Join top executives in San Francisco on July 11-12, to hear how leaders are integrating and optimizing AI investments for success. Learn More
The internet has put the whole of human knowledge at our fingertips. Unfortunately, finding just the right piece of information quickly and easily has become like finding the proverbial needle in the haystack. In an era when so much content is so readily available, we’re forced to ask ourselves: How do I choose what to click on first? Is this a trusted source with reliable information? And how much time do I want to spend looking?
As a regular person looking for a basic answer, this flawed process adds time to your journey. As a consumer, a broken knowledge management strategy can make interacting with a brand frustrating at best — which in turn can mean an abandoned purchase, a degradation in brand loyalty or even outright anger that can translate into negative reviews.
The good news is that a solution exists right under our noses: By taking a cue from the gold standard of search (Google) and instituting a system of knowledge graph-driven information management, brands can provide customers and their support teams with the answers they need in the most straightforward way possible.
What is a knowledge graph?
The concept of knowledge graphs is intuitive to humans because it’s based on understanding the context of different segments of a question. For example, if I ask a friend, “Do you have a recommendation for a pediatrician in town who speaks Spanish?” they understand that a pediatrician is a type of doctor, that “in town” means “nearby,” and that Spanish language proficiency is required.
Join us in San Francisco on July 11-12, where top executives will share how they have integrated and optimized AI investments for success and avoided common pitfalls.
But making these connections has been difficult for machines until relatively recently. Enter knowledge graphs: A way of organizing and connecting different categories of related data — known as entities — so they can be easily “understood” by various search algorithms.
Think of these entities as databases of information unto themselves that a search query can draw from. To give another example, if you were searching for information in a school system, separate entities could include personnel, classes, extracurriculars, buildings and class numbers. With this framework, a knowledge graph connects disparate groups of data based on the context of the search query.
If a user were to search for: “Where is Mr. Johnston’s third period history class?” a knowledge graph will use each part of that question in different ways: “where” denotes location, “Mr. Johnston” denotes personnel, “third period” and “history class” denote time and schedule.
Connecting all of these different datasets into one query — based on the natural language of the user — enables the search engine to combine the data in just the right way to deliver an exact answer. In traditional search, this query would simply pick out key terms and deliver a list of results, which may simply be links to articles or other information sources, rather than a straight answer.
For brands, knowledge graphs are vital for connecting informational content of different types that exists across numerous platforms, including content management systems, customer relationship management platforms and other information sources. With brands investing so much in content, it’s frustrating for everyone when a customer needs to reach out to support because a search wasn’t sophisticated enough to find answers that already exist within the site.
Making answers findable and knowledge discoverable
When knowledge graphs are deployed successfully, they make answers findable. But what exactly does that mean?
Again, we can look at Google for the answer to that question. When you provide Google with a specific question, it has the ability to give you the answer in a featured snippet along with a structured info box of related information. This is a feature you’ve seen time and again; searching for “How tall was Andre the Giant?”, the results present a simple response with his height — 7’4” by the way — rather than a series of links to articles and websites that contain a reference to his dimensions.
On a brand website, these dedicated info boxes can pull from a knowledge graph built off of information contained in product manuals, articles, FAQs, support documents (and more) to offer usable answers in context for the customer. So, if a customer were to search a manufacturer’s website for “how to clean a microwave” they will be presented with step-by-step directions instead of links to articles that may or may not answer the exact question asked.
When these answers are easy to find, users avoid contacting customer support or spending precious time sorting through unstructured content to arrive at an answer. It also avoids the worst-case scenario of the customer actually leaving the website to ask Google their question and possibly getting directed to a competitor or a third-party site with questionable intentions.
It’s important to remember that, these days, quality of search is not measured in a silo. A customer isn’t going to compare individual brands based on their search; instead, the best search experience is now considered the standard for everyone. When Google, Amazon, Apple and other experienced leaders make it easy to get the right answer quickly, we ask ourselves, “Why can’t every brand make it easy too?”
When answers to questions are made available, it also enables knowledge to become more discoverable. But what is discoverability?
Whereas findability provides usable answers in context, discoverability means that users can more easily encounter information that isn’t immediately sought out. Again, building off of knowledge graphs can provide context for recommended content that understands a user’s intent and offers further relevant information to enrich their experience.
Both findability and discoverability are important for customer experience, and knowledge graphs serve as a foundation for delivering that enhanced experience.
Building a better search experience for everyone
While Google has for years been the gold standard of applying knowledge-graph structures to search, the technology itself isn’t walled off just for Google; it’s accessible to any brand wishing to utilize it. Instituting a knowledge graph-based search system is an endeavor a brand can take on, customized to whatever products, services and information resources the company uses. Building this better search system aggregates enterprise knowledge by connecting disparate systems of information into one usable engine that works for both customers and support teams.
With analytics, support and experience leaders can review common search queries to identify points of friction across the entire customer journey. A knowledge graph-based system complements these insights to form a powerful knowledge management tool. Businesses can analyze customer engagement and sentiment with search analytics, all while having access to scalable content infrastructure that can rapidly address and close knowledge gaps. This level of actionable insight is invaluable in improving the overall customer experience.
Brands invest heavily in content. Knowledge graphs turn this into the most actionable version of itself, improving resources so that answers are findable and deeper insights are discoverable.
Joe Jorczak is head of industry, service and support at Yext.
Welcome to the VentureBeat community!
DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.
If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.
You might even consider contributing an article of your own!