Product discovery is the moment a shopper decides something is worth a closer look. Most e-commerce brands spend their budget optimising checkout flows and ad targeting, then wonder why their conversion rate stubbornly refuses to climb. The problem is usually much earlier in the journey. When customers cannot find the right product at the right moment, no amount of checkout polish will save the sale.

Key Takeaways

  • Poor product discovery kills sales before a shopper ever reaches a product page.
  • Most SMBs treat search and navigation as a one-time setup rather than an ongoing investment.
  • Personalisation at the discovery layer can materially lift average order value and repeat purchase rate.
  • Catalogue architecture matters as much as the products themselves.
  • Discovery problems are operational and strategic, not just technical.

Why Do Most E-Commerce Brands Get Discovery Wrong?

The short answer is attention. Checkout is easy to measure. A drop-off at payment is visible in any analytics dashboard. A shopper who browses for ninety seconds, fails to find what they want, and leaves is almost invisible. Google Analytics 4 will show a bounce. It will not tell you they were ready to buy.

According to a 2023 Forrester Research report, around 68% of online shoppers abandon a site because they cannot find what they are looking for. That figure has remained stubbornly high for years despite the widespread adoption of search tools and filtered navigation. The infrastructure exists. The implementation rarely does.

There are three core reasons brands fall short here.

  • They treat site search as a feature to install, not a system to tune.
  • They build navigation around their internal catalogue logic, not how customers think.
  • They invest in paid acquisition to compensate for poor organic discovery, which compounds the problem.

What Does a Discovery Failure Actually Look Like?

Consider a mid-sized Australian outdoor and camping retailer. They stock around 4,000 SKUs. Their navigation was built when they had 400. The top-level categories still reflect that early structure. A customer searching for a compact sleeping bag for a toddler types "sleeping bag small child" into site search and gets zero results. The product exists. The tags do not match the query.

This is not an extreme example. It is common. Catalogue growth outpaces taxonomy maintenance in almost every SMB e-commerce operation. The team is focused on buying, marketing, and fulfilment. Nobody owns the search experience.

The same issue appears in different forms across markets. A Singaporean fashion retailer might have a navigation system built for desktop that collapses into an unusable accordion on mobile, where over 70% of their traffic lands. A Canadian supplements brand might have twelve subcategories under "protein" that mean nothing to a new customer who just wants something for post-workout recovery.

Is This a Technology Problem or a Strategy Problem?

Both, but strategy comes first. Better search technology will not fix a catalogue with inconsistent product data. Personalisation algorithms cannot recommend what is not properly tagged. The most common mistake is reaching for a tool before diagnosing the underlying structure.

Platforms like Shopify, BigCommerce, and Salesforce Commerce Cloud all offer increasingly capable native search. Third-party tools like Klevu, Constructor.io, and Searchspring layer personalisation and merchandising control on top. These are good products. They work better when the catalogue beneath them is clean and logically organised.

Before evaluating any search or discovery tool, brands should audit three things.

  • Tag and attribute completeness: Are all products tagged with the attributes customers actually filter by?
  • Synonym coverage: Does search understand that "trainers" and "sneakers" refer to the same category?
  • Zero-results rate: What percentage of searches return no results, and what are those queries?

The zero-results rate is the single most diagnostic metric in discovery. A well-run e-commerce operation should sit below 5%. Many SMBs sit at 15 to 25% without knowing it.

How Does Catalogue Architecture Affect Discovery?

Catalogue architecture is the invisible infrastructure of e-commerce. It determines how products are grouped, what attributes they carry, and how they relate to each other. Most brands build this structure early and rarely revisit it.

The problem compounds with scale. When a brand adds 500 new SKUs per year without a governing taxonomy, the catalogue becomes a patchwork. Navigation paths get longer. Filter options multiply without coherent logic. Customers get decision fatigue before they reach the product.

A clean taxonomy has a few key properties.

  • It reflects how customers shop, not how the warehouse is organised.
  • It is shallow enough to reach any product in three clicks or fewer.
  • It uses consistent attribute naming across all products in a category.
  • It allows meaningful cross-category discovery, for example "shop by activity" or "shop by occasion" that cuts across traditional product lines.

Cross-category pathways are where significant revenue lives. A customer shopping for a birthday gift does not want to browse "Men's Accessories" and "Women's Accessories" separately. A curated gift guide path, or a filter for "suitable for gifting", serves that intent directly.

What Role Does Personalisation Play in Discovery?

Personalisation at the discovery layer is one of the highest-return investments an e-commerce brand can make. McKinsey research from 2023 estimated that personalisation at scale can lift revenue by 10 to 15% for retailers who implement it effectively. For SMBs, even basic personalisation achieves meaningful gains.

Basic personalisation means showing a returning customer products relevant to their previous browse and purchase history. It means surfacing "recently viewed" items prominently. It means adjusting homepage hero content based on a customer's known segment or acquisition source.

Advanced personalisation goes further. It reranks search results based on a customer's inferred preferences. It adjusts the order of category page results. It surfaces complementary products based on what similar customers have bought together.

Most SMBs stop at "recently viewed." That is a missed opportunity. The brands that grow quickly in saturated categories are typically those that make discovery feel effortless for returning customers. A shopper who feels like the site "gets them" will spend more and come back sooner.

Are Recommendations Doing the Work They Should?

Product recommendations appear on almost every e-commerce site. Most are underperforming. The default "customers also bought" widget, served without personalisation or context, converts at a fraction of what a well-tuned recommendation engine achieves.

There are four placement zones where recommendations create meaningful lift.

  • Product detail pages: Show complementary items, not just similar ones. "Complete the look" or "pairs well with" framing outperforms "you might also like" in most categories.
  • Cart and checkout: This is not just an upsell opportunity. Recommendations here can surface an item the customer was browsing earlier and forgot to add.
  • Zero-results search pages: When a search returns nothing, a recommendation block showing trending or related items keeps the customer on site. A blank page is an exit.
  • Post-purchase emails: The first email after a purchase is opened at two to three times the rate of a standard campaign. A well-placed recommendation in that email generates disproportionate return visits.

Each of these placements requires a different recommendation logic. Treating them all the same is a common mistake.

Visual search has moved from a novelty to a genuine commercial channel. Pinterest Lens, Google Lens, and platform-native tools on Snapchat and TikTok all allow customers to find products by pointing a camera at something they like. For apparel, home decor, and beauty brands, this is no longer an edge case.

Brands that optimise for visual search need high-quality product imagery from multiple angles, consistent image naming conventions, and schema markup that connects images to product data. These are backend disciplines that most SMBs have not implemented, even when their category is inherently visual.

Social commerce is accelerating this trend. In markets like the US and Australia, TikTok Shop and Instagram Shopping are driving significant discovery volume for SMBs, particularly in fashion and beauty. A shopper who discovers a product on TikTok and taps through expects to land on a page that confirms the product, shows it clearly, and makes purchase obvious. If the landing experience is disjointed from the discovery experience, the sale evaporates.

If you are building or reviewing your social discovery strategy, a structured content planning system helps maintain consistency across channels. Lenka Studio's free social media toolkit is a practical starting point for teams managing content across multiple platforms.

What Should SMBs Fix First?

The order of operations matters. Brands with limited resources should sequence their investment carefully.

Start with the zero-results rate. Pull the data from your site search tool. Identify the top 20 queries returning no results. Many of those queries represent real demand. Fixing those specific gaps, through better tagging, synonym rules, or new landing pages, returns faster than any platform upgrade.

Then audit your top-level navigation. Ask five people who do not know your brand to find three different products using only the navigation. Watch where they get stuck. Navigation failures are usually obvious once you watch someone encounter them in real time.

After that, look at your mobile browse experience. If more than 60% of your traffic is mobile (which it is for most SMBs in Australia, Singapore, Canada, and the US), your category pages, filter UI, and search input need to work at thumb scale. Many desktop navigation systems degrade badly on mobile without anyone noticing.

Finally, evaluate your recommendation engine. If you are using default platform recommendations, test a third-party tool on one high-traffic category page and measure the lift over 30 days. The data will tell you whether the investment makes sense at your scale.

Teams at Lenka Studio regularly encounter discovery problems when helping e-commerce brands audit their digital experience. The pattern is consistent: the checkout is polished, the acquisition spend is significant, and the product discovery layer is three years out of date.

Frequently Asked Questions

What is product discovery in e-commerce?

Product discovery refers to every moment in a shopping journey where a customer finds, evaluates, or considers a product. It includes site search, category navigation, filters, recommendations, and social or external discovery channels like TikTok and Google Shopping.

How do I know if my e-commerce site has a discovery problem?

Check your zero-results search rate, your bounce rate on category pages, and your session depth. If customers are leaving after viewing fewer than two pages, or if more than 10% of searches return no results, you have a discovery problem worth investigating.

Is product discovery different from SEO?

They overlap but are distinct. SEO drives discovery from external search engines like Google. On-site product discovery refers to how customers navigate and find products once they are already on your website. Both matter, but on-site discovery affects customers who are already ready to buy.

How much does fixing product discovery typically cost?

It depends on the approach. Fixing catalogue tagging and synonym rules can be done with internal resource. A third-party search and personalisation tool typically costs between $300 and $2,000 per month for SMBs, depending on catalogue size and traffic volume. The return usually exceeds the cost within the first quarter if implemented properly.

Do I need a developer to improve product discovery?

Not always. Many improvements, like fixing product tags, updating navigation labels, and adding synonym rules in your search tool, can be done by a non-technical team member with platform access. More advanced personalisation and visual search implementations typically require development support.

Ready to Fix Your Discovery Layer?

If your conversion rate is not moving despite strong acquisition spend, product discovery is worth examining closely. The team at Lenka Studio works with e-commerce brands to identify where customers are dropping off and what structural changes will have the most impact. Get in touch to talk through what you are seeing in your data.