As 2025 comes to a close, we’re wrapping up the year with meaningful improvements for all LupaSearch users!

Beyond the updates we’ll discuss in a second, we’ve been investing in performance, analytics, and internal systems — all to ensure faster, more reliable search and recommendations.

To close out the year on a high note, here’s what’s new.

Personalization, powered by purchases

We’ve just rolled out a powerful personalization update across both Search and the Recommender.

By using a shopper’s past behavior, you can now super-personalize homepage suggestions, product page recommendations, and even search result ranking. This is especially impactful for returning visitors, where relevance matters most.

Here’s how it plays out in real life:
a shopper searches for “women’s running jacket,” views a few running jackets, and interacts with several products. The next time she visits your store, the homepage automatically adapts — suggesting running-related clothing and accessories.

Behind the scenes, we’ve also extended the data collection window from 30 minutes to up to a day. This allows personalization to reflect longer-term interests, while still giving the highest weight to the most recent searches and actions.

Smarter filter ordering (It’s live!)

Filters are now more logical and easier for your shoppers to use.

Numeric filters (such as width, weight, size, gigabytes, or any other automatically recognized numeric units) are now automatically sorted in a proper order. Values are displayed from smallest to largest, exactly how users expect to scan and compare them.

On top of that, we’ve unlocked custom filter ranking. This gives you full control over how filter values appear when alphabetical or numeric sorting isn’t the best choice.

For example, you can:

• Show the most relevant cities first (e.g., London, Birmingham, at the top);
• Prioritize specific sizes your business wants to highlight;
• Ensure clothing sizes appear in the correct order (XS, S, M, L) instead of an incorrect alphabetical sequence.

Search Query access

From now on, all configured queries can be queried publicly if their query key is known (which is great, as we want as many users as possible to use your e-commerce search).

At the same time, we know not every business wants this data to be publicly reachable, especially if search is used for internal systems or non-public catalogs.

That’s why we’ve added optional authentication for Search Query access.

That said, you can now choose between:

• Public access (available by default);
• Secured access that requires authentication (if your business applies limitations).

Smarter autocomplete suggestions

Autocomplete suggestions are now more intentional and genuinely helpful for your shoppers.

You can now set up rules in the console that combine brand + category in autocomplete. For example, when a user starts typing “Samsu…”, suggestions can appear as “Samsung TVs” instead of generic or random brand-only results.

Now, autocomplete suggestions guide users toward a concrete category (or any other field combinations that you can select), helping them reach relevant results faster. This is especially powerful for categories such as shoes and apparel, where intent is often brand-led but category-specific.

Instead of sending users into a broad Adidas product pool, autocomplete can guide them directly to “Adidas shoes”, “Adidas jackets”, and similar focused paths.

More control over Recommendation Rules

Introducing: more flexibility to recommendations.

A common use case — when a shopper is viewing Nike products, many businesses want to recommend only Nike-related items (or even avoid cross-brand suggestions entirely). You can now set this up cleanly and intentionally.

At the same time, you can create targeted rules for other scenarios. For example, if a user searches for ‘Samsung’, and your business wants to highlight two specific products — say, a drone and another trending item — you can define those precisely.

The key improvement is how these rules interact. You can now set conditions like:

• Show specific recommendations for everyone
• Exclude Nike-related keywords
• Show specific products, excluding Nike, alongside all other products.

This ensures rules don’t conflict with each other and gives you full control over when and where certain products appear.

Additional notes

With this release, LupaSearch finishes the year with clearer control, better relevance, and a stronger platform overall. We’re looking forward to what we’ll build together next.

We remain committed to improving our e-commerce search solution and delivering the best experience for our customers.

If you encounter any issues or have suggestions on improving it, contact our team at [email protected]. It will help us further improve our service.