Param Jaggi is CEO at Agora, a search engine for ecommerce products. Agora has indexed millions of products from online stores like Shopify and WooCommerce, and brought them together in a single search & discovery experience that lets users browse & compare options from different stores.

Param founded Agora with a straightforward premise — in a world with so many stores selling products online, it should be a lot easier than it was to find exactly the product that you’re looking for. In Param’s case, it was finding red shoes for a gift for his wife. So he did what a developer does, and started building that better way. Agora indexes millions of products available on platforms like Shopify, and has built a new kind of user experience around the data with search at the center. Within 24 hours of launching, Agora had served 50,000 users. Within another 24 hours that number had doubled. It was becoming (urgently) clear to the team that a lot of people were also looking for a better way to find the products they were looking for.

Over the next few months Param and the team would isolate two critical challenges in delivering a great search experience for their users — making Agora search really fast, and making Agora really relevant. With the switch to Objective, the team didn’t have to compromise on either — typically a very hard thing to do in search and something that, historically, would take a lot of meticulous engineering and tuning. Not only has the team seen a 5.2x increase in retention, they also saw a huge increase in overall relevance.

Making search feel magical.

Initially, the Agora team launched with the search system that was built into their Mongo database. User engagement was climbing, but the user experience was hard — each user search was taking between 5 and 10 seconds. As Param described it:

“A big problem we faced building search ourself was that everyone uses search on Amazon and Google — some of the biggest companies in the world, with massive teams and massive infrastructure. But that’s the fundamental expectation when it comes to search. You type something in, and you get results that are instantaneous & relevant.”

So, the team explored their options, and settled on a migration to Typesense. But growth continued, and user feedback uncovered the next obstacle - understanding what a user means, not just what they said. “Our search experience was still failing when the user intent mattered. If you search ‘king of the jungle’, you  don’t want to see results for kings, or jungles - you want to see lions. We were seeing this over and over again. People would search for something, but mean something totally different. We needed to figure out how to bridge that gap and show them the products they were looking for.”

This is what results looked like for “king of the jungle” looked like before, with Agora’s legacy search systems, and after with the switch to Objective:

Team started investigating building an in-house AI stack to address the limitations in semantic understanding they were seeing with their Typesense usage.

And the team started working directly with merchants that hosted products on the platform, where they found merchants also wrestling with the same unforgiving search experience. Sellers were having to be creative with product names & descriptions on shopping platforms like Shopify and WooCommerce to try and anticipate the exact words someone might use to find them, rather than being able to let the search system do the work for them. They might call it a lamp, but if someone else calls it a “lantern” or a “light” when they searched for it, that merchant was just out of luck. The team was starting to feel like text-only search just wasn’t going to cut it.

“We had a lot of search providers reach out and pitch us. We evaluated 5-10 vendors, including Elastic and Algolia, but when we found Objective it just seemed like an AI-powered version of everything else in the market.” Param said. “It was really clear to us that you all ship quickly, which aligns with our ethos as an engineering team around shipping velocity. We were coming from a self-hosted world, where we were dealing with hardware configuration, data sharding, memory management — all of those headaches. We think it probably would have taken us 6 months to a year to build on our own what we ended up integrating in 3 days with Objective.”

The result of that 3-day integration didn’t take long to materialize. The Agora team saw a 5.2x increase in retention almost overnight, and saw their overall relevance go through the roof.

This is what Agora search results for “yellow ball gown” looked like before, with their legacy search provider, and after with AI-native search powered by Objective:

“Before, our search felt like it just wasn’t there yet. Since we turned on Objective, it feels like search is just working the way it should have always worked. We’re able to deliver results really fast and the results are exactly what the user expected. We’re able to give users the experience they expect from some of the biggest companies in the world.”, Param explained.

Passing the mom test

And while all of those metrics make everyone over here at Objective HQ smile, the thing we’re most proud of (and we’re guessing Param is as well) is that he finally passed the mom test.

Param beamed, “I think our search finally passed the Mom Test after today. I showed it to my mom and she said it actually worked as she expected.” ❤️