a nominal cost. If that premise is held true, then every Participant basically is saying, yeah, I don’t care how much value other companies like the consumer portals generate from the collected MLS data, I just want it to be free for everyone.
Not only does this argument wildly undervalue the MLS data itself, it breaks the value chain of cooperation, because, sooner or later, Participants are going to stop cooperating when the value differences reveal themselves and the MLS will be accused of “giving away” all of our data. Have you ever heard that before? This is why I’m advocating that NOW is the time to start licensing the MLS content for the value it has and that such licensing terms need to be targeted at specific use cases and the value of such use cases.