OpenBoxMLS: An Open Source Acorn

“All of us are smarter than any one of us.” – Jim Zemlin, LINUX Foundation

Remember this day: May 14, 2019.

This is the day when BoxMLS announced OpenBoxMLS, the first real open source software initiative in real estate. I don’t want to embed the entire email, but for real estate technology geeks, these two paragraphs ought to be a jolt of energy:

San Francisco, CA—May 14, 2019: Today, BoxMLS, a real estate technology company providing search and collaboration tools, officially announced OpenBoxMLS, an open source software initiative to modernize the software available to MLSs, brokerages and technology companies in real estate.

“If we look at other industries, the future of software is open source,” said CEO Kevin Hughes. “Open source allows for collaboration and cooperation even among competitors for mutual benefit, and that’s how the real estate industry works as well. We are excited to take this step forward for an industry we are so passionate about.”

In the interest of full disclosure, while BoxMLS is not a client, I am an advisor to BoxMLS and have known Kevin Hughes, the CEO, for many years. I’ve talked to them and advised them on the open source initiative. My talk on open source software at Clareity earlier this year has apparently borne at least one shining apple.

And I could not be more thrilled, or more proud of Kevin and his team for taking this step. It is a monumental thing they did, even if it doesn’t immediately appear that way.

Now we get to see if the real estate technology space can come together and create what we all deserve, a modern, modular, plug-and-play system owned by no one and owned by everyone… or if real estate technology is doomed to incompatible silos, increasingly outdated systems, and ultimate domination by one or two major tech platforms.

Let’s get into it, shall we, with a spirit of celebration and congratulations to the team at BoxMLS.

What is OpenBoxMLS?

I’m hardly the right person to provide details about an open source software project, since I can barely keep my own website up and running without major bugs. But I figure, the real techies have already headed over to the OpenBoxMLS webpage, and then over to the Github repository to check things out for themselves. The rest of us who are business people, MLS executives, brokers and tech-savvy agents might benefit from this layman’s attempt to explain what I think OpenBoxMLS is and could be.

If I’m wrong, I’m sure someone far more technical will correct me.

Maybe it’s easier to start with what OpenBoxMLS is not.

  • It is not an MLS.
  • It is not a free and open source of MLS data.
  • It is not even, as I understand it, a full MLS system.

As the announcement puts it:

“There is quite a lot of confusion and misconception within the MLS industry about what open source is and is not,” said Hughes. He explained that under the OpenBoxMLS initiative, current industry business structure will remain intact. MLS’s, brokers, and third-party service providers will continue to own their respective proprietary data: licensing oversight, compliance, MLS rules, policies, and governance are not affected. [Emphasis mine]

So… uh… what the hell is it then?

From the website:

OpenBoxMLS is a nationwide, scalable RETS aggregation engine that outputs standardized RESO data, synced to any database. Our back-end poller provides a foundation for MLS’s and brokerages ready to own and operate their own database.

Yeah, I didn’t understand those sentences either. But I have had the advantage of talking with the team over the weeks they’ve been preparing this. So my layman’s understanding is that OpenBoxMLS is the “pipe” between the MLS database and the applications that want to use the data from said database.

The stories I have heard from real estate technology companies over the years about the difficulty of trying to create cool new applications and tools have all had an element of something like this:

“See, we have to get the data from 115 MLSs. To do that, we have to go to their MLS vendors and request a feed. Sometimes, we get sent Excel spreadsheets. We then have to take all that data, import it into our database, then normalize it — which means manipulating different datasets so that they all kind of line up. Some MLSs have one field for “Address” so it’s “1234 N. Main St., #301” while others have “Number”, “Directional”, “Street Name”, “Unit Name” so it’s “1234” then “North” then “Main St.” then “#301″. We then have to output that data so that our software can use it. It’s a major pain in the ass, and I’ve got three data guys who do nothing but take in MLS data feeds and make them usable.”

As far as I understand it, OpenBoxMLS solves this problem.

The Github repository has some more information:

  • Real Time Data
    Listing updates continuously pushed to your systems instantly. Price changes, status changes, new listings, everything – instantly. Any slower than real time and you’re losing out on opportunities and deals.
  • Compatibility
    Work within the existing Real Estate data model, compatible with all legacy MLS platforms, allowing for easy migration.
  • Extendability
    Further enrich incoming data instantly with a variety of data transformation and enrichment functions, including address validation, geolocation, travel time analysis, and listing quality scoring.
  • Freedom and Ownership
    Open BoxMLS is open-source software and that puts you in complete control. You’re free from vendor lock-in, free to use the software without restriction and for any purpose. As open software, you’ll benefit from the stability, security, and innovation that comes from other businesses and developers using and contributing to the project.
  • Flexibility
    Many products don’t integrate. Confusing software fragmentation between MLS – Broker – Agent platforms.
  • Cloud Technology
    Lower your cost of ownership and maintenance.

Okay, even I can understand those benefits. Price changes and status changes pushed to the software instantly? What is this, the 21st century? Cloud technology for lower cost of ownership? I say again, what is this, the 21st century?

The coders can tell the rest of us how exciting this is, because I can’t tell:

Simplify event delivery
Eliminate polling—and the associated cost and latency. With OpenBoxMLS, event publishers are decoupled from event subscribers using a pub/sub model and simple HTTP-based event delivery, allowing you to build scalable serverless applications, microservices, and distributed systems.

Acorn, Not the Oak Tree

The way I see it, OpenBoxMLS is an acorn, a seed. The start of something pregnant with possibilities. It is not the oak tree, not a fully functional MLS system you can go download tomorrow and run your MLS on. And I think that’s intentional on a couple of levels.

First and most obvious is that BoxMLS’s main product is an MLS front-end-of-choice that works across multiple MLSs — useful for users who have to belong to multiple MLSs. Presumably, BoxMLS would like to keep being able to sell that front-end-of-choice product. They open sourced the data utility backend that makes their front-end work, which every software company in real estate has to have, but doesn’t add much value to the actual product that customers want to purchase.

Second and less obvious but more important is that the team at BoxMLS truly understands and embraces the open source philosophy: the power of self-forming community. I know this, because I’ve had long conversations with them about it.

Absolutely essential to the open source philosophy are collaboration and community. If BoxMLS had released a full-blown MLS system, then there is less of a need for other developers, other database programmers, other coders to get involved.

I would go on and on, but since I covered almost all of these concepts in my Clareity presentation, let me embed the video I made of that presentation again for you all:

Power of self-forming communities. Code poets coming together to collaborate by insulting each other and engaging in flame wars, but backing those up with code of their own so that the whole is improved and enriched for everybody. Application developers creating their own special proprietary code for the things that really add value to the customer, rather than having to redo the grunt work that everybody has to do.

This is what BoxMLS is trying to unleash with OpenBoxMLS.

This Idea is a Long Time Coming

In putting this post together, I found something from the historical archives of this blog. It turns out, I wrote a post back in 2012 that was… somewhat skeptical of the idea of open source MLS software. And I did that in response to something written by my friend and colleague Jeff Corbett, the X Broker. Since Jeff was the first person to really introduce the idea to the industry, I think he deserves some credit and recognition for being the original visionary.

Jeff wrote a post on the RETSO blog titled, “Shift Happens: A Case For An Open Source MLS.” Go read the whole thing. The key grafs:

Central to open source projects is the idealism that no one entity owns or controls the tools. Yes, there’s money. Money is made in an open source environment by offering services around the technology.

Think: WordPress.  There is an entire economy around providing services around the platform, not by charging for the platform itself.  More on this later.

Architecting an open source MLS would require a community of willing developers.  Considering the agenda at hand, its overall potential and understanding the developer culture, I can’t help but think finding these folks wouldn’t be too difficult.   Knowledgeable peeps in the current MLS arena would need to sign on as well.  Some would probably require some compensation for their time… so a micro fund of sorts could be created to subsidize costs.  In any case, ‘Many Hands Make Light Work’.

Real, meaningful standards would be created.  Everything from data consistency and quality, to front and backend UI/UX, to eliminating the multitude of loopholes currently exploited in current MLS systems (i.e. manipulating days on market by removing and re-listing a property).

My criticism back in 2012 centered around the question of, “Who would create these real meaningful standards?” It rang true back then.

But today, we have RESO, the Real Estate Standards Organization, whose new CEO, Sam DeBord deeply understands the connection between the MLS and the brokers and agents on the ground and their consumer clients. There are dozens of companies and hundreds of people who have worked hard and collaborated (even as competitors) to create real meaningful standards.

We have had seven more years of technological progress, not just in real estate but in overall information technology, from Cloud computing to AI to Big Data — all of which rely on open source software.

The timing is right for open source software in real estate.

So, What Now? What Happens Next?

The short answer is that I don’t know. Because I’m not a coder. I’m not an engineer, not a programmer, not a data scientist.

What happens next is up to the code poets of real estate, and to a lesser extent, to the bosses and business leaders and MLS executives and brokerage owners.

If the coders and engineers and hackers who work in this crazy industry of ours come together on the base that BoxMLS has released, and start working together on common problems they all need solved, then we will have growth and energy and value. If they do not, then we won’t. Then all we will have is a nice data utility from BoxMLS.

The coders are primary, because they’re the only ones who know what problem they’re working on that the outside world simply doesn’t see and doesn’t care about. Nobody buying their tools cares whether they use AMQP or Zapier; they just care that it works, that it’s easy to use, and that it’s cost-effective. So it is these individual programmers who have to form communities to collaborate on solving common problems. They could do it in their spare time if they really wanted to.

Or those coders might go to their managers and say, “So, I could spend the next two days writing code for XYZ… or get involved with this open source project and just download what others have done, improve it, tweak it, and have it done in four hours. Your choice.” That seems like an easy decision for anybody who has to pay programmers.

And not every code poet in real estate necessarily works for a company in real estate. Who’s to say that there isn’t a Google engineer who has great ideas about speeding up property search? Who knows if some brilliant Slovakian programmer might want to get involved, with dreams of one day creating an MLS platform in his country?

So the first burden, the first responsibility, but also the first opportunity from OpenBoxMLS belongs to the coders, to the engineers, to the hackers.

But to a lesser extent, but not any less important, the business leaders have a role to play.

For one thing, they can get their entire companies and organizations involved.

Maybe someone like Corelogic or Remine would think about contributing some Add/Edit code so that an open source listing entry engine could be built. That’s a basic utility now; customers are not buying some MLS system because of basic Add/Edit module. They might buy your very cool, very slick, beautiful and easy to use UI for an Add/Edit, though. That such a thing would dramatically cut the cost of updating, maintaining, and improving a listing entry engine would not be lost on those companies.

And MLS executives, brokerage leaders, and franchise executives could support open source software initiatives because those initiatives would solve problems and save them money. (Hint: you should; you really should.) At the very least, businesspeople need to get educated on open source and what it means for their business bottom line so they can make informed decisions.

Three Things Not To Do

In closing, let me take the impertinent step of advising the non-coding leaders of the industry on three things they should not do.

One, do not write off open source software immediately because OpenBoxMLS doesn’t solve all of your problems today. It is a seed, an acorn. If the coders come together, it will grow into a mighty oak tree. At least educate yourself on what open source is, what it isn’t, what its possibilities and limitations and opportunities are. There are plenty of people far smarter than you and I on that subject, many in the industry. Ask them for their thoughts and get informed on this important development.

Two, resist the urge to be narrowly self-interested. Self interest is fine; it is actually what drives open source. There is a lot of money to be made in open source, both in cost savings and in actual business opportunities. But don’t go sabotaging shit because open source threatens some product in your lineup. Think bigger about what possible opportunities there may be if that product is open sourced.

Three, do try to avoid the urge to control this. It isn’t hyperbole to suggest that our industry sometimes acts like a bit of a control freak. We’re constantly trying to mandate this, or certify that, or create rules and committees and enforcement mechanisms and forms and hoops to jump through. Resist the urge. There will come a time when the open source software movement in real estate will need guidance, some structure, some organization. Now is not that time, because all we have today is an acorn.

For now, just liberate the code poets. Let the coders code. Let the engineers argue with hackers and let them duke it out with code. Let’s see what the power of self-forming community can do with all of the chaos that creation entails… and then let’s talk about organizing it.

Remember this day: May 14, 2019. May it go down in history as the start of something wonderful. Thanks and congratulations to Kevin and his team at BoxMLS.

Now it’s up to you, code poets of the real estate industry.

-rsh

Get the latest posts via email

7 Comments

Join the discussion and state your opinion. Some comments may be held in moderation. I try to get to them as soon as possible, but may be traveling or unable to approve comments immediately. I do not censor comments, but reserve the right to remove anything that looks like spam, trolling, or just outright inappropriate.

  1. Very cool and a next great step from CoreLogic Trestle (aggregates MLS data + proprietary) and WolfNext (https://wolfnet.com/solutions/real-estate-data/) that are closed-end.

    I’ll have to dig in but would be exciting to see BoxMLS (or whomever!) work in other API’s to really create an incredibly robust offering. Thinking on piping in agent reviews (Zillow, Yelp), photo classification (restb.ai), local city data, CMAs (Cloud CMA), document management (DocuSign, DotLoop)…so many options now that the framework is exposed.

    1. That is certainly a direction the Open Source community can take the Openbox. The possibilities are open ended.

  2. BoxMLS presented to our CRMLS focus group and we were all impressed. They are bing considered for front-end of choice. It still needs work but is definitely a breath of fresh air. In speaking with one MLS exec, I get the sense that MLS organizations are sick and tired of MLS vendors slapping lipstick on a pig and calling it good.

  3. Looking forward to seeing where this goes.

  4. The simplistic analogy is that OpenBoxMLS is like the MLS version of the ZIP-code standardization tool of the USPS. The problem is, MLS, like the USPS themselves, is so ancient and useless in the big data world that doing things like normalizing it with RESO is a waste of time given all the issues surrounding its use. A modern real estate data system like the forthcoming #OREDS is the only way out of the decades-old MLS logjam.

    It’s not necessarily true that there’s a lot of money to be made in OSS. It’s a labor of love and morals, with the thought that “software should be free” but the sticky fact that coding big systems takes a LOT of money and a LOT of frustrating coding/debugging cycles. Providing compelling differentiation, a fundamental aspect of capitalism, requires layering SOMETHING on top of OSS: Services, features, support, etc.

  5. Nerd moment…did no one notice rob’s image of Sepheroth! Nice article Mr Hahn

    1. Hehe, Sephiroth was kind of a mascot of this blog for a while 🙂 Good to see someone else notices these things. #kindredspirits

Comments are closed.