Does Your Chatbot Deliver the Answers Your Customers Need?

As search engines continue to improve their results and learn what we're actually searching for, we keep getting more specific with our queries. Now we use long strings of keywords in natural language phrasing (like "dermatologist in New York City who takes Cigna") and even ask actual questions ("Where's the nearest emergency room?"). And we bring this expectation for answers wherever we go. So whether we're typing into Google, asking Alexa a question, or searching on a brand's website, we're using phrasing that increasingly resembles the questions we'd ask a person — and we're expecting the machines to answer accordingly.

What does that mean for chatbots? Chatbots have played an important role in reinforcing this desire among consumers to ask complex questions — and have also helped brands meet that demand. As brands have recognized that consumers want to engage conversationally, they've used chatbots to help with questions that the rest of their website may not be able to answer. According to Search Engine Journal, chatbots "are a very well-timed middle ground between automation, AI, and the changing use of the internet, in a way that could make it extremely effective in doing business online and ushering in some of the other future technological advancements we've all been waiting on." When consumers can't find answers on your website by browsing or searching, they are likely to turn to your chatbot — hoping it will be able to help them. They expect it to answer complex questions like a human would. But while consumers bring their conversational habits to the chatbot, they're often disappointed when it doesn't meet their expectations.

Consumers come to your website for three reasons:

1. They want information
2. They want to contact you
3. They want to take action (e.g., to transact, to book an appointment, or to register for an event)

Of course, it's not always as simple as doing just one of these things. Sometimes you need information before you can transact, like when you're researching doctors or financial advisors before booking an appointment with one. A chatbot works best when it makes its capabilities clear by telling users which of the three functions it serves, so that users won't be disappointed if the chatbot can't deliver what they want.

Site search, on the other hand, can deliver information, help consumers contact you, and let them transact with you, depending on the results it returns. It presents an opportunity to fill the gaps that chatbots can't cover — and to exceed consumers' expectations whenever they're seeking answers. People are used to receiving direct answers to their questions from search engines like Google and Bing. We can ask Alexa for the phone number of our nearest pizza place; we can get a doctor's address easily from Bing; and Google can show you a "buy now" button for that pair of sneakers you've been eyeing. Search engines have evolved with consumer behavior to deliver more accurate and immediate information in response to the consumer's specific need — and your website can do the same thing.

A website with robust search functionality improves the consumer experience by helping people meet their own needs the first time — without having to default to a chatbot or call your support hotline when they can't find the answer. That leads to a higher rate of transaction because people are able to find those calls-to-action in fewer clicks, immediately after they've found the information they were looking for. And support costs are lower because consumers aren't using costly resources to get answers they could have found themselves.

So what does a great search experience entail? First, it's able to parse out a consumer's intent using natural language processing so that it can deliver information about what the person actually wants — not just a list of pages that are tangentially related because they contain similar words. Second, it should deliver direct answers wherever possible, so if someone asks for "Dr. Adam Kunin's phone number," the first and most prominent result is Dr. Adam Kunin's phone number. Third, the answers it returns should allow consumers to take action right there in the results page. The user should be able to call a doctor, get directions, book an appointment, RSVP for an event, or place an order immediately. By structuring your brand data in a knowledge graph, you can respond to this demand and weave direct answers into every consumer experience on your website — from a doctor finder, to an events calendar, to a store locator.

Natural language processing will continue to evolve and improve until one day we'll be able to have a real conversation with a machine. Until then, site search solves the most immediately pressing problem brands face in giving consumers the ability to find information, to contact them, and to transact with them using an interface that's familiar. In a world where every customer journey starts with a question, and reputation means revenue, robust search functionality is an investment that pays off every step of the way.

Read more: The 5 Elements of Great Site Search