Algolia acquires »

Read More
Case study

Learn how RentPath boosted search conversions for their marketplaces

Real estate site search case study

Founded in 1989, RentPath is a digital marketing solutions provider that operates multiple sites to connect property owners with renters. More than 10 million unique monthly visitors are able to search across nearly 250,000 rental properties on, ApartmentGuide, and RentPath was acquired by Redfin in 2021. 

RentPath offers a variety of tools to help property owners run their business — all the way from listing management and discovery to a completed rental application. The process begins with site search. 

The scope

The team had been using Elasticsearch to run site search, but felt it had become too burdensome to manage. Extensive, complex rules and customizations were required to get certain types of searches to work. 

“We wanted a more scalable and cleaner implementation for ranking logic. Over time we had added a significant number of rules and logic that had become difficult to manage,” explained Teo Harrison, Senior Product Manager for Search at RentPath. 

RentPath also wanted a better way to utilize business data to drive better search results. For example, are the sales leads leading to better outcomes for advertisers? Is there a better way to provide exposure for properties that are “lead deficient” and give renters a wider array of choices? They wanted a search solution that could leverage data from successful outcomes to deliver better results for future searches to help renters and property owners. 

Finally, as a two-sided marketplace, RentPath needs to provide balance in its results to work for both property owners and renters. Property owners need better leads across all their units, and renters need to see a wide variety of options to match their needs. 


The team at RentPath selected to replace Elasticsearch. offered a vastly simpler and more powerful approach to optimizing search results to meet their criteria. 

“We wanted to be sure that we were distributing leads at least similarly to before. We wanted to have some assurance that the transition would be smooth, but not at the expense of our clients or renters.” 

The company first introduced the new search engine on ApartmentGuide. RentPath tested different search ranking algorithm settings to achieve the same baseline as it had before with Elasticsearch. After reaching the new baseline, the team was able to do additional A/B test experiments to deliver even more value. With, search algorithms can be easily adjusted, saved, and versioned via pipelines or relevance settings

“With, we’ve simplified the search model. We can tell the app which business goals are important and know that we’re actually optimizing for those goals. It has the added benefit that we can send even more data to the model to provide even more optimization.” also provided natural language processing (NLP) algorithms to allow renters to search more easily. Rather than simply requiring users to enter a number of bedrooms, bathrooms, and a zip code, visitors can now type in natural language queries — such as “2 bedroom in Los Angeles near the beach” — and get great results. 

Results are in

The company was able to measure two clear outcomes very early on. Not only were they able to increase the number of leads they were distributing by a whopping 18%, but they also improved distribution of “lead deficient” properties by another 5%. For RentPath’s customers, that meant properties which were being underserved before are now getting better placements and leads.

RentPath have also seen a better redistribution of leads to higher-paying advertiser tiers, and improved the number of leads per property. 

“We’ve introduced a model that allows us to vary the ranking order. We can redistribute leads to ensure that no one is getting too many or too few,” added Teo. 

As he explained, with a two-way marketplace, there’s a balance between delivering results for property owners and renters: “If we over-index results for our clients, then we might force renters to express interest in properties that they’re not actually interested in. On the other hand, if we only focus on renters, then they can polarize to only a certain set of listings without getting the full range of options.”

Every company and website has data and success measures that are unique to it. For RentPath, there was a need to deliver balanced results for both types of user while also increasing overall success outcomes. Features in gave the team at RentPath the ability to easily adjust the search algorithm and leverage site data to optimize better outcomes. 

“We’ve really enjoyed working with the team. We meet with your team regularly and they help us to dig into the data and learn from the data to provide areas of recommendation,” said Teo.

The company is now rolling out on their other sites following the same successful process as they used on ApartmentGuide.

Not only were they able to increase the number of leads they were distributing by a whopping 18%, but they also improved distribution of “lead deficient” properties by another 5%.

What are you waiting for? Get in on the action