Search and Discoverability: Boosting Trust in the Public Sector

Public trust in government is a complex, multidimensional phenomenon shaped by decades of experience, cultural context, and institutional behaviour. But in the digital age, an increasingly important component of that trust is transactional: can I find the information I need? Can I access the service I am entitled to? Does the government's digital infrastructure work for me? When citizens encounter a government website and cannot find an answer to a straightforward question, the failure is not experienced as a technology problem. It is experienced as a government problem.

This ebook examines the relationship between search quality and public trust, explores the concept of discoverability as a public service obligation, and provides practical guidance for public sector leaders seeking to improve both the search experience and the trust outcomes that flow from it. The evidence is clear: better search builds trust, and poor search erodes it. The good news is that AI-powered search technology now makes it possible to deploy dramatically better search experiences quickly, without lengthy implementation projects or significant disruption to existing content workflows.

The Trust Deficit in Digital Government

Government digital services consistently underperform against citizen expectations set by consumer technology. Citizens who use Google, Siri, and Amazon every day come to government websites expecting search that understands natural language, tolerates imprecision, and returns direct answers. What they typically find is a keyword search bar that returns a list of official documents in response to a conversational question.

The consequences extend beyond user frustration. When a citizen cannot find information about a benefit they are entitled to, they may not claim it. When a business cannot find the permit application they need, they may delay or abandon a legitimate activity. When a researcher cannot find a public record, the government's commitment to transparency is undermined by practical inaccessibility. In each case, the failure of search has a consequence that extends well beyond UX metrics.

"Every time a citizen cannot find information on a government website, it costs twice: once in the contact centre call or office visit that follows, and once in the incremental erosion of the belief that government digital services are worth using."

— Keyspider Public Sector Practice, 2024

Discoverability as a Public Service Obligation

The concept of discoverability in the public sector context is this: if a service or entitlement exists, citizens should be able to find it. The fact that information about a service is technically published on a government website is not sufficient if that information is effectively inaccessible due to poor search, confusing navigation, or terminology that only insiders understand.

This framing has implications for how government agencies think about their digital responsibilities. Publishing information is necessary but not sufficient. Ensuring that information can be discovered and understood by the people it is intended to serve is the actual obligation. Search quality is therefore not a technical nicety but a core component of the public service mission.

The Equity Dimension

The citizens most harmed by poor government search are disproportionately those with the greatest need for government services and the fewest alternative means of accessing them. Wealthier citizens have professional networks to ask, have higher literacy and digital confidence, and have more flexibility to spend time navigating bureaucratic websites. Citizens experiencing housing insecurity, navigating the welfare system, or dealing with healthcare crises often have none of these advantages.

For these citizens, a failed search is not an inconvenience. It is a barrier to accessing services that may significantly affect their circumstances. AI search that understands natural language, tolerates imprecision, and returns direct answers is therefore an equity intervention as well as a technology improvement. Digital leaders who frame AI search solely as an efficiency measure are underselling the case to their stakeholders.

From Keyword Search to AI-Powered Discoverability

The technical shift from keyword search to AI-powered semantic search is well understood: vector embeddings replace string matching, enabling the system to find semantically relevant content regardless of terminology alignment. But the practical implications for discoverability go further than most technical descriptions convey.

AI search does not just improve the results for queries that previously returned poor matches. It fundamentally changes the nature of the search interaction. Citizens can ask questions in their own words, receive direct answers with citations, and follow up with clarifying questions. The search interaction begins to resemble a conversation with a knowledgeable, patient public servant, rather than a lookup in an unfamiliar filing system.

Building Trust Through Search Design

Transparency and Citations

AI-generated answers in government search must include clear citations to source documents. Citizens who can see that an answer is grounded in a specific, verifiable policy document are more likely to trust that answer than a response that appears to come from nowhere. Citations also serve an accountability function: if the cited document is wrong or outdated, the error is identifiable and correctable through normal content management processes.

Honest Handling of Unknown or Out-of-Scope Queries

A trustworthy AI search system should clearly indicate when it does not have sufficient information to answer a question, and direct users to appropriate alternative channels. A system that attempts to answer every question regardless of whether it has relevant information will eventually give incorrect answers, and incorrect answers from a government AI system are corrosive to trust in a way that a simple 'I could not find information on that topic, please contact us at...' never is.

Accessibility and Inclusion

Trust in government digital services is also shaped by whether those services work for all citizens, not just those with high digital literacy and no accessibility needs. WCAG 2.1 AA compliance for search is not merely a legal requirement but a statement that the service is designed for everyone. Search interfaces that work correctly with screen readers, keyboard navigation, and assistive technologies communicate a commitment to inclusion that builds trust across all user groups.

Measuring Trust Outcomes from Search Improvements

Measuring the trust impact of search improvements requires looking beyond search metrics to outcomes. Contact centre volume for information enquiries is the most direct indicator: if citizens can find answers themselves, they do not call. Citizen satisfaction surveys, particularly questions about how easy it is to find information on the agency's website, provide a direct trust measure. Task completion rates in user research studies, measuring the percentage of citizens who can successfully complete defined information-finding tasks, provide a behavioural measure of discoverability.