Here at Objective HQ, we’re on a mission to make search more human. A big part of that is applying this modern generation of AI retrieval technology & techniques to the oldest problem in software — finding things. But a big part of making search more human is also about putting search & retrieval in all of the places people already are in your product experience. Great search experiences come from bringing search to the user, not making the user come to your search widget.
There are four places in particular we’re pretty excited to see people building some fascinating things with Objective:
- Making discovery as important as ‘search’
- Upgrading the search you already have with semantic understanding
- Adding AI-driven recommendations and dynamic category pages
- Driving Product recommendation emails with AI-native retrieval
Making discovery as important as ‘search’.
Historically, e-commerce search has relied on users who already have specific purchase intent interacting with the big search bar on your website. But there’s a whole world of user interaction you’re missing out on in helping users browse & discover before they know what they’re looking for. Styl.digital has built a business entirely around discovery, with a tinder-like swiping experience for fashion exploration - and usage is exploding.
The folks at Styl have integrated AI-native search from Objective to power the recommendation and exploration engine they’ve built, that enables new kinds of discovery like their ‘Search by Vibe’ feature. When your search is able to understand human intent in a search query rather than just individual words, the possibilities for new discovery experience get pretty incredible. Browse dresses for a summer wedding? No problem. Show me coats with big buttons? No problem.
Upgrade the search box you already have with semantic human understanding.
You might be a product & engineering team that wants to start modernizing your search incrementally — a great place to start the journey is just improving the experience your users already have by replacing your legacy keyword-lookup search system with an AI-native upgrade. You’ll see three major benefits in AI-native platforms (like Objective!) — semantic understanding, machine learning, and real multimodal search that understands all of the data types in your Objects together, instead of just smashing separate results together.
Developers tell us the biggest hurdle here is the integration effort of legacy solutions — so we’ve put a lot of effort into making this dead-simple. Adding your products to an Objective Index is a few lines of code, and shipping AI-powered search in your app today is literally as simple as:
Pod Foods is a company that started their journey into great search relevance this way, going from a home-grown PostgreSQL installation that eventually started to hit a major ceiling in user experience, and have now fully migrated to an AI-native solution. Keyword search only goes so far when your users are searching things like hard to spell food ingredients & allergens.
They tested other legacy keyword-lookup solutions like Algolia and Elastic, but finally broke through the ceiling they had hit by switching to semantic search with Objective. The first big jump in relevancy they saw was just from making the switch to an out-of-the-box Objective Search index. And they saw another whopping 19 percentage-point boost in relevancy when they started finetuning their search.
The ability to finetune is what really unblocks search teams that hit a relevance ceiling with a legacy provider — those legacy sytems are frozen, and can’t learn. With Objective Finetuning, Pod Foods was able to teach the system what good and bad results look like, and let the system learn alongside their user behavior. That means less time manually boosting & deboosting, and more time focusing on creating great user experience.
Product search is also the area that real ‘multimodal’ search shines — the ability for a platform like Objective to holistically understand an item in your product catalog, both semantically understanding the text and ‘seeing’ the meaning inside images, together. The result is that your users can search like they speak — if they’re looking for a coat with big buttons and you have one, they’ll find it. Even if the only indication of ‘big buttons’ is visually in the product images.
Adding AI-driven You Might Also Like and smart SEO-focused category pages
After you’ve upgraded your core search experience to an AI-native approach, you start to look at all the other places your retrieval could be smart. A lot of times the next low-hanging fruit is You May Also Like recommendations — check out these other products like this one. With an AI-native search platform like Objective, adding AI-powered retrieval is just as easy as upgrading your core search experience was.
And you can take that same idea to your ‘category pages’ — the pages your team builds out for various product categories to capture SEO traffic around different themes of products. Traditionally, you’d be left to manually promote items to each of those pages (and there could be hundreds of them), or rely on internal category taxonomy to dictate which pages exist at all. Both of those things make a category page targeting “Dresses for a Summer Wedding” a really hard task.
AI-native search does this without breaking a sweat. It’s all still ‘search’, but it’s taking that powerful AI-native search tool and making it portable to every layer of your user experience.
Product recommendation emails users look forward to opening.
And last but not least — you might already be sending product recommendation emails to bring users back to your experience. Just like You May Also Like & SEO category pages, AI-native search can be a powerful tool in your toolbelt to dynamically retrieve products based on season, user preferences, or almost anything else your developers can dream up.
We can’t wait to see what you build next!
And if you already have built things like this, let us know! We’d love to see it.
