Defining Two Types of Local Searches: Recovery and Discovery
by Gib Olander

When it comes down to it, there are two main types of local search queries that a search engine needs to be able to handle. John Battelle’s book The Search describes them as follows: “This perfect search also has perfect recall. It knows what you’ve seen and can distinguish between a journey of discovery, where you want to find something new, and recovery when you want to find something you’ve seen before.”

In this article, we’ll explain how these two searches work and why they’re both important for businesses.

Recovery searches

There are many times when people using a local-search application are trying to recover a bit of information or the address of a business that they know exists. It’s easy for consumers to find an unhelpful recovery search dissatisfying because they know the item exists.

A 2002 survey of search-engine users by iProspect drives home the importance of effectively fulfilling recovery searches. The iProspect survey noted: “If a search is deemed unsuccessful by the user, 27.2% of users immediately switch to another site/search engine.”

The problem is that in today’s world, nearly 50% of all businesses do nothave a Web presence. Without a Web site or HTML-based profile, these local businesses don’t have any listing that the various spiders, bots and slurps of the world can find. In other words, if left to their own devices, the search portals would fail at a recovery search.

To address this problem, search applications have begun licensing databases (such as the Localeze local search repository). These databases contain the “base” information on all US business locations, whether they have a Web presence or not, allowing a search application to fulfill recovery searches. The business attributes that make up “base” data center around business names, business address, phone numbers and business classifications.

Discovery searches

The second type of search, a discovery search, is much more difficult to address but it is just as vital for a successful local-search application. In the local-search sense, a “discovery search” is when a person has a need and knows where, geographically, she or he wants a product or service. But the user doesn’t know which business to turn to in order to get that particular need met.

For example, consider a person who wants to find a one-hour dry cleaning service in ZIP code 60606. This person knows what they want (one hour dry cleaning) and knows where they want it (ZIP code 60606). However, he or she doesn’t know WHOprovides that particular service in that ZIP code. After all, all dry cleaners do not provide one-hour service. The person using the local-search application is trying to discover the answer to her or his question.

Answering discovery-type local search queries poses problems for any search application without a complete local-search database. This problem again stems from the fact that most businesses still do not have a Web presence.

Those businesses and their unique qualities are invisible to a Web-crawled index. To successfully fulfill a discovery-type local search, an application needs more than just the name, address and phone number.

A local-search application needs to understand the fabric of a business and identify the particular thing that makes a business unique. People making local-search queries want context that allows them to understand what each business does to distinguish itself from other businesses in their same category and same geographic space.

The information needed to understand what is unique about a business range from the simple — hours open and payment types accepted — to the more complex attributes like a business’ specialties and the products it carries.

Does your local data work?

A local-search application should have an understanding that there are two types of local-search queries to answer. This level of technology will help you evaluate the quality and depth of your index or database. Ask yourself:

• Can you answer both a recovery or discovery query?
• Will your application allow your users to recover information they know exists today? For example, can they recover the phone number of their favorite pizza place or dry cleaner?
• Is your search index deep and complete enough to answer unusual queries?
• Will your search result offer a user relevant data beyond a category list of businesses?
• Can you help your users discover the one-hour dry cleaner in their neighborhood?

If you can’t answer these questions with a “yes,” then the Localeze local-search database will help.

If you have a business location, or you represent a business with physical locations, please take some time to evaluate whether or not you have made enough information public for a search application to find you for recovery and discovery searches. If you don’t, you could be discouraging users from returning to your site rather than building their trust.