What if the secret to increasing conversion rates didn’t lie in UX or A/B testing? What if it all came down to how you merchandise your category pages?
No matter how many best practices your online store adheres to, or how many rounds of optimization you’ve performed, you’re not going to generate revenue unless customers discover the products they crave. And if a customer is dead-set on acquiring your goods, she’ll put up with the clunkiest of checkout flows — even get on the phone to buy.
The biggest missed opportunity in ecommerce is optimizing the presentation of products in your category and sub-category grid pages.
Showing the “best” items first impacts revenue, basket size, customer loyalty, inventory turnover and profit — especially on mobile, where space is limited and fishing through hundreds of products is painful.
The Other-other CMO – Category Merchandising Optimization
For too many brands and retailers, the ecommerce storefront is little more than a container for the product catalog, with every category and sub-category behaving essentially the same (save a few differences in filters and facets).
But customers shop for different products in different ways, and not all category pages should be merchandised the same way. We shop for high ticket electronics in a very different way than accessories. We shop for clothing in a different way than jewelry, though a brand may offer both categories. The purchase criteria for home decor is very different than home improvement supplies. The list goes on…
Category Merchandising Optimization (CatMO) is the discipline of developing and testing product ranking strategies at the category (and sub-category) level, based on what factors most influence a customer’s purchase decision in a given buying context.
Category merchandising strategies
Your ability to customize and optimize category page design and product list ranking depends on your ecommerce platform. If you’re on a home-grown system, you may be able to support category merchandising strategies with custom development, though it may be cost-prohibitive and difficult to make updates to your ranking rules at the speed of business.
If you’re on a commercial ecommerce platform, a degree of category-level merchandising features may be available to you out-of-the-box, such as manual pinning of featured products or simple attribute-weighted ranking. Options like category-specific default sort may be possible with custom extensions to the platform.
There are several third-party, bolt-on merchandising tools like Jirafe, BloomReach, Peerius and Merchandising.io that offer their own sets of merchandising capabilities, such as boost-and-bury (merchandising rules for weighting product list results by attributes), A/B testing, predictive search, personalized filters and facets, and social integration. These tools sit on top of the ecommerce platform, and put merchandising control in the hands of the business user without the need for IT.
The following tactics are categorized into what you can do at a basic level (with minimum merchandising features in your ecommerce solution), intermediate (with out-of-the-box platform features) and advanced (with a third-party solution).
Basic Category Merchandising Tactics
CAMA, filters and facets
Begin with good CAMA (Catalog and Merchandising Architecture), with the most appropriate filter and facet options for each category.
I recommend beginning with a category-by-category experience audit, conducted by a team member or consultant that understands consumer behavior within the category (think like the customer)! You may choose to include your buyer or head of merchandising in the audit, or a top in-store salesperson for each department.
Look for filters and facets that may be missing. Hudson’s Bay’s Earrings category contains 1215 items, but the refinement options doesn’t contain style filters like stud, hoop, chandelier, drop, or clip-on — all attributes that matter, even when “just browsing.”
Use your site search reports to uncover any filters and facets your site is missing. When category browse fails, shoppers turn to search engines to narrow their product set to their purchase criteria.
Hudson’s Bay’s Tie category does apply style. Site search reports may reveal that materials like silk, vegan-friendly imitation silk, linen and cashmere are important attributes not included in category refinement options.
Consider promoting frequently searched for attributes, and frequently clicked filters to sub-categories to further improve your CAMA (and boost SEO).
Default category sort
Depending on your ecommerce platform’s capabilities, you may be able to control default sort at the individual category level. If you can only apply a single default, consider what’s the best for your product type and customer across all categories.
Best Selling is a popular default (it’s Amazon’s sort-of-choice), but keep in mind that for thematic categories (those constructed not by product type, but by themes like “Back to School,” “Gifts for Him,” or “Mad for Plaid”) often contain products from a number of categories, and their relative popularity can give them an unfair (and perhaps undesirable) advantage over more “relevant” themed products.
For example, Oriental Trading created a thematic category for the adult coloring trend. Though adult coloring books are more relevant (and higher dollar value) than coloring pens, the default “Best Sellers” sort returns equal felt pen results as books.
If, in this scenario, the merchandiser cannot control the experience with pinned, featured products or otherwise control the ranking (for example, setting the “Recommended” rule to put highest price first), a workaround could be to remove tools from the category, and instead show them as cross-sells on the coloring books’ product detail pages.
In Oriental Trading’s case, there is a “Relevance” sort option, which ranks coloring books first. Ideally, this would be the default sort for this and other thematic categories.
Note that “Relevance” is an irrelevant sort option in categories, it’s more “relevant” to search results. Consider using Featured, Our Picks or Recommended in lieu of Relevance or Position.
There is rarely a case where sorting by brand A-Z ever makes sense. Brand is a filter. (Much less should it ever be a default sort).
In the example of a Ukeleles category (with 16 pages of paginated results), the first products to appear sorted by brand are lower priced starter kits.
If enough kits were shown before actual instruments, a first-time or mobile visitor may conclude the site doesn’t carry ukeleles, and abandon the site.
Another pitfall of showing A-Z brand by default is it can impact engagement with and sales of some brands that always appear last in the list.
London Drugs’ Cameras page ranks by A-Z, and unwittingly ranks products pre-order products first. (Audit every category!)
As a rule of thumb, Best Selling and Featured (your custom ranking) are typically the best default sort options for non-thematic categories, and Featured is the best for thematic across the board. If you have the ability to set custom defaults for each category, you may sort Clearance categories by percentage off, or apparel by New Arrivals, depending on your access to sales trend data and buying behavior.
Beware of ranking by “Top Rated” by default, as low sell-through products that happen to have one glowing five-star may outrank more relevant and popular items.
During the audit process, you may determine some categories would benefit from custom “landing page” treatment, with different design, content or featured filters (like graphic sub-category tiles) to help guide the customer to more relevant results.
Like a good salesperson, Oriental Trading’s Wedding category page’s sub-category tiles suggest “we’ve got a LOT of stuff here, how about you tell me what you’re looking for.”
Spotlight content can also be featured contextually (or con-productally) in your category list. Benefit’s cheeky callout for its bundle builder and Lululemon’s waterproof sub-category are examples.
Toys R Us features spotlight selections, but visually separate them from results with an ad slot. (Advertising for other retailers anywhere on your site is a conversion killer, avoid it).
If you call out specific product features, ensure they are selected based on very strong merchandising rules and are intuitive as to why they are featured. Adding context like “top rated” or “our picks for you” help would be appropriate here.
Intermediate Category Merchandising Tactics
Depending on your current ecommerce platform, you may be able to apply some or all of these intermediate tactics to your category experience.
The ability to manually pin featured products to the top of category results provides control, but is labor intensive. Unless a merchandiser is dedicated to regular management, the relevance of pinned products can decay over time. And unless they’re selected by solid data or reasoning, they may not be relevant at all – though some platforms may support A/B testing of pinned products.
Boost and bury
“Boost and bury” allows merchandisers to algorithmically weight product results based on specific attributes, which can greatly improve the user experience and improve conversion rates. For example, a retailer may choose to “bury” items for which most size or color variants are currently sold out, or “boost” a retailer’s house brand or other higher margin products.
Boost and bury rules may also be applied seasonally. Oriental Trading gives Easter candy a temporary boost in results in late March.
Similarly, an apparel retailer may wish to “bury” spring and summer apparel during the fall and winter season, and vice-versa.
Some ecommerce platforms support “boost and bury” algorithms out of the box or through marketplace extensions built by third-party developers. Third-party tools like Jirafe track product-level engagement and revenue trends, such as click-through, sell-through and revenue per visit. These metrics help merchandisers better select which products to boost and bury, and which to promote in email, paid search and social campaigns.
Default sort, by category
Ecommerce platforms that don’t support boost and bury may be able to support category-specific default sort rules out of the box or with some custom coding.
A merchandiser may consider ranking categories differently based on why customers buy them. For example, Clearance and New Arrivals categories would benefit from default sorts “Percent off” and “Newest,” respectively.
Work boots are less about fashion and more about function. A shoe retailer may choose to sort for best selling work boots, but newest dress shoe styles. A beauty retailer may show best selling perfumes first, but newest cosmetics. (If you’re tracking engagement with sort features at the category level, your customers may be telling you what matters most to their purchase decision).
Ninja Category Merchandising Tactics
These tactics typically require extensive custom builds or third-party tools to execute.
Tools like Visenze and Allyke support “more like this” filters from product page results, which both re-ranks products based on visual similarity, and remove styles that are not similar.
Planet Shoes uses Allyke to surface similar styles with one click from the product list.
Visual filters are much faster than applying multiple faceted navigation selections, such as the combination of heel height range, boot height, color family, shape and pattern, which may refresh the page after each selection is made, and depend on the quality and consistency of product data. They’re also helpful to international customers that may not understand facet labels like “mary janes,” “lug bottom” or “slingback.”
Visual technology also does the job of ranking results in order of similarity to what the customer likes, intelligence that simple faceted navigation doesn’t apply. More relevant results leads to faster and more confident buying decisions (especially on mobile). For Planet Shoes, customers that interact with “view similar” buttons convert 10% higher than those who don’t.
Contextual boost and bury
Some ecommerce platforms and third-party tools support contextual boost and bury. For example, bury online-only products for mobile shoppers geolocated inside or near your physical stores, and boost items in-stock in store. Likewise, bury brands that can’t be shipped internationally to international visitors. Or, bury colder-weather clothing for visitors in warm cities, and boost them in colder climates. In advanced tools, rules can be written to algorithmically rank products by sales velocity within a geographic region, or across similar user demographics or profiles.
Tools like BloomReach combine machine learning with rules-based logic to algorithmically optimize category results in real time – even down to 1:1 personalization based on retailer-defined segments, past purchases or on-site behavior. For example, if a visitor filters the Michael Kors brand in Dresses, Michael Kors purses receive a boost when she visits Handbags. Other contextual cues include sorting by price, engaging with facets like color, size or style, visiting the Sale section first, “liking” certain products, or adding them to a wishlist.
Benefits of Category Merchandising Optimization
Maximizing merchandising productivity simply boosts profits. Move full-priced inventory more quickly, and quickly discount laggards before they become unsellable.
Customer satisfaction and competitive advantage
One way to compete against Amazon and your online competitors is to deliver a tighter, more relevant set of products across your site. The better the product discovery experience, the more satisfied your customers, and the more likely they’ll purchase and come back.
Better marketing and merchandising campaigns
Data gleaned from merchandising tools tracking engagement and sales can be used to improve email, social, paid search and remarketing campaigns, and can influence the creation of on-site look books and thematic categories.
Need help with your category merchandising strategy? Drop me a line.
Ecommerce Illustrated is a project of Edgacent, an ecommerce advisory group.