Are Competitive MLSs Possible or Desireable?

Oct 21, 2007  |  Michael Wurzer

This post really is a look back at what I think are some of the key posts from the FBS Blog, with the hope that I can make everything a bit more cogent and focus in on the strengths and weaknesses of the case for competitive MLSs in contrast to monopolistic MLSs.

First, let me start with a more recent post, The Future of MLS Is Now, which makes the case that this time in the history of the MLS and the web is crucial, that the rules of community and engagement are being drawn right now, and we can either participate in this defining time or not. Importantly, the web is not just redefining real estate but much of society, or how we interact with each other.

This latter point is made robustly in You Must Read This, If You’re Interested In The Future of MLS, in which I apply Clay Shirky’s excellent post A Group Is Its Own Worst Enemy to MLS and conclude: “I truly believe th[ese patterns for success in social software are] the framework for reinventing the constitutions of the MLSs. The seeds are here. We need to grow them. Now.”

The design patterns for social software outlined by Clay Shirky were not about MLS, but they resonated with me because they were so similar to my post, MLS Is More Than Technology, which posits that the genius of MLS is bringing competitors together to cooperate just enough to aggregate listings together, which undeniably serves consumers in important ways. Thus, the question becomes, as the rules for MLS are reinvented, can it be done without destroying the cooperation that allowed for the listing aggregation in the first instance?

There are two primary fronts on which listing aggregations (MLSs) are being attacked today: (1) demands that MLS aggregations get bigger and cover more territory; and (2) demands that individuals within the MLS need more control over the aggregation to do with it what they want. As you’ll see below, these issues are inextricably intertwined, but have a common solution, focused on data and rule standards coupled with competition.

The former challenge to MLSs was addressed in my early post Raging Regionals and my suggested solution was presented in Regionals, Part II. The quick summary is:

  • The demand for regionalization is based on real pain (inefficiency) being experienced by regional brokers dealing with multiple MLSs having disparate rules and data sets across many boundaries.
  • Forming regional MLSs, however, only moves the boundary and shifts the pain, at great cost in mashing MLSs together.
  • A more efficient approach would focus on national standards, which already is occurring for data with RETS. Another way of thinking about standards is that it’s just like forming a national MLS, but without all the challenges of trying to merge MLSs together. Call it a virtual MLS, if you will. The big benefits of this approach are: (1) a lot of work has already been done and focusing more attention and energy on RETS at this time will bring that work to fruition ; and (2) focusing on national standards provides a much more definitive, long-term solution than inching along painfully with forming many regional systems that will still have boundaries.
  • Once deep and broad national standards are developed, there are many possible ways of making use of them, and the focus should be on encouraging competition to produce the best solutions. Leveraging the idea of a national repository, I suggested in Regionals, Part II that one use of national standards would be to provide a path for existing MLSs to populate a national repository, which, in turn, could be used to allow the local MLSs to compete with each other if reciprocal rights of download were part of the repository agreement. In other words, if an MLS contributes data to the repository, then it also can download whatever slice of data it wants from the repository. In this way, each local MLS can try to serve as big or small of a territory as it wants and individual brokers will be able to choose from the one that serves them best, knowing that the data they enter once will end up in the national repository and available to all MLSs.

I still think this is a good solution that would leverage existing resources to solve the real pains of brokers dealing with overlapping markets of MLSs with the long-term benefits of a competitive market for MLS services.

At the same time, as I’ve thought more about it, there clearly are gaps in the proposal. First, what is the definition of an “MLS” that would entitle it to mutual download. Taken to the extreme, could an “MLS” be a single broker such that if they contribute one listing they are then entitled download the entire national repository? If so, what about buyer brokers who do not have listings? How does this right of reciprocal download relate to IDX or advertising? What kinds of things can the “MLS” do with the data when it is downloaded?

These questions quickly return us to the beginning or the heart of the issues at hand, namely the blurring between MLS aggregations and listing advertising. Importantly, as discussed in MLS Is More Than Technology, listing aggregation (a key purpose for MLS) is different than advertising. If the only issue is where to advertise listings (e.g., Google, Trulia, Zillow, etc.), competitive brokers naturally will choose different outlets, resulting in no complete aggregation and a dis-service to consumers as well as brokers. This is the fundamental reason that claims of the “death of the MLS” miss the mark. Without the cooperation engendered by the MLS, brokers and agents would advertise their listings here, there and everywhere, but not all in one place like they do now with the MLS.

The issues surrounding listing advertising were further explored in Listings, Leads and Losers, Oh My!, in which I concluded: “The value proposition seems clear: (1) consumers need listings to learn more about the market; (2) brokers can supply the listings; and (3) agents can help educate the consumers on what the listings mean. The technology needs to facilitate this value proposition in as complete and unobtrusive manner as possible. If we can do that, there need be no losers from listings or leads.”

All of this leads me to conclude that an important part of the solution is to focus on the “terms of use” upon which brokers are willing to allow use of their data by others. I broached this issue in Is Data Free? Or Is There No Free Lunch, by examining the terms on which Google allows others to use their data. In focusing on these “terms of use”, distinguishing the who, what, when and where is very important. Who is going to use the data for what and when and where?

The biggest challenge here is that a collective “terms of use” is important, but collective agreements are not appreciated by the Department of Justice. However, even the DOJ recognizes, that justice requires balancing the benefits to consumers of listing aggregation against any harm from collective agreements. In this case, if the DOJ were engaged in the discussion on what the terms of use should be, perhaps that balance can be struck from the beginning.

That balance is required takes us back to the beginning of this post, where I mentioned the Future of the MLS Is Now and the patterns for developing social software discussed in Clay Shirky’s A Group Is Its Own Worst Enemy. MLS was born of the benefits of competitors working together. The model proved so successful that it has grown and grown. The challenge of that growth is described well by Shirky:

And, finally, you have to find a way to spare the group from scale. Scale alone kills conversations, because conversations require dense two-way conversations. In conversational contexts, Metcalfe’s law is a drag. The fact that the amount of two-way connections you have to support goes up with the square of the users means that the density of conversation falls off very fast as the system scales even a little bit. You have to have some way to let users hang onto the less is more pattern, in order to keep associated with one another.

MLS currently is pushing the boundaries to where “scale” may kill the conversation. You see this same pattern with the success of a site like Facebook, which is all about giving individuals control over who they talk to and when. Facebook is a walled garden much like the MLS. I made this case in my post MLS 7.0, which suggests that the MLS is one of the first very successful social networks. So, today, the participants are defining the boundaries of that social network and yet the discussion is so much about the technology and so little about the substance of defining the rules on which the competitors are going to agree to work together.

To address this gap, I started up the Future of MLS Wiki, hoping a discussion about the “terms of use” on which data could and should be exchanged would ensue. Unfortunately, with other events intervening, I haven’t given it the time it needs and the main participants so far have not been the brokers that need to be in the conversation. Anyway, the wiki remains and perhaps the conversation will take off with a little love and attention. There is no doubt that the issues raised are what need attention, much more than the monopolistic power plays currently gaining attention.

20 Responses to “Are Competitive MLSs Possible or Desireable?”

  1. […] FBS Blog Conversations about the MLS industry, creating software, and employee ownership. « Are Competitive MLSs Possible or Desireable? […]

  2. David Harris says:


    FMLS has been in a competitive situation with Georgia MLS for almost all of our 50 years of existence. I can attest to the fact that we are constantly looking for “points of differentiation” between us and them. This inevitably leads to a broader choice of tools and services for our agent/broker members.

    On the other hand, we are constantly reminded from many of our members that the two MLS’s cause them double work, but not necessarily double exposure.

    So, like everything else, you are in a balancing act between the pros and cons of choices you make. Where some see variety, others see complexity…


  3. David, are listings entered in one of the MLS automatically updated in the other? Or do those belonging to both having to double-enter?

  4. David Harris says:

    Ah, the $64 question. The answer is… they have both options. Due to the benefits of the RETS Update Transaction we have a third party company that sells software to brokers to allow them to enter in both services simultaneously. This software has made many decisions on behalf of the agents on how to map non-exact matching fields, so many agents/brokers chose to enter the listings separately to retain full control of their data.

  5. So, the $64 million question is whether the new RETS schema will be broad and deep enough to allow someone to develop an independent listing management tool that could syndicate listings out to any RETS compliant MLS. If not, what’s needed in the schema to make that possible?

  6. Put another way, I’m wondering if “blind” update is possible, where the RETS schema is broad and deep enough that the client need not know all or any of the business rules of the MLS other than what is described in the standard itself and could still get a listing validated?

  7. David Harris says:

    unfortunately, most of the arbitrary differences have been mitigated. The ones that remain are either deeply tied to important business situations, or even worse, local regulations.

    For example, we are forbidden in Georgia from discussing or supplying square footage, thus when we transfer a listing to a system that “requires” sq ft, our listing would be rejected.

    Data mapping issues can become further complicated when integrated with the different stages of a listing. For example, we do not allow a listing to move to pending or “under contract” status unless the agent supplies a “proposed closing date”. Thus, an external, previously entered listing must already conform to these rules to be allowed into the system.

    While I agree that many rules are instituted to reflect the business at that point in time must be revised as times change, we have the rules we have to provide the cleanest, most accurate, up to date data as possible. Because without trust in that, the MLS serves no other purpose.

    Many of the consumer facing real estate data providers don’t typically adhere to the same level of data integrity, happy to supply the data they “poached” weeks ago along with a hefty disclaimer all for the goal of a quick ad click or impression…

    wow, thanks for letting me vent on your blog 🙂

  8. Just because you’re prohibited from discussing square footage in Georgia doesn’t mean your users couldn’t benefit from entering that data for transport to other systems. Similarly, if a “proposed closing date” is important, why can’t that be part of the national standard?

  9. David Harris says:

    How far should the data set be expanded beyond those fields that your local market needs? Should I allow “really big” to be entered in a sq ft field that I will not use? Whose rules on these “external system fields” shall I support/enforce? Should I store values of 1500 miles for “distance from valcano” so I can integrate with Hawaii? 🙂 (Just using extreme examples to indicate the issues that must be addressed)

    The proposed closing date isn’t an issue of inclusion, more of an issue of validation/enforcement. If their are differences in rules, how do you compromise?

    I am enjoying our public conversation on this subject. We have discussed these issues in our RETS sessions for many many years, and I am always open to new ideas…

  10. I’m enjoying it, too.

    No, “really big” should not be allowed let alone required. Does anyone disagree? Let them speak. (Silence.)

    On distance from volcano, no, that can be derived from our lat/lon. 😉

    I’m addressing the specifics only because I can, but I realize there are some tricky spots here.

    My point, though, is that I wonder about the goal of the update transaction seeking to discover every rule from every MLS and whether that jives with the reality we’re facing now or certainly soon with trying to aggregate data widely. It seems we should be defining a standard that is wide and deep enough to allow every MLS to accept a listing that follows that standard, and wash over these local issues at least to that extent. There very well may need to be a different “class” of listings for “less validated” or “unvalidated” or something like that, but I think we can and should get as close as possible to allow for universal entry. This is what our customers are demanding.

  11. David Harris says:

    It really comes down to the basic purpose that the MLS is designed to serve…
    1) A highly efficient marketplace focused on getting the word out on a house for sale (with enough data to attract the right buyer) or
    2) an accurate property database with accurate sales histories and descriptions

    Most MLS operators are in a daily struggle with these two (somewhat) separate purposes. I know agents are frustrated when they take a client to a 4 bedroom house that really has 2 bedrooms and 2 storage rooms in the basement (closet rules not withstanding); or when a buyer wants in a top school district, and half the “affordable” homes in that district are actually outside of it. But for every one of those calls we have calls from agents upset that they house has to be listed as 20+ years old even though it has had a huge renovation last year.

  12. Kristen Carr says:

    I have a few things to add to this discussion:

    For more than 10 years, Regional MLS (RMLS) in Florida has been exchanging MLS listing content with the Southeast Florida MLS (SEF). At first the process was faulty and expensive. We used different MLS systems and have different business rules. Even though we are geographicly next to each other, we do things a bit differently.

    Two years ago RMLS made a vendor change and selected the same MLS vendor SEF uses. This change made the delivery of data a little simpler but didn’t fix the business rules issues. We have successfully mapped about 99% of our RESIDENTIAL data which I think is a pretty good match.

    Agents/Brokers who belong to SEF and to RMLS can access ON and OFF MARKET listings from Miami all the way north through St Lucie county (which is only about an hour from Orlando). Our inventory is doubled because of this. Collectively we service about 50k agents. The SEF MLS is made up of 5 Realtors Associations while RMLS is owned by 3 shareholder associations. We have, in our data share agreement, rules stating the nonnative MLS can not distribute listing content to third parties (meaning we can’t ship SEF data to but we are working on allowing a broker who participates in both MLS’s to obtain one data access for IDX, etc. This, again, is a business practice, not a technological limitation.

    Fortunately for our subscribers, RMLS and SEF have worked very well together. We don’t view each other as competition. We don’t try to “steal” subscribers. We have actually found the number of subscribers hasn’t really changed, people pay fees in the area in which they live generally. RMLS and SEF offer some different products/services. We don’t do everything the same but we work very, very well in trying to provide the best answers for our subscribers.

    With some of the new products out there for managing data via RETS we may be able to take big, huge leaps in what we’ve already managed to do. We may not have to think about physically shipping the data back and forth, worrying about whether an update will go to the other system. We’re all very excited about some of these options. That said, we have been doing this for more than 10 years. The panic and controversy have been the talk of the conventions for 2 or 3 years but we can tell you it is very manageable. There is a heck of a lot more involved in the politics of the whole thing rather than the technology. I’m sending this post to some of the people in SEF who may also want to comment.

    Mike – thanks for giving space to air these things. It’s nice to be able to voice what is happening in this industry which is RIGHT rather than the fear so many people like to spread!

  13. Kristen, thanks so much for commenting. The background you provide is excellent. I think most interesting are the terms of use you’ve established to enable the data sharing. Rules like that will be critical to a national repository.

    I’m also curious about your statement about not shipping data back and forth. Is this an indication that you all are thinking about a regional common back-end? With what front-end? 😉

  14. Brian Bell says:

    Nice reading.
    Our question as the heads of the few largest MLS Operators in North Carolina meet next week, is regarding the standardization of rules and compliance.
    With each MLS Operator having its own set of rules and regulations, how is that best and most efficiently accomplished (best case practice) in order to make all parties happy on a state-wide level ? How is that “policed” ? How is that best accomplished, and did that make any sense at all ?

  15. Brian, I think rule standardization is the next frontier that needs to be addressed today. I know NCREX in California has gone through this effort and so perhaps talking with Peter Spicer might help.

    From my perspective, I think the first step is to determine what business rules are different for substantive reasons and which are not. Many times rules are just different because of history, or, “that’s the way it’s been.” With those separated, moving towards “best practices” may be easier. My suggestion is to use a wiki to document the issues as you go and allow those interested to participate in the discussions so that the need and reasons for the changes are clear to everyone.

    Also, I think RESO could move into a “best practices” set of business rules as well. Some of this is already occurring with payloads, which are describing important business rules like status, property type, etc., and a logical extension is to move into this area next.

  16. Kristen Carr says:

    Thanks Mike!

    To answer your question, we are not considering a common backend. We are considering a hybrid solution. It has served us very well to have our own systems. We’re looking at feeding our information into a middle tier and running ancillary services from there. We’d each still have our own unique MLS system but could offer any addition products running off that.

    For instance – the idea of “bring your own front end” was something I learned as a Petchling. Way back when, 8 or 9 years ago, Gregg talked about the idea of a single database and users could select the client application. I love that concept. I talked about it with an MLS provider probably 4 years ago and was laughed at. Oh, times are a changin! How great would it be for large brokerages to be able to plug their own software into the MLS system. In my opinion, the part which is currently available to all of us is the query. Update is coming but realistically is not here yet. I have also had the unpopular belief that sometimes it may be better for a company to not try to do too many pieces of the puzzle. Putting all your eggs in one basket is a bit scary to me. In a perfect world we could use whatever system we wanted for UPDATE, feed everything into an awesome database and let our users select what they use for pulling the data out. I have always been a huge fan of FBS as a company, wouldn’t it be nice if we could offer Flex, MLXchange, Rapattoni? Our agents would have lots of freedom.

    I’m sure you’re aware, the Florida Association of Realtors has offered a statewide MLS solution for many years. This past January their BOD voted to offer it to all FAR members for free. Several MLS’s signed on and are contributing their listing content. There are still so many things to think about, work through, etc. There are a few technology issues which could be improved upon but mostly there are many business and politcal issues.

    I’m excited to see what’s popping up. It’s like finally, finally people are coming together and the answers are right on the horizon. What made people laugh just a few years ago is getting people fired up to pull together and do great things. One of my favorite things about RETS meetings is seeing people who, in other circumstances, would be considered competitors collaborating, helping each other.

    I have many strong opinions on this.

  17. I agree that enabling choice to agents is important, and I think RETS could be at the center of that opportunity. However, I still have questions about the repository approach, primarily because there are so many pieces of data RETS does not handle and I wonder if innovation will ever allow a repository to keep up. This is why I’ve been thinking more and more about the universal property ID concept and blind listing entry.

  18. Brian Bell says:

    The AE group from the largest Association Offices in North Carolina met in Raleigh yesterday. We had a very positive meeting and are now moving forward with discussions in regards to a statewide data exchange…. Fast Forward… Times are changing indeed….. With good times ahead.. !!

  19. […] course, these exact same issues have been and are being confronted by the real estate industry.  Interestingly enough, however, the real estate industry is well ahead on the path toward data […]

  20. […] As Jim notes, the ideas really aren’t new.  In fact, I’ve been writing about the future of MLS here on the FBS Blog for some […]