Why Government Websites Need a Knowledge Graph

According to a recent McKinsey study, customer satisfaction with federal and state government websites had the lowest rating of all industries studied. Averaging 5.9 out of 10, the public sector has a failing grade when it comes to, well, interacting with the public.

But why?

With the passage of legislation like the 21st century IDEA Mandate and the new President's Management Agenda, it appears as if the government is trying to take customer experience seriously. Following the pandemic, the public's need to get answers from governing bodies like CDC, NIH, SBA, and individual States became paramount. But government websites haven't been quite ready for the challenge — and a tremendous amount of misinformation circulated in part because the public could not search for and find a direct response from the authoritative source. (Make no mistake, the correct information exists within Government websites, but finding that answer is like finding one needle across ten or more haystacks.)

In the public sector, information typically resides within any agency in multiple silos. This is by design, to keep data separate for specific uses within a department. But it results in it being difficult to surface information from across disparate parts of the organization, and thus nearly impossible for government websites to answer citizens' questions when they search for something.

The solution is smarter search — built on the foundation of a knowledge graph.

How a knowledge graph helps you return better answers

According to a recent Yext Survey, 84% of people are likely to trust answers they got directly from a government's website and 79% of that population are likely to go back to the website if they get their questions answered directly. By employing a knowledge graph — a brain-like database that is structured to connect entities and relationships — government and state agencies can improve their ability to deliver the right answer to any given question at any time.

Let's illustrate the value of a knowledge graph with a use case.

Imagine a veteran goes to a VISN website to search for a doctor, and that they need a doctor who practices a specific type of medicine (treats diabetes, for example), and who speaks Spanish.

All of that relevant information resides within the VISN, but it's not tied together. So, instead of returning a search result that shows a specialist who speaks Spanish, the VISN's search experience would likely return multiple blue links to pages that merely mention the keywords "diabetes" or "Spanish." That's a bad experience for that user.

But, instead, when an organization uses a knowledge graph to link these types of attributes ("entities") to one another – with flexible, bidirectional relationships – they can understand each users' question and answer it. Just like Google does.

This is the beauty of an AI-powered search solution based on a knowledge graph. It can do the heavy lifting, making service to each searcher far more effective and efficient. The same holds true for any other public sector example: when platforms built on knowledge graphs can relate and connect entities, the answers rise to the surface. This leads to higher public trust and more satisfied constituents.

Also worth noting: costs begin to drop as more and more of the public are able to self-serve with better search. From the same Yext survey, 48% of people who cannot find what they need on a government website will opt for a more cost intensive way of getting the needed information. But when they can answer their own questions via search? They will.

Let's put the power in the hands of the people: to have the same search experience in the public sector that we've all seen as consumers using Google or Amazon.

Learn more about better site search in the public sector here.