NETTING IT OUT

Vendors of premium search technology claim their ecommerce site search solutions will produce a better search experience and differentiate your store from those of your competitors. Our study of four online department stores, Bloomingdales.com, Macys.com, NeimanMarcus.com, and Nordstroms.com, indicates that premium search engine technologies produce equivalent results. We performed a small set of searches across each of the sites to demonstrate the search experience of each.

These stores use four different premium search engines. And yet, they exhibit essentially the same site search behaviors, and score both hits and misses for seeker experience.

The differences lie not in the search engine technology deployed, but in the management of the search experience. The stores vary in their search philosophy, with Macys.com valuing precision (showing exactly what the customer asked for) while the other three opt for breadth (showing everything the store has to offer, irrespective of relevance). While we understand the urge to “show ‘em what we have,” we love the precision of Macys.com and are frustrated by the sloppiness of the broad results. When we ask for men’s trousers, we don’t want women’s sneakers or baby clothes. The stores also differ in their navigation philosophy, with Macys.com and Nordstrom.com opting for the most consistent options for search refinement, and Bloomingdales.com the most adaptable.

DIFFERENTIATING CUSTOMER EXPERIENCE VIA SEARCH EXPERIENCE

Search Experience Is a Critical Component of Customer Experience

Search experience (SX) is fundamental to the online customer experience (CX). If a customer can’t find the answer, the information he’s seeking, or the product he needs, as far as he’s concerned the supplier doesn’t have what he’s looking for. If he does find what he’s looking for, then his relationship with the supplier deepens.

Our definition of a terrific search experience has these features:

* Supplier appears on first page of Internet search results, and the transition from Internet search to the right page on the Web site is seamless

* Site search delivers the right item in the first five results of the first search

* Site search offers search-specific navigation choices that deliver the right item in the first five results of the customer’s first navigation choice

Our definition of a terrific SX is pretty stringent, so not surprisingly, few customer search experiences stack up. Of course, each of us would like to be the company delivering the terrific results, while our customers found our competitors’ search experience fruitless. Customers would flock to our site because we not only have the goods, we can deliver the right information about them.

But what does a company do to deliver a great search experience? There are several components you must get right in order to deliver a great search experience, and each has the potential to make a difference. The key components are:

* SEO and SEM (search engine optimization and search engine marketing)

* Content and product assortment

* Search engine technology

* Design and integration of the search experience

* SX management

Search technology vendors would like to be the key to differentiation, but the evidence to date does not entirely support that claim. In our research, we see that premium search technology is critical to a terrific seeker experience. But the choice of premium search engine seems to have only slight impact on the quality of seeker experience.

SEARCH EXPERIENCE COMPARISON: DEPARTMENT STORES

Let’s look at some examples. Four prominent U.S.-based department stores, Bloomingdales, Neiman Marcus, Macy’s, and Nordstrom’s use four different premium search engines: WebSideStory, IBM WebSphere Content Discovery Server, Mercado CSN, and FAST ImPulse. The design of their sites, search results pages, and product detail pages have the same behaviors and features. Navigation tabs reflecting the store’s departments march across the top of the page. Site search results provide a column of navigation choices on the left. Products are represented by thumbnails, and clicking on the image brings you to a product detail page, where you can zoom, read more copy, and view colors. See Illustrations 1-4.

In our example, we actually started our search looking for “men’s plain front chinos.” This proved to be too much for several of the sites, and after some experimentation we arrived at search terms that produced results at all four Web sites. We started with “men’s plain front chinos,” then tried “men’s trousers” and finally “men’s slacks.” We chose to avoid “pants” in deference to our U.K. audience, who wear “pants” under their trousers. We also tried our search for chinos on Google. A summary of our searches is presented in the Table.

The experiment revealed quite a bit about how each store manages search. Here’s what we saw.

Bloomingdales.com

Bloomies had utterly consistent results across all three searches. The thumbnails included non-trouser items, but at least they were men’s items. We can observe that Bloomingdales.com merchandising staff has made sure that synonyms exist for trousers, slacks, pants, and chinos.

One might surmise that rules controlling search at Bloomies dictate that the “men’s” search term is so important that it constrains the search to the men’s category. However, a search for “men’s polo shirts” produces a list of women’s blouses. Scratch that theory. And scratch our opening comment on consistency.

As for the transition from Internet search to site search--Bloomingdales.com doesn’t appear on the first page of Google results for “men’s plain front chinos.”

Macys.com

Macy’s merchandisers failed to establish synonyms equating trousers, slacks, pants, and chinos, thus driving the customer to find her own synonyms.

Once past that hurdle, however, the search experience was top notch. The search results for “men’s trousers” were all men’s items, which is good. The navigation options let you zero in on precisely what you’re looking for––in this case, “plain front.” The results also included a shirt and sweater that one designer features with his jeans, which is a great cross- and up-sell offer––the items are potentially extremely appropriate to the customer’s wardrobe.

Other searches at Macys.com demonstrate that the left-side navigation choices are fairly consistent within the category, but differ markedly from category to category. Clearly, merchandisers have decided that having a consistent set of navigation choices within a category is more comfortable for customers than item-specific navigation. Thus, in men’s trousers results, “sleeve length” is a navigation option, and in men’s polo shirts results, “plain front pant style” is an option. The latter option, by the way, takes you to the same items you’d see searching on “men’s plain front.”

Macys.com had the best showing on our Google search, showing up in third position, and as the lone organic result for “men’s plain front chinos.” Score one for Google: it found the items that the Macys.com site search failed to see.

NeimanMarcus.com

Search results at Neiman’s are all over the place, and navigation doesn’t help much. The search for “men’s trousers” produced a list comprised mostly of trousers; the searches for chinos and slacks produced mostly non-trouser items.

The navigation options allowed you to narrow the results by “him” or “her,” by “maternity,” by “sneakers” or “loafers,” and among four price ranges of $3,000 spans. What sort of men’s trousers would be “maternity,” “sneakers,” “for her,” priced “$12,000-$15,000?”

The navigation choices offered for the “men’s slacks” search differed markedly from the navigation choices offered for the “men’s trousers” search. Clearly, the consistent navigation valued by Macys.com is not prized at NeimanMarcus.com.

Neiman’s, like Bloomies, hates to impose any constraints on search, apparently in the belief that the more they show, the more you’ll buy. If you take the “pants” navigation option to refine your search for “men’s trousers,” you get 11 pages of women’s slacks. Not a single menswear item in the lot....