The Empathic Search Engine

Over the years observing customer behavior offline and online, we’ve noticed many similarities in the way stores and Web sites work. There are many things you can learn from offline best practices to improve the customer experience online. One of them is how you can improve your Web search engine by giving it some of the behavior a good staff in a store would have.

Let’s compare the shopping customer experience in 2 stores.

The Bad Experience

A customer walks in a grocery store and decides to get the help of a staff. No staff is to be found near the front of the store. The customer then walks to the back to find a staff who is busy organizing some products on the shelves.

  "Excuse me", the customer tentatively asks the staff.
  "Just a moment" replies the staff, as he finishes placing 2 more products on the shelves. After 10 seconds, the staff just looks up inquisitively. Taking this as a cue she can proceed with her question, the customer asks: "Do you have any crapers?"

The staff just looks at the customer, bewildered, and simply replies: "Hunh? Nah, none of that, I don’t even know what that is."

The customer leaves the store, disappointed.

The Good Experience

The same customer walks into another grocery store. As she seems to be looking around for something or someone, a staff greets her.

  "Hello, can I help you with something?"
  "Oh, yes, thank you. I’m looking for some crapers."
  "Crapers? Do you mean crepes, or capers?"
  "I don’t know, they’re small salty things shaped like peas, people eat them with smoked salmon."
  "Oh yes, capers, we have them, they are near the canned products and pickled vegetables, in the canned food section. I will show you."

The staff also suggests that with smoked salmon and capers, people also usually buy some lemon, and onions, which they will cut in very thin slices. Pleased with the staff’s helpfulness, the customer buys a small jar of capers, as well as onions and lemon. She will come here again.

Main Difference: Understanding Customers

What was the main difference between the 2 shopping experiences?

It was not the staff’s knowledge of the products and their location. The staff in the first example knew what capers are, that the store has them and where they are located. But he wasn’t open to customer needs.

The quality of the customer experience between the 2 shops depended on the staff’s empathy; his or her attention to customer needs, his or her ability to understand what the customer wants, what the customer means, and to ask the right questions to better understand the customer.

Making a Search Engine Understand

On the Web, the search engine replaces the staff. If people can’t find what they are looking for using your navigation, they might use your search engine.

For a search engine to be truly helpful and provide a better user experience, it has to go beyond the simple ability to find exact matches. It has to understand users.

Although you can’t have human staff answer Web search queries on the fly, there are a few ways to imbue your search engine with a little more empathy.

Although there are a number of excellent commercial search engines available, most are based on crawling and indexing content pages, are not very customizable, and do not cater well for database-driven product searches. If you have your own search engine, or if you can modify the one you bought, implementing some of the following recommendations could make the difference between customers buying your products, or finding them elsewhere.

Make the Search Easy to Find

For customers to get help in a store, they have to find a staff in the first place. The search on your Web site should be easy to find. If you can’t place your search on every page, at least place the search box right on the homepage and link to it from every page.

Simple Search

Unless you are sure 80% or more of your customers use multiple search criteria at the same time, offer a simple search to start with. Too many search options will just be confusing to the users. Relegate advanced search options to an advanced search page.

In stores like on the Web, most of the time people will either ask for a specific product ("M2000"), or a product category ("phones"). A simple search box will cater to both.

Understand Users’ Language

How to name a concept might differ between the company and customers and between customers themselves. You should allow users to enter what they are looking for in their own terms, and provide results with your matching products. For example if you sell computer display adapters, you might want to match the terms "display adapter", "video card", "graphic card", "video adapter", etc.

One way to do this is by implementing a thesaurus that checks for synonyms of search terms. If you have a limited range of products or categories, you can build a thesaurus-like application yourself by manually adding to your search engine alternative wording of the same product.

Forgive Mistakes

Inevitably, some users will enter a search term with the wrong spelling, and might not notice they did so. Instead of giving them an error message, be forgiving and understanding. Either you can add common misspellings in your thesaurus, or implement a spell-checker separately. This is how many search engines are able suggest another search with the correct spelling. For example, if you search "recepies", you will get "Did you mean: recipes?"

Help Your Customers Understand Your Categorization

If there is more than one category matching the search term, don’t simply display a list of unsorted, uncategorized search results that point directly to a product or content page. You can help your users pinpoint their product by providing an option to refine their search with a list of related categories above the search results. This will also help them better understand the structure of your site and classification of your products.

For example, if you are a bank and have insurance products, the results for the term "Insurance" can provide links to categories like "Business Insurance", "Travel Insurance", "Home Insurance", etc.

Learn from User Behavior

In some cases, a computer can do a better job than a human could, especially in statistics… Record the user activity with on search engine; what users have typed, the number of results and the ensuing clicks or searches. Analyzing the data manually can immediately reveal improvement opportunities.

One thing you should focus on first is the most common no-results wielding searches. These might be different names for products you do have. It might also provide you a hint on what users want, which you do not yet have.

You can also somewhat automate the process of suggesting alternate searches by performing longitudinal (over time) search pattern analysis. For example, when faced with little or no results, most users will attempt a second search with an alternative wording. Recording all of these you will give you enough patterns over time to automatically provide alternate search terms and results.

Most companies regard their search engine as a simple lookup tool to which they pay little attention, assuming that people will browse using navigation. But at the same time, it seems that most companies still don’t understand how to group things from the customers’ perspective, resulting in customers never finding what they are looking for.

Just like in stores, some users on the Web prefer asking for help (using search) right away, while others search only as a second option. Regardless of the quality of your site’s information architecture, you could increase your revenue opportunities by implementing these simple user-understanding features.

Tagged: | search engine, User Experience, Web