Optimizing Email Conversion for
Real Estate Portals

Make All Your Data Work For You

Your real estate portal is not just about your core business. Consumer monetization products make up a large part of your revenue. From helping banks find leads for mortgages to linking relocation companies with potential customers, your user data enables a wealth of additional revenue streams. But how much you earn depends on how well you target your users.

Let’s say you’re trying to identify users who might be interested in relocation services. You have a ton of user data, and you already have a stellar email campaign ready to go. You’ve come up with several rules to select by: any user who has looked at more than five URLs for relocation companies in a the last two days; any user who has viewed more than 20 listings and in the last two days no more than five – these people are serious about relocation. So you apply these rules to your targeting system, and any user who exhibits these behaviors will be added to the list and receive your email campaign about relocation services.

So far, so good. Then the statistics start coming in. Open rates, click rates, conversion rates – how many of the users who receive the email actually fill out the form to request a quotation for relocation services? And how can you optimize this conversion rate?

The simplest way to optimize your email conversion rate is to optimize which users receive the campaign. Open, click, and conversion rates depend heavily on who you send the offer to. You might have a perfectly crafted email, but unless the email gets to the most interested users, you won’t sell it. This means the accuracy of the system that selects the users is key. If you improve that system just a little bit, you’ll get a significantly higher conversion rate.

Your real estate portal may have hundreds of thousands or even millions of users, but only a small fraction of those will be interested in a relocation service right now. Most people sort out their relocation needs in about a week. It’s a time-sensitive issue, and you need to make sure your email campaign reaches the right users at exactly the right moment.

As a data-savvy real estate portal, you’re already aware that your user data is a potential goldmine. You’ve manually created several rules which you’ve applied to your targeting system. It’s a time-consuming process: a manager has to decide which rules to apply, and an engineer has to build the system to automatically select which users receive the campaign. Once you get up to seven or eight rules, things get complicated – it becomes too much to handle.

What’s more, even if you have seven or eight rules, there are still thousands of attributes and behaviors your targeting system isn’t considering. Your data can tell you when your user registered, where she’s from, what kind of device she’s using to access your portal, and every page she’s ever visited, but there’s no way to integrate all that information into your selection process manually. So you’re using only a tiny fraction of the data you have to select the users for your relocation campaign. And this means that your conversion rates are nowhere close to optimized. You’re potentially missing out on a huge amount of revenue.

Don’t Leave Valuable Data on the Table

You have big data, you’re storing it, and you know you should be using it. Machine learning is the only way to efficiently extract thousands of valuable patterns from the data you already have.

Your real estate portal collects a lot of unique behavioral data: your users come back again and again, and they spend a lot of time on the site. Maybe at first they’re just thinking about moving, checking out the neighborhood and whether it’s affordable to buy or rent. Then a few weeks later they come back with more serious intentions and spend a month searching and making inquiries. Then they take the plunge and choose their dream home, but until they sign on the dotted line, they keep searching on the side, just to be sure. This is when you want to target the right users. You need to take into account all the data you have over an extended period of time – which user logged in when, from where, on what device; which pages he visited or which advertisements he clicked on. All this data is potentially important for your relocation campaign, but you need a sophisticated machine learning system to optimize your conversion rates and maximize your revenue.

Same Emails, Double the Revenue

We’ve helped large real estate portals double, and on some products even quadruple, their revenue from emails.


How Does It Work?

First, Scoring AI goes back and analyzes all the users who previously requested offers for relocation services. It sees every aspect of their activity on your portal and makes connections between certain user behaviors and conversion. The machine analyzes these interactions with incredible granularity, depth, and objectivity. This means the patterns it finds are much more solid, fine-tuned, and varied. Scoring AI takes a huge tree of thousands of rules and treats each user individually based on her specific attributes.

Once you’ve trained the machine to turn behavior analysis into a prediction of who will fill out the relocation form, you can apply that machine to your current users and their current behavior. The machine then spits out how likely each user is to convert.

Then you take the cream of the crop, the users with the highest score, and you send them your stellar email campaign for relocation services. Since these users are selected based on their complete behavioral history on your portal, you’ll get a significantly higher conversion rate.

How high? We’ve seen revenue improvements of up to 400%. The average for all the scorings we’ve built is 250%. This translates into additional revenue of up to 200,000 € annually for a single product – like your relocation campaign.

Working With Us

This all sounds great, but how will Scoring AI affect your campaign management system? Our job is to feed optimized information back into your system, no matter which system you use. We make sure the information gets delivered exactly where you need it, when you need it. The score will surface as one new attribute in the same tool you’ve already been using. From there, you can use the data any way you’d like.

How long will this take? We build scorings for real estate portals in less than two months.

How much will it cost? Great question. A custom Scoring AI solution costs about 10% of the price of building it in-house. We make sure every scoring solution pays itself back in additional revenue within 3 - 6 months. Our average payback period is 3 months.

When is Scoring AI not right for you? It all depends on the data. If your users don’t log in, or if they only come to the site once or twice and never come back, scoring will not significantly raise your email conversation rates. Likewise, if your analytics tracking can’t connect a user ID to an email address, then machine learning can’t help you make the connection between what a user does and how to contact him with your relocation campaign. However, in this case you can still use scoring for cookie-based retargeting.

The more user data you have, the more consistent histories and complex information you’re gathering, the more patterns the machine can find, the better it can target, the greater rise in revenue you can expect.


The Perfect AI Partner

We a string of success stories and happy customers that testify to the ease of our process and the benefits of our experience. We’ve built more scoring projects than most data scientists do in their whole careers. We understand the challenges you’re facing, and we know just how to help you optimize your email conversion rates with Scoring AI. Contact us to find out more!

Rheinsberger Straße 76/77 10115 Berlin
+49(0)30 99256788 - info@datarevenue.de