6 tips to improve onsite search in Mercadona Online
Last week, I enjoyed reading ¿Por qué no hay SEO en Mercadona Online? or ‘Why is there no SEO on Mercadona Online?’. I sincerely applauded Jose Tarheel for openly sharing key insights on Mercadona’s product roadmap and digital strategy. Big kudos to his organisation for embracing a culture of transparency.
I found the critical role that search and discovery play in their product agenda quite remarkable. Naturally, as an enthusiast and a professional working in the search space, a topic I regularly write about on this channel, I firmly sympathised with what he had to say.
#1 “Buscador para permitir que el usuario encuentre los productos que quiere comprar de la manera más rápida y cómoda posible.”
#1 “Search enables the user to find the products they want to buy in the fastest, most comfortable way possible.”
As Jose highlights in the quote above, search is the easiest way for users to quickly add things to their shopping basket. Its primary role is to minimise the effort required and friction created in finding what they’re looking for. However, I’d argue that search goes well beyond that. Searching and finding are only the foundations. Maximising discovery is what adds the value. So, how can search lead to discovery? Finding to exploring? Satisfying a need for locating a product to sparking curiosity and nurturing serendipity?
It may seem that search and discovery are two use cases with opposing objectives, narrowing queries vs expanding queries, but that’s a missed opportunity. The design of an irresistible and functional search experience can change behaviours, and satisfy both needs, transforming food catalogue browsing into an enjoyable and pleasurable shopping experience.
At EmpathyBroker, we believe that search has the potential to be the best shopping ambassador for brands. Search is capable of acting as the perfect shopping assistant. It can help people fill their shopping basket with new products based on their family’s needs, favourite foods and brands, nutritional profiles, seasonal products, promotions, etc. As a system that listens to every customer’s demands and understands shopping patterns, who better than search to recommend products to help us lead a healthier lifestyle and save some money at the same time?
The road to getting search and discovery correct is a long one. However, I wanted to use this post to highlight some interesting steps that can point anyone looking to start the journey in the right direction. If I was asked to suggest a handful of tips to Jose on how to design and evolve onsite search at Mercadona, these would be my top six pieces of advice:
- SEARCH-AS-YOU-TYPE SUGGESTIONS (SAYT): Search suggestions help users not only save time and typing effort they also help them discover new search options, in many cases sparking new purchase ideas. These search autocomplete suggestions are automatically extracted from past user behaviour. They play a critical role in recommending seasonal queries and overall customer demand, practically in real time.
In fact, if you try searching for the same keyword in different grocery stores, you’ll easily recognise what type of consumer each brand has and what their most demanded products are (see below).
2. PERSONALISED SEARCH RESULTS: Once the user has performed a query, browsing through dozens of search results is a tedious and time-consuming task. How can search results be as relevant as possible to every single buyer? Understanding context, such as the user’s previous purchase behaviour, their nutritional profile, and other affinities is fundamental in terms of search results accuracy. This understanding could be used in a number of ways:
i. Display “Last purchased” items on top of the search results list and save valuable time from unnecessary filtering and scrolling. This is highly recommended for all sort of products that are purchased on a periodic basis, such as cereal, milk, olive oil, soap, etc.
ii. Show only products that fit the user’s profile. Items that match a user’s profile, such as nutritional preferences and allergies, favourite brands, promotions, etc. should be displayed on top of search results. Appropriate facets/filters should also be automatically applied, further reducing friction in their search experience.
3. RELATED TAGS: Related tags are a relatively new thing in eCommerce search. However, soon they will be widely used among all online grocery stores. Related tags are extracted from past users’ search sessions and are displayed in the SERP after a query has been performed. By clicking on related tags, users refine their original query and formulate new ones without the need to type. This way users can quickly check different product options/attributes, (e.g. flavours, popular pack sizes, packaging types, recipes, etc.), in a joyful and entertaining way.
Once again, the goal here is to reduce friction in the SERP and to provoke a change in behaviour, sparking curiosity and further catalogue exploration.
4. FILTERS/FACETS: Filters play a critical role in grocery shopping. Filtering, also known as faceting, is widely used across all eCommerce verticals and allows users to reduce noise from results. Filters should be carefully designed from an experience perspective. For example, they should be easily accessible from anywhere in the SERP, they should be grouped and displayed in a way that allows multiple selections, and they should always anticipate the number of results before they’re applied.
With regards to facet options, they should also go beyond traditional product attributes, (e.g. brands, subcategories, packages/size, price, etc.), and allow filtering through categories such as nutritional characteristics (e.g. gluten-free, nut-free, etc.), promotions, delivery options, etc.
5. NEXT QUERIES: Using search as an assistant for filling up the shopping basket, similar to a smart list creator, is also quite innovative. The collective analysis of millions of search journeys allows predicting what the next query of a user journey could be, thus anticipating search demand. Although this feature can be implemented in multiple ways, let’s focus on a simple use case for now:
A user performs a search for “eggs”, adds eggs to their basket. They go back to the search box and, before they start typing, a list of recommended and “contextual” queries is displayed (e.g. milk, bread, bacon). The user then adds bacon to their basket and the search box follows up by suggesting sausages, mushrooms, etc. next.
6. SEARCH TRENDS/ DISCOVERY WALL: The search box is the start of a journey of discovery. Once the user clicks on the search box, it’s the right time to start recommending new things. Search trends are a great way to do that. “Trending Now” discovery walls are becoming increasingly popular, thanks to the likes of Instagram, Twitter, Pinterest, etc.
Discovery walls are great for showcasing seasonal products, promotions, etc. based on popular demand and almost in real time. Search trends should not only display the most popular queries, but also the queries with the highest growth and propagation rate within the last 48hrs.
Those are my six top tips to getting search and discovery improvement journey off the ground. As always, there are many more things that could be looked into, but this is a great start whatsoever. If any of you have other suggestions, please share them in the comments section below. I’d love to hear what you think.
Thanks for reading!
