Shopping carts often provide a search capability. A customer can search for a product on your Web site. What do these points have in common? They all use a search "query" (question) to work with a database.
What`s a database?
A database, simply put, is a list of things. Each thing is categorized by a number of descriptive lines. So, for example, an umbrella and a giraffe are two things. When they`re listed in a database, someone has to figure out what "lines" (fields) to create that will describe all the items in the database.
Remember that game 20 questions? A common starting point asks, "Is it animal, mineral, or vegetable?" Why that question? Because it separates living from non-living things.
In the case of our umbrella and giraffe, we would have a "line" (field) in the database with a dropdown box. When we list our item, we describe it in that line. The umbrella is "non-living," and the giraffe is "living." The question is whether or not this particular field is "searchable." Is it included when the database goes out to try and find on your question?
A database is a software program, and there are many of them in existence. Not all databases are well designed, but ALL databases include some way to ask questions---to form a query. When you see anything on a computer screen that offers you "Search:" and a window, you`re dealing with a database and its query "engine."
Not all databases are equal. Some of them search every word in every field in every record. A "record" is the whole description of the thing, but only that single thing. (The database is the group of all records.)
Google goes out and gets every word on at least the first three pages of your Web site. So if you search for a word and it`s anywhere in those pages, you`ll get a "result"---something found. On the other hand, eBay only searches the main title (not the sub-title). It appears not to search the descriptions at all.
Some databases allow "wildcard" searches. If you search for man* (with an asterisk wildcard), you`ll get results for "man, mannequin, mandible, manifest, Manchurian," and so forth.
Google uses programming for "smart searching." If you put in `man` (without quotes), Google looks for the letter M-A-N in whatever form. It also rates the letters by relevance, assuming that you mostly want to find a human being. So "man" is ranked higher than "roman," or "manifest."
Constrained or Loose searching
A good database doesn`t much care about the order in which you type words in a question (query). So if you put in "flag signal navy" it`s the same as "navy signal flag." But many online search engines and e-commerce databases do care!
That means if your product is "navy signal flags" (with an `s` on `flags`), and someone types in "navy flag" the database is "too stupid" to find your product. Some are so dumb, they differentiate between uppercase and lowercase letters as two, entirely different "things."
When you create a title for an item, then a description, think about how database programming works.
- Does the e-commerce database search allow wildcards? Most of them don`t, nowadays.
- Does the order in which words are typed matter? In many e-commerce sites, exactly that order is the only way an item can be found.
- Does the site search the rest of the description, or only the
title? Many e-commerce sites only look at the title and perhaps 1 other
field, usually not the description.
Above all, think about the title for each of your products! Include as many "key" words as you can, even if it makes the title seem a little strange or repetitive.