Feature Guide

What is Product Discovery?

Your guide to product discovery for ecommerce

Download the guide →

The return of retail therapy


This is where search is critical for the shopper who isn’t too sure what they are looking for. They will need to be guided and inspired in their browsing and presented with recommendations during this type of visit.

Getting the right balance between straightforward usability and the less predictable journey of discovery hasn’t always been easy – often requiring time-intensive tools, system integrations and manual effort to implement and manage.

However, new technologies like artificial intelligence and machine learning now make it easier than ever for shoppers to discover new and exciting products. And they give online retailers a distinct advantage over brick-and-mortar stores by combining customer profiling with deep algorithmic insight to present visitors with new product recommendations that feel personally curated.

In this guide, we’ll take a look at some of the technologies that enable ecommerce providers to quickly guide shoppers on journeys of discovery through often vast inventories. We’ll see how AI is helping online retailers to display laser-targeted recommendations through personalised search listing pages, custom email marketing and more. And we’ll find out how new data-driven approaches to product discovery can help providers beat KPIs, boost revenue and build fiercely loyal customer bases.



Product discovery explained

Put simply, product discovery is the process of guiding customers to their desired items and ranges consistently across all channels.

In the past, sites relied on contextual information like categories, keywords, and other metadata to achieve a scattergun but relatively reliable approach to selecting and displaying tempting new products. But as new technologies like artificial intelligence and machine learning provide more efficient approaches to capturing and analysing data at scale, ecommerce retailers now have an array of product discovery tools at their disposal.

Artificial intelligence uses data to learn how to interpret users’ intent, preferences and behaviour to personalise the product discovery experience across multiple channels in real-time. The process involves collecting data at each of the many touchpoints from discovery to post-purchase. This includes browsing behaviour, cart events, search history and more. As the collection process continues over repeat visits, shopper profiles will come to incorporate more comprehensive data like demographic information, purchase history, style and brand preferences, reviews and more.

testing-2x 3

By comparing similar shopper profiles, algorithms can predict which products will generate the most interest on a user-by-user basis with reliable accuracy. Machine learning and AI can also gain similarly deep insight about products beyond the semantic, keyword-driven approach of old.

Visually similar recommendations like those used by Attraqt client Superdry, for example, show available alternatives for selected items identified by AI. The process enhances recommendation relevancy and increases the likelihood that shoppers find exactly what they are looking for. Additional operational benefits are uncovered for the Superdry team as their merchandisers leverage this AI automation to remove the manual task of creating relational links and cross sells so they can focus on more value adding tasks.

For Shop the Look, Attraqt’s AI-powered recommendation engine displays products in a styled model shot. It lets shoppers browse visually similar products without the need to physically tag each item or redirect visitors to individual product pages.

Product discovery stats

98 %

Reduction in zero-result searches

80% %

80% of turnover is generated via 20% of products that have personalised recommendations applied

135 %

Increase in conversions from visually similar recommendations

Mapping the touchpoints


It starts with a search

Even the most leisurely online shopping experience usually begins with a search. Attraqt’s study showed that 49% of consumers will search for products on branded websites before search engines. This is a golden opportunity for sites to grab attention, deliver relevant results and guide shoppers towards their goals. However, it’s an opportunity that too many online retailers miss. Stats show that an average 20% of all search queries yield zero results – and that’s an ideal challenge for AI.

Attraqt’s AI-powered search uses a host of intelligent techniques to elevate product discovery from the first keystroke. It begins with predictive searching that autocompletes users’ queries as they type. However, Attraqt’s AI-powered search does more than just save time. It serves up a selection of related products, brands and accessories that deep-learning algorithms have carefully selected to appeal to the shopper, individually.

Because Attraqt’s AI-powered search combines deep learning with natural language processing (NLP), it continually improves at serving up more accurate suggestions.


Chatbots and AI assistants

AI and NLP technologies can nudge the buyer towards new products even when shoppers arrive without a specific goal in mind.

Virtual assistants use product and user data alongside a decision tree workflow to make fast, highly personalised suggestions from site inventories.

Attraqt’s API’s can empower chatbots and virtual assistants letting shoppers talk about their buying goals in a fun and conversational way. They can respond to slang, abstract language and other idioms in a way that feels like the user is chatting with a live assistant.

French retailer Nature & Découvertes’ Gift Finder lets users describe the gift’s recipient and answer questions to narrow the search from a +30K inventory. This unique and powerful feature now generates 9% of the site’s overall turnover.


Search results, filtering and personalisation

Traditionally, search results have been generated by matching search terms to keywords and description data in the product listing. However, Attraqt’s product discovery tools combine this with deep learning insight about both users and products to generate personalised search result pages with products ranked by relevance. And when items are unavailable or out of stock, personalised search results pages replace bounce-creating no-result pages with a range of appropriate alternatives.

Attraqt’s product discovery tools also incorporate smart filtering to help shoppers discover new products by brand, colour, size and more.

These powerful techniques are a formidable tool for boosting KPIs like click-through rates, conversions, average order value (AOV) and lifetime value.


Checkout and beyond: Omnichannel product discovery

Attraqt uses the same suite of technologies and techniques to suggest complementary items, accessories, and other intelligent cross-selling or upselling opportunities at checkout.

Product discovery continues later in the journey, with personalised post-checkout social ads, emails and product landing page (PLPs) funnels packed with curated items.

Product discovery challenges


Implementing successful product discovery strategies is labour, time and cost-intensive

Attraqt provides ecommerce and merchandising teams with painless deployment solutions that integrate with ecommerce platforms to deliver value from day one. In addition, the suite of discovery tools automates the most time-consuming and costly elements of product discovery and is easily managed afterwards using Attraqt’s dedicated dashboard.

Retailers don’t have sufficient product or user data to personalise product discovery

Attraqt features a package of strategies that online retailers can use straight away without requiring masses of customer or product data. Going forward, data collection and collation is fully automated, so teams can focus on establishing and meeting KPIs.

ab-test (3)

AI and data science is just too complicated – especially at scale

With Attraqt, retailers get access to pre-trained algorithms to aid product discovery at every touchpoint with minimal management. Attraqt features self-learning algorithms that constantly A/B test strategies to automatically determine and deploy the winning experiments. It can work with vast datasets and inventories and is designed to scale as brands grow.

The power of product discovery


Meet and beat KPIs

Personalised product recommendations help ecommerce teams and merchandisers consistently deliver on every KPI that matters. With custom search results, intelligent suggestions, virtual assistant support and more, Attraqt’s AI-powered product discovery tools boost click-through rates, conversions, AOV, lifetime value and more.


Get deep market insight

Attraqt’s product discovery tools reveal a seemingly infinite variety of new and relevant items to shoppers. For ecommerce brands, AI and deep learning techniques deliver unsurpassed insight on new market segments and product niches and trends that help teams identify new marketing and sales strategies.


Turn shoppers into loyal fans

For users, product discovery is more than just enhanced UX. It also creates a fun and exciting experience that shoppers want to return to again and again.

See Attraqt in action

Want to find out how new approaches to product discovery from Attraqt could help your ecommerce brand? Ask us about seeing a demo today.