According to Econsultancy, 30% of consumers use site search to find products — and those users are 50% more likely to convert on products compared to visitors who just browse. Those are big numbers.
But there’s a problem:
What if users don’t find the products they’re looking for using your site search? What if the products they see in your search results are “relevant” to their query, but don’t appeal to them? What if the results aren’t attractive?
If your search isn’t providing users with products they want to see, the statistics above don’t mean much. Achieving real results with site search requires more than placing a search bar at the top of your page, plugging a basic TF-IDF algorithm into it, and letting it sit.
And that’s where searchandising comes in.
What is Searchandising?
Searchandising at it’s core is simple:
Search + Merchandising = Searchandising
In other words, searchandising combines the use of proper merchandising techniques with search experiences: it’s the strategic placement of items within search queries with the goal of optimizing for a business metric, whether it be unit sales, lifetime value, or something else.
There are, however, some slight misconceptions derived from this definition.
Searchandising is more than just eyeballing search results and tuning them to what you (or what an external rater) would call “relevant,” and it’s more than just providing users with basic navigational tools like search facets (although navigation is still important, which we’ll go over shortly):
“Fundamentally, searchandising is about augmenting established search techniques – faceted search and navigation, autocomplete, recommended products, recent searches, related queries, etc – with behavioural data and automation in order to create a seamless, personalised, unique, and highly profitable product search experiences.”
This article will cover every updated searchandising strategy — from advanced autocomplete to personalization algorithms — to give you everything you need to optimize your search with users in mind.
If you’re interested in learning more about attractive results and the limitations of optimizing search for “relevance,” check out our recent webinar.
Let’s start with the basics:
These tools allow merchandisers to control what — and when — products appear for certain search queries, as well as what products appear on other parts of the website (home page, cart page, etc.). Giving your merchandisers manual searchandising tools comes in handy for:
Creating custom, timed collections
Want to advertise themed products to users around holidays or other special events? Merchant tools allow you to quickly create customized collections you can show to users when they search for specific queries.
For instance, if a user searches for “christmas gifts” on your site, you could redirect them to a custom-made collection based around Christmas gifts.
Boost new (or own-brand) products
If you have a new product (or an own-brand product) you’d like to advertise to users, merchant tools should allow merchandisers to quickly slot those items in the top position for queries relating to the item.
Boosting high-margin or high-inventory products
In the same way merchant tools can be used to boost new products, they can also be used to boost products that generate high profit margins or products with high inventory you’re looking to get rid of.
While manually boosting these products can be great for driving impactful business metrics (like high-margin unit sales), it’s important to closely monitor user behavior after altering search results for any query. Some manually boosted products can negatively impact your search experiences leading to an opposite of the business metric you’re trying to drive.
Updating product catalogs
Great merchant tools should allow merchandisers to alter product catalogs as needed through easy-to-use interfaces, FTP clients, or API endpoints.
Solving frustrated searches
Your site will have search results users aren’t happy with — and when users aren’t happy, they’ll either leave the site or refine their search. We call these frustrated searches.
Intelligent search systems today record these frustrated searches and present them to merchandisers. For instance, if 500 users on your site search for “sun block” and 90% of them leave the page immediately after, that query will be presented to the merchandisers as a frustrated search. These systems also show merchandisers the queries that users refine their searches to and the queries they convert on (for instance, “sunscreen”). Merchandisers can then manually set synonyms and redirects to solve those frustrated searches.
Using Merchant Tools Correctly
Humans aren’t always the best at serving the right products to the right users at the right time. While merchant tools do give you control over the products that show for queries, you’ll sometimes end up telling your users what search results they should consider “good” for a query, rather than listening to what they have to tell you.
The right answers almost always come from your users’ clickstream data.
When you know what products are most likely to be clicked and purchased for any query, you can make informed decisions on how to position those products for those queries. This is the basis of serving “attractive” search results. To succeed in today’s eCommerce search landscape, merchant tools must be used in tandem with more advanced tools. \
Want to learn more about serving users attractive search results and the limitations of optimizing search for “relevance?” Check out our recent webinar here.
So, what are some ways we can use advanced tools, algorithms, and data to searchandise?
Boost/Bury Rules Around Availability or Arrival Dates
Of course, your users can’t purchase a product if it’s out-of-stock.
In modern searchandising systems, automatic boost and bury rules can be applied to those products based on their availability (i.e. they can boost in-stock or high-inventory products and bury out-of-stock products). You can also do the same with products that haven’t arrived yet; however, there is a case to be made for boosting products that haven’t yet arrived to increase pre-purchases or to promote an upcoming release.
In our detailed guide on eCommerce merchandising, we outlined the ways in which modern autocomplete systems do more than estimate basic user queries:
- Correct phonetic misspellings, keyboard proximity typos, punctuation nuances, and omitted character typos. Is it “blue ray”, “blue-ray” or “bluray?” It’s officially spelled “blu-ray”, but you shouldn’t require users to know this to get good results.
- Detect synonyms to serve relevant results to users. If a user is searching for “suntan lotion,” good autocomplete will recommend “sunscreen.”
- Identify word importance to better understand what results will most likely lead to a conversion.
- Add support for multiple languages to give the best search experience to any visitor.”
Embedded Product Listings
Embedded product listings work great for encouraging clicks through search.
But that’s not all: our own data tells us users are twice as likely to convert on products they’ve clicked from embedded product listings.
Automatic Results Re-ranking (High-Level)
In a previous example, we discussed how optimizing your search results around “relevance” isn’t the best approach for driving a business metric like unit sales. Instead, one correct approach you can take is to optimize for attractiveness.
But what exactly is attractiveness?
Let’s say you’re a cellphone retailer trying to optimize search results for the “smart phone” query. If you’re optimizing for relevance, you could technically show any number of smart phones to users in the search results — even if they’re 8 years old and cost a fortune — and they would all be relevant.
If you’re optimizing for attractiveness, however, the smart phones that appear would have proven click-through and conversion rates with similar visitors. Instead of showing random “relevant” phones, you’re showing phones you know performed well with other visitors.
Using the clickstream data from your other users (click-through rate, conversion rate, as well as what phones were ignored), modern search systems can now automatically re-rank products for a specific query based on what it thinks will lead to a positive outcome for the retailer.
Personalized Results Re-ranking
In our last example, we showed you how products can be re-ranked at a high-level based on their attractiveness (i.e. how likely they are to lead to a desired business outcome). But here’s a question:
What if we know one of our users has a preference towards iPhones? Maybe, in the past, they’ve searched the same query and only clicked on iPhones. Maybe they’ve been to the site before and had an interest in iPhone cases. Would it then make sense to show them a plethora of Android phones (even if they are proven to convert well with general visitors in the past)?
The answer is no — and smart searchandising systems know this.
Instead of showing a wide variety of phones to the user, you should show phones you know are tailored to their interests (i.e. iPhones). This is the basis of personalized attractiveness in searchandising.
Category Page Personalization
Earlier in this article, we discussed how some retailers redirect users to category pages based on the queries they search. This strategy can be used in combination with personalized results re-ranking to give your users a totally customized category page based on their searching and purchasing habits as well.
Filters and Faceted Search
Filters and facets allow users to find products they’re looking for faster by narrowing their searches by specific product details.
“Once we were able to build faceted search into our navigation, sales exploded” – to the tune of 2.7% lift in average order value.” – Rollie Nation representative.”
Just like your search results, however, your filters and facets should be dynamic based on what users search for. For instance, if a user is searching for “blazer,” you may decide to rank facets like “fit” and “jacket style” higher in the navigation rather than leaving a basic set of static clothing filters. If they’re searching for “pants,” facets like “inseam” and “front style” would be ranked higher.
You can also combine personalization algorithms with filters and facets. Let’s use our “iPhone” example once again:
When a brand new user searches for “smart phones,” you may decide to show that user a “Brand” filter somewhere in your navigation with relevant brands ranked by popularity. If we have a user who prefers to shop for iPhones, however, the facets within that filter can be re-ranked to make it easier for users to find the phone they’re looking for — an iPhone.
In addition to the general re-ranking of products based on their likelihood to drive a desired business outcome, some systems can also re-rank products based on a user’s geography.
“Advanced solutions allow you to set merchandising rules at the geographic level. For example, boost the Kardashian Collection brand in Miami and LA, and bury it in New York and Chicago.”
While the above example mentions setting merchandising rules through merchant tools, the same goal can be accomplished automatically with enough data. If your searchandising systems recognize that users in Miami and LA tend to purchase a different lines of clothing compared to users in New York or Chicago, boost and bury rules can be applied automatically to serve each user the most attractive results.
If used correctly, banners can also be used to drive users to the right search/category pages (as well as to increase interest in specific promotions you may be running):
In addition to banners above search results, badges can be added to specific products to indicate sales or promotional offers:
It’s an unfortunate fact: there will be queries your users search that lead to “no-result pages.” But you shouldn’t let these no-result pages stop your users from finding the products they’re looking for.
For example, if a search terms leads to a no-result page on House of Fraser, a list of recommended products appears directly under the no-result message:
Want to learn more about optimizing on-site search to drive real business metrics?
For years, retailers have optimized search solely on “relevance,” hoping the results that users want to see (and the results that drive important business metrics) appear at the top.
This is exactly wrong — and it’s not how companies like Google and Amazon optimize their search.
We just hosted a new webinar discussing the limitations of optimizing search results for “relevance,” and the 3 attributes retailers should focus on to drive real business results from search. Check it out here: