Is there a national MLS?

MLSs in the US are like these fishing lures

Running a real estate API company I’ll often get this common question/request from eager real estate technology startups about a National MLS.

“Hi we’re trying to create a real estate portal, how can we get access to all of the listings in the United States. Can your API provide us with this data?”

If you’re not familiar with the real estate industry in the United States, it doesn’t sound like an unreasonable request. In 2016 there HAS to be a way to get all of the listings from a single source right? So does a national MLS exist?

No. There’s no nation-wide MLS or central database for real estate listings in the United States or Canada. A national MLS probably won’t happen anytime soon. There are between 700–800 MLSs in the United States and no national MLS or central data source for all of those listings. To get all of the listings in the United States, you’ll have to contact the 700-800 MLSs in the US one by one.

There’s no nation-wide MLS or central database for real estate listings in the United States or Canada.

No National MLS? That’s crazy! Why is it setup that way?

It boils down to two things (aside from politics), rules & regulations of private real estate data, and schema.

There’s been some progress made towards consolidating MLSs into larger regional databases thanks to attrition. It’s harder for smaller real estate associations to keep up with the cost of maintaining their own systems, so there’s been a trend of larger MLSs absorbing smaller MLSs to keep them from failing.

However even this level of consolidation is going to take time because each MLS has their own rules for data licensing and their own fee structures. Every MLS requires a data licensing agreement before you can access and distribute the real estate listing data. This agreement usually outlines the following:

These conditions vary from MLS to MLS. One MLS might only allow active listings for public display, while an different MLS might be okay with active & pending listings. Also the fees can vary wildly as well. Some MLSs charge a one-time fee, others monthly (that might scale depending on the # of customers you add), or a combination of both.

If you’re a vendor, you must work on behalf of a member of the MLS (brokerage, agent). In addition to signing a data licensing agreement, you will usually have to have a brokerage or other member of the MLS sign the data licensing agreement as well. To get access you’ll need to be working with a brokerage or agent already before an MLS will even consider giving you access to the real estate data. MLSs don’t want to see their data distributed to third parties without oversight. Afterall, much of this data includes private addresses, contact information, public and private remarks and other sensitive info that could cause real harm if distributed in the open. They have a legal requirement to protect the data.

Another block to national MLS listings is due to data schemas. Each MLS organizes their data in different schemas; fields for listings, photos, and other data stored in the MLS can vary widely. Thankfully the industry is starting to move towards a common schema called Data Dictionary (www.reso.org), however with 700+ MLSs it will take a while for the Data Dictionary standard to become widely adopted.

What about ListHub? Realtor.com? Zillow? They have national data right?

ListHub has a lot of data, about 600 MLSs but only if brokers explicitly choose to share their listings. ListHub will charge you as a publisher $1000/mo per MLS (capped at 6 MLS). Making this a pretty expensive option, however the feed is in a common schema. Aside from the cost there are two downsides, one is that you’ll still have to approach each brokerage to convince them to syndicate listings to your application. Second, while 600 MLSs is a lot, there are huge gaps in coverage (for instance only 30% of Chicago listings are on ListHub).

Realtor.com probably has the most listings (800 MLSs), but they only keep the data for their site; their parent company Move, Inc (now owned by News Corp) does not distribute the data. Realtor.com has been around since 1995 and has had connections w/ MLSs stretching back before some Y-combinator startup founders were even born.

Zillow used to source their data from ListHub, now they’ve had to go to each MLS directly for data since they ended their contract w/ Move Inc in April 2015. They’ve signed on quite a few MLSs including Georgia MLS, but they don’t have national coverage.

What about Upstream / Advanced Multi-List Platform (AMP)?

Upstream & AMP are two technologies being developed that could lead to the closest thing to a national MLS database (if they succeed). The projects are backed by the NAR and are being developed in a partnership with RPR and a collection of national brokerage franchises. The idea is a developer can build an application that ties in with Upstream/AMP, getting access to an MLS still involves working out a deal with the MLS, but the process is streamlined and managed through the connection with Upstream. This is pretty close to the model ListHub uses.

Upstream/AMP is in private beta right now, but by the end of the 2017 it could be available to the public. Still it may take a while to get fully to market. Upstream is shooting for 50% of national listings in five years. That’s a lifetime in tech. There’s still a question of fee structure and how much vendors will have to pay for access. I’ll post an update when I know more about how the system will work (hopefully I’ll get a login myself).

So my app/startup ideal is dead on arrival?

Not necessarily, just because there’s no easy way to get national MLS data that doesn’t mean you can’t build a great business in the real estate industry. It just requires a bit of strategy.

The advice I give startups is to build their MVP first in a large real estate market. Considering the work and cost it takes to get access to data to an MLS you’re better off concentrating on a larger market with over 100K listings rather than trying to scoop up a lot of smaller MLSs. California Regional Multiple Listing Service (CRMLS), Arizona Regional Multiple Listing Service (ARMLS), and North Texas Information (NTREIS) all have 100K+ listings and lots of agents/brokerages you can serve.

The advice I give startups is to build their MVP first in a large real estate market.

A second option is to establish a good relationship with the MLSs closest to you and the brokerages/agents in that MLS. Some MLSs are more favorable to startups/vendors than others, but if you’ve got a compelling service that you can demonstrate would be a benefit to the members of an MLS you should be able to get access.

So how does Rets Rabbit get the data? We get the data from either brokerages or customers who have their own RETS access to their MLS, or we’ve secured a data license because we’re working with a specific brokerage. We may have the data for the MLS you’re looking for, but per the MLS licensing rules we can’t give that out to a third party.

What do I do once I get access?

Once you get RETS access in a big market (the common standard for connecting and pulling real estate data from MLSs) we can import your listings and serve them through the Rets Rabbit API. We’ve got experience working with almost all of the major MLS software providers and we can save you thousands of dollars and months of effort on data importing. We can even normalize listings across MLSs to the Data Dictionary standard and geocode listings if that’s not provided by your MLS.

I hope I’ve given you a clear understanding of the current state of the real estate tech industry. There’s a lot to be hopeful about I think, but it’s still a challenging market. The good news is that the barriers to entry are high enough that the casual wanna-be founders stay out.

If you have any more questions about contact me! I’m happy to answer any questions you might have about real estate tech and MLSs.