Using search to boost sales in ecommerce : What to look for in an internal search engine (episode 2 of 3)

In this second episode of our series, we delve into the challenges of relevance in search engines. You'll discover how an effective search engine must accurately interpret user queries, provide relevant results, and know when to broaden the suggestions if needed. We also discuss the importance of guiding customers by suggesting appropriate terms and facilitating their navigation for more satisfying results.

Meeting the challenge of relevance

Your search engine is capable of finding many products via a multitude of attributes. However, when internet users type the words “extension cable”, do they really want you to offer them a lawnmower that comes with a 25m cable?

Often, less is more. Your search engine must know how to extract the right information to refine the relevant results. It must be able to extend the search if needs be.

When the customer has made the effort to put forward several words, you need to be able to take this into account and respond specifically. If this search doesn’t return any suitable results, then the search engine must be able to broaden the range of products offered.

For example, if the client searches for “black faux leather trousers”, it should never return black trousers, purple leather trousers and bags in a single page of results.

Customers need to feel that their request has been understood. What can be more frustrating than to be offered items that are completely irrelevant?

Of course, the first impulse for all online merchants is to offer something, even if it doesn’t correspond exactly to the product that the customer is searching for. This is not a good idea.

Putting forward just any product shows that you haven’t understood the request. It’s better to admit that you don’t have what they are looking for.

Of course it’s possible to offer alternatives for a specific search which has not brought up any results from your catalogue, but these should always be presented as such.

For example, if you don’t have any “brand X vacuum cleaners,” you can, having first told your customer that you don’t have any, offer other vacuum cleaners in the same style.

Help the customer to explore your product offering (suggestions, auto-completion, navigational aids)

Customers don’t want to be left alone with a search box, they want assistance!

“sports shoes” Which sport are we talking about? Are they for daily use? What’s your favourite brand? You can’t just bring up all of the 3,000 types of trainer that you have in your catalogue to give your customer what they want. Your search engine needs to suggest the right terms, based on the complete knowledge of your catalogue, in order to allow the customer to refine their search.

Just as a good salesperson might notice that many of the customers who look for sports shoes end up buying running shoes, the search engine should learn from the behaviour of previous users. It should be able to identify the various different choices and offer them to new customers, in order to make their search easier.

Through the search engine, encourage the internet user to specify what they want. Ask qualifying questions and open-ended questions to avoid them ending their search because of a misunderstanding.

Some answers will be firm choices, others will be an expression of preferences. The search engine needs to differentiate between the two. Choices must be definite criteria, while preferences will be optional.

The design and usability of your search interface are the keys to success!

The customer must be in no doubt about the implications of a choice or a preference. You should differentiate filters visually, for example, or suggest words that are more open.

At the same time, be careful not to depend too much on “faceted results”. Faceted searches currently have an average usage rate of 10 to 15%. Your search engine must be cleverer and offer advice in a more specific manner, without getting lost in a set of overly precise criteria. It’s important that your customer has confidence in your ability to provide a relevant result. Customers need to feel that the engine understands each of their actions all of the time. The response time of the search engine is a major factor; instantaneity is crucial.