Most people looking for a product on a website start off with the search box. These customers experience bad Findability in 4 ways:

1. Too many results
2. No results
3. Missing results
4. Wrong results

1. Too Many Results

This is the most universal, as well as obvious, problem. A customer visits your website, enters a one or two word search and is drowned by an overwhelming list of products to examine.

 Most E-Commerce Sites that use faceted navigation (or guided navigation) to allow the filtered refinement of an initial search provide both Brand and Price. These are easy to provide as price is universal and almost every product has a Brand – usually and conveniently provided by the manufacturer in a separate column (field). Unfortunately, neither of these two filters alone or in combination is typically sufficient to narrow a product selection, and the extent and quality of filters beyond these two basic ones are highly variable.

a. Few sites provide all of the filters relevant to a customer’s decision

Even those that do rarely provide any guidance as to what choices are appropriate to a customer and what the trade-offs between different choices are.

b. Filters are often generic

Classic examples are “Type” and “Features.” These filters are so broad that virtually any attribute can be placed there. What this means to the customer is that filter values refer to more than one quality of a product; e.g. scent and color, in this case meaning that customers can’t pick first by scent and then by color.

c. Filter values are too numerous

We have seen as many as 165 choices in one filter. The more choices that exist in a filter, the less likely the filter is going to be useful to a customer in reaching a decision.

d. Filter values are not linked

On one shopping engine we looked at recently, a customer searching for “coffee maker” got over 2,500 results. One of the values under the “features” filter was “Burr,” indicating coffee makers that included a built-in Burr coffee grinder. Clicking on “Burr” displayed only 1 coffee maker. We found 45 Burr coffee makers without too much difficulty and if we’d gone through all 2500 results we’d have doubtless found more.

2. No Results

The other half of the “million or none” problem, returning no results for a product that you actually sell – typically when a customer uses a highly specific search (see Known Item Search) – mean no sale.

3. Missing Results

In a sense, this problem is more pernicious than no results. A customer that doesn’t get any results may try another search, but how many are going to recognize that some results are missing? In the coffee maker example above, although we counted 45 results for coffee makers that included a built-in Burr coffee grinder, a search for “burr coffee grinder” yielded only 19 results.

4. Wrong Results

At least with wrong results, a knowledgeable customer will realize that the search didn’t work. But it sure makes you look foolish! A recent example we saw is on a site that sells stationery supplies. A search for “pens” yielded paints – even though navigation by category indicates that there are hundreds of pens available.