Showingly Transforms Real Estate with MongoDB Atlas and MongoDB Realm

Buying or selling a house is difficult. There are more steps than you could imagine, and each one feels harder than it should be. The improvements that technology has made in the past decades seem to have passed the industry by.

Showingly is trying to make the process less difficult, starting by making it easier to see houses you might want to buy. Buyers can browse listings and book showings with no fuss. Sellers and their agents can make listings available on the app, making it easier for buyers to find them. Agents and other real estate professionals can simplify their workflows.

Showingly built its full stack on MongoDB, using MongoDB Atlas and MongoDB Realm. I caught up with Andrew Coca, Co-Founder and President, to learn more.

Tell us a little bit about your company.

Showingly is a real estate platform for buyers, sellers, and professionals. Of course, it does everything you would expect: sellers can have their house listed on the platform, buyers can browse listings and see all of the important details, agents can manage their listings, and so on.

What sets Showingly apart is that consumers can actually drive the showing process, instead of merely searching for homes. On most real estate platforms, if you find a house you’re interested in, you might see a list of times and dates to schedule an appointment, but you’re not actually booking a showing. Instead, the platform sells that to an agent as a lead, and the agent has to follow up with you – they may or may not be able to show you the house at that time. Showingly is an actual showing platform: when you book a showing, we’re really scheduling it for you in our backend. Until now, the home showing process has been prioritizing agent convenience instead of consumer convenience. Now, for the first time, consumers can have the transparency of directly booking the showings they want.

We also have features built for agents and other professionals. For example, agents are able to delegate showings. If you’re too busy, you can find another agent to show one of your listings; on the other side of the coin, if you have spare time, you can turn that into money by picking up delegated showings.

We’ve been building Showingly for two years, and we launched publicly five months ago. We’re currently live in Alaska, Arizona, Colorado, Hawaii, Massachusetts, South Carolina, Tennessee, and Utah. We continue to integrate with more multiple listing services (MLSs) to launch into new markets.

I have to ask: What has your experience been launching a real estate application during a global pandemic?

Early on, it made raising funding a little harder, but we were able to find investors who wanted to invest during COVID. Surprisingly, it then made hiring easier: people who had lost their internships or jobs came to work for us. We grew from a couple engineers to a dozen, and we’ve probably built as much in the last two months as we did in the year before.

It hasn’t had an enormous effect at a business level. As you might know, real estate markets have reacted in very different ways to COVID. In the Northeast, the market is cautious, but here in Colorado, it’s booming.

What was the genesis of Showingly?

We wanted to start Showingly after learning the ins and outs of the industry through real estate sales. I’ve been an agent, I’ve managed a team of agents, and I’ve experienced a lot of parts of the process. Real estate is such an archaic industry. Technology has done so much in the last 10 or 20 years, and the real estate industry hasn’t materially improved. Sure, there have been some new applications, but they don’t fundamentally change the process – they just put the old process into a pretty website. We saw a big opportunity to actually transform the industry.

How is Showingly using MongoDB?

Showingly is built fully on MongoDB. We’re storing all of our data in MongoDB Atlas, running on AWS. We have one main production cluster, plus a few dev and test clusters.

Our application backend is built entirely on MongoDB Realm, using Realm Functions. It’s really nice to have it all in one place. We use functions for data movement, like retrieving listings from an MLS, as well as application functionality. When you take any action in the app – displaying listings, displaying showings, creating a new showing, updating a showing, and so on – that’s calling a Realm function to access the data in MongoDB Atlas. For frontend, we have a cross-platform mobile app built with React Native, as well as a web client for agents and other professionals. It’s easy to connect to Realm and Atlas from those different clients.

What made you decide to use MongoDB Atlas and Realm’s application development services?

We went to MongoDB World last year, and learned about the document model, MongoDB Atlas, MongoDB Realm, and all of your other products. We knew then that it was the right way to build a simple but powerful architecture. MongoDB has made it easy to get started, but it will also scale with our business: we could go nationwide or worldwide, and MongoDB makes that easy. There’s no reason we would ever need to change our backend.

As I mentioned, my background is in real estate, not technology. But over the past couple of years, I’ve gotten quite good at working with MongoDB Realm and MongoDB Atlas. The simplicity of JavaScript on the frontend and MongoDB on the backend makes it an easy stack to work with. And getting expert help from a consulting engineer quickly taught us the best practices for developing with MongoDB. Working with MongoDB let us develop a great application quickly.

And how would you describe the benefit of MongoDB to your business?

Time to market is certainly part of it, as I mentioned. But I wouldn’t have picked MongoDB just for that. Building for the long term is more important to me than time to market. I wouldn’t take on technical debt up front just to be able to move more quickly. I want to build a structure that lasts, and I’m confident that what we’re building with MongoDB Realm and MongoDB Atlas is just that.

You mentioned that the ability to handle scaling in the future was key for you. What are your plans for scale?

We actually use auto-scaling in Atlas, so our production cluster automatically scales up and down depending on workload. Right now, it’s usually either an M10 or an M20, fluctuating between them as needed. If and when the workload of our application increases beyond that level, Atlas will continue to scale up to match. It’s so easy to set up auto-scaling, so why do it ourselves? And we know that if we need to move to a multi-region cluster or global cluster, that’s very easy to do.

And of course, MongoDB Realm is serverless and we don’t need to worry about scale on that side at all. We just define functions, and they run when needed at any scale.

You said you worked with a MongoDB consulting engineer. Can you describe that process?

We’ve had several Flex Consulting engagements in the Design & Develop and Optimize tracks. Flex Consulting has been the key to making the best use of MongoDB Atlas and MongoDB Realm for our application. We actually covered a number of different points in our consulting engagements.

First and foremost was getting the schema design right. For example, our consulting engineer helped us model the data structure of listings and showings (e.g., embedding vs. linking information), and how to represent data in a way that matches how the application uses it. Getting that design right the first time definitely helped avoid more work down the road.

Advice from the MongoDB engineer also helped us control data quality when multiple people and processes can update records. We fetch listings from MLSs, and if all we had to do was present listings, it would be simple. But of course we’re also dealing with showings tied to listings, we’re enriching those records with other fields for our specific use, and there are cases where an agent might be modifying some of that extra data. So when we refresh listings from the MLS or make updates from the application, we need to make sure that those updates aren’t clobbering other data. Our consulting engineer helped us design Realm functions that would have the correct upsert behavior in all cases.

These consulting engagements were almost like school for us. We spent time with our consulting engineer understanding the technology and making the right design decisions, which was super valuable.

What’s next for Showingly?

In the short term, it’s about growing use in the markets where Showingly is live, and launching into new markets. On the product side, we’re adding some social elements so that agents can see what their peers are up to. We’re also very eager to do more analysis of our data, integrating machine learning to do things such as improving pricing for some of our agent features.

What advice would you give to someone considering MongoDB for their next project?

First, take advantage of consulting. MongoDB’s consulting is the best way to build your project not only quickly, but correctly for the long term.

Second, you’ll get the biggest benefit from utilizing the whole stack of Realm and Atlas together. There’s an enormous amount of convenience from having everything in one place.

In the long term, we want Showingly to be the real estate platform. To date, real estate hasn’t had a good platform that facilitates the process. If you’ve ever bought or sold a home, you know how convoluted it is. You should be able to do everything in a single platform: finding potential homes, getting pre-approvals, scheduling and going on showings, writing an offer, signing contracts and other documents, even closing and getting insurance. We want Showingly to be that platform, to turn a 30-day process into a 3-day process.