Algolia acquires »

Read More

How to Optimize Product Browsing and Discovery

How to Optimize Product Browsing and Discovery

With thousands or even millions of listed products on many retailer websites, you might expect e-commerce website visitors to exclusively use site search. However, even on large marketplace websites a large percentage of visitors still prefer to browse. The numbers vary from site to site, but half or more visitors will still browse through the site nav and merchandising banners. 

In this article, we’ll talk about how to configure the browsing experience to optimize for product discovery conversion. 

What is e-commerce product discovery?

Customers on your site can find products one of two ways: search or discovery. Search refers to site search — a search engine that has indexed your site. Discovery refers to browsing. 

For the purpose of this article, we’ll use the term “product discovery” synonymously with browsing. Although browsing would seem to be quite different from site search, it’s really the other side of the coin. Like search, browse pages are dynamic where results can change or be customized for each visitor.  

Take The Gap, for example. In the screenshot below, the new arrivals in the baby section is a dynamically generated page. When a customer clicks the “baby” navigation item from the menu, the page is populated with products in that category that meet some conditions.  

ecommerce browsing

What if we did a search on their site instead? Here are search results for a “onesie.”

retail website search filters

In both cases — search and discovery — results are displayed in a 4 column layout and customers can filter the items by attributes such as gender, price, size, etc. Results are delivered via a database with merchandising rules that can be dynamic. Results may change from visitor to visitor, or for different sales, inventory levels, or other conditions. 

Anatomy of a category page

website design eye tracking
Image via Nielsen Norman Group.

The order and format that products are displayed on a page can make a big difference in conversion rate. Visitors will scan a page from top left to bottom right, often following an F-pattern. Your most important products and messages should be placed where customers can find them. 

anatomy of ecommerce category design

Category or product listing pages may include a combination of content that is curated by hand, driven by rules, dynamically inserted for a sale or campaign, or listed organically. It can include sale banners above or below the main content, or inserted into the content grid. 

For example, here’s a list of ways content might be ordered and prioritized as customers browse your site:

  • Best sellers. Sometimes it makes sense to push best-sellers into the top positions because, well, they’re best sellers.
  • By margin. The best match to a query, and even the highest-priced match, might not have the best margin and therefore not be the most profitable.
  • Highest or best reviews. Product pages with at least one review have conversion rates over 350 percent higher than those with none, so making those items more visible makes sense.
  • Newest. More visitors than you may realize are visiting your site to find out what’s new, and such items are good candidates for pushing to the top.
  • Discounts and on-sale items. Offering a discount or putting a product on sale is often a tipping point for potential buyers. It’s also a way of moving items with a short shelf life or dealing with overstocked items.
  • Return rate. Returns are expensive, and it makes sense to boost products with the lowest rates of return.
  • Site branding. If your brand competes on price, like Walmart, you might want rankings to reflect that. On the other hand, if your goal is to be perceived as upscale and exclusive, price will not be a factor, and ranking should put more elegant (and therefore more expensive) items at the top.
  • Shelf life. Food products are an extreme example of time-sensitivity. Anything susceptible to spoilage can benefit from high rankings. 
  • Stockouts. About 15 percent of customers who encounter an out-of-stock item will go to another site, so it’s clearly best to bury those items.

This are just some of the attributes you can use to order products on the page. In the next section, we'll explore some ways to use Merchandising, Dynamic Boosting, and Personalization rules to display products on your site. 

Browsable dynamic content

The same rules, merchandising, product filters, and personalization settings you’re using for site search can be leveraged for discovery, too, to deliver a consistent experience. Every platform will have a different approach, but for we give customers the ability to change results via promotions rules (also called merchandising), dynamic boosting and reinforcement learning, and personalization. 


Often collections (category or brand pages) are manually curated by merchandisers. Significant time can be spent arranging products in a specific order, often based on popularity, discounts or inventory that needs to be moved. Depending on how fast inventory moves, the same needs to be done again week after week.

With, customers can schedule campaigns and populate results either manually or dynamically. Manual ranking consists of hand-selecting the products you want to display and dragging them into the order they should appear. Alternatively, merchandising rules can be used to display results dynamically based on conditions such as product popularity, margins, stock levels, discounts, by brand, etc. 

merchandising drag and drop
Drag and "pin" products in the order you want to display them on site search results and category pages, or use rules to dynamically display results. 

For example, if you have a sale on a specific brand or category of product, you can create a rule that promotes higher inventory items within that brand or category. 

You can also add merchandising banners and use our merchandising templates to display results. 

These various merchandising features allow you to optimize the order and rank of results to drive conversions whether someone is searching or browsing your site.  

Dynamic boosting

Search engine try to improve results over time to push higher converting and most popular items to the top. For, this process is done via reinforcement learning and dynamic boosting. Reinforcement learning uses signals such as clicks, purchases, reviews, ratings, etc., to automatically improve ranking. Over time, it pushes more popular and higher-converting products to the top of the search results. 

Related to reinforcement learning is another feature: dynamic boosting. Where reinforcement learning is incremental and evolutionary, dynamic boosting is instant and revolutionary. Together, they continually push the best results in front of customers. 

Dynamic boosting is a feature specific to It can massively increase sales performance by automating the order of products to maximize business outcomes such as conversions, revenue, or profitability. We’ve seen businesses quickly lift revenue by over 10% using only this feature.

dynamic boosting

Dynamic boosts were built for site search, and in fact, the data we’re able to glean from search will inform results for the browse pages as well. Dynamic boosts analyze your customer behavior such as searches, clicks, cart-adds, and purchases. It then builds a model that determines what search or browse results are the most relevant for a given search query.


Personalization starts with data. The more demographic and psychographic data you can collect about your customers and visitors, the more sophisticated your personalization can be. This includes stuff such as:

  • Location (geo)
  • Age
  • Gender
  • Past purchase history
  • Site search history
  • Pages viewed
  • Social media likes
  • Member status (if you have a membership or rewards program)

In, personalization works in the same way; by setting rules for merchandising or dynamic boosting. Simply select the attributes you want and the system will do the rest. 

mobile design for ecommerce sites
Don’t forget to optimize browsable content for mobile, too. In 2022, mobile phones are expected to account for more than 60% of web traffic.

Dynamic filters

The order in which products are displayed makes a huge difference. Equally important are the kinds of filters customers can use to find what they want. 

Your site’s product filters can be static (e.g., your category pages) or dynamic based on the context of the products customers are viewing. Dynamic filters are also known as facets. 

Different facets are displayed depending on what the category is. For example, you might sell shirts and shoes, and both of those types of products will have size and materials attributes. The size and materials filters will change, however, depending on which category someone is browsing.

Filters and facets can be generated automatically when your site is indexed by your site search engine. Both filters and facets help users refine their search query, which allows e-commerce and information-rich sites to deliver a better onsite search experience.

Final takeaways

Search and discovery results can be made to work identically so customers can get a consistent result however they navigate your site. By using intelligent boosting rules and merchandising, you can improve conversions, revenue, and customer experience. 

To learn more about browse features, sign up for a free 14-day trial or schedule a demo with our team.

AI conversions ebook

Similar articles

Best Practice

8 Actionable Tips for Improving Your Search UX and Usability

Best Practice

Why Good Site Search Is Important for Government Websites

Best Practice

How to Fix the Out of Stock Product Problem Using Onsite Search