Why your intranet search is broken: A survey of research
Most intranet search problems are structure problems. Before you sign a contract on a new application to improve your intranet’s search, it's important to understand why search engines fail when built on a foundation of disorganized content.
While I have years of experience witnessing employees have bad search experiences, it helps to have research to back up experiences and business cases. This post reviews that research to help you understand the issue and make informed decisions to improve your intranet.
Three things to keep in mind as you read:
The cost of bad findability is real and measured in hours, not annoyance. Studies have found that knowledge workers lose a significant share of time hunting for information.
Search is only as good as the structure beneath it. An internal search engine retrieves and ranks what's already there. If it’s not supported with good structure, it can't impose structure on the content.
"We need a new search tool" is usually the wrong diagnosis. The more accurate framing is “Our search sucks because our structure sucks.” You’ll need better information architecture (IA) and taxonomy to support any new search tool.
The pain is real and costly
The frustration of a broken intranet search isn't a minor irritation; it's a measurable drain on productivity. Without interviewing and surveying your employees, how do you quantify the productivity drain? Here are some reports that point you toward some statistics:
Employees lose a large share of the workweek to searching. McKinsey Global Institute's 2012 report on social technologies is one of the most widely cited sources in this space. Knowledge workers spend about one full workday per week searching for and gathering information. The same report found that implementing better tools for finding, sharing, and using information could raise the productivity of high-skill knowledge workers by 20 to 25 percent. (1,2)
Although this report is from 2012, this still matches my experiences. Employees still can’t find information on internal systems, so they can’t leverage it in new ideas, products, and services.
The frustration is universal and employees blame the system. A 2023 Adobe Acrobat survey (3) of 1,118 employees found:
48 percent had difficulty finding documents quickly
47 percent say their company's digital organization system is ineffective and hard to navigate
71 percent say poor digital organization interferes with their ability to work effectively
95 percent have felt frustrated searching for a document
Notably, employees pointed to structural causes: difficulty finding documents, inconsistent naming conventions, and colleagues not following organizational protocol. Those aren't search-engine complaints; they're information architecture and content governance complaints.
In my work on intranets, employees have a hard time making connections between how they organize their documents and this makes documents findable. They’re not trained on the basics of metadata and organization, they don’t follow standards, or no standards are established. Individual employees can’t create standards on their own; these need to be created and disseminated at an organizational level.
Why good search tools still fail
Why do searches keep coming up with so many results that just aren’t right? Because a search engine is a retrieval system and surfaces what exists. It doesn’t invent the structure, labels, and relationships that make content findable in the first place.
Often, we think that search will just find the document that’s relevant based on a full text search, but this is actually a much more complicated subject than we realize! We need to support the search engine, but we don’t.
Search can't rank what isn’t labeled. An Early Information Science report shows us just how deep labeling goes. Pre-AI enabled search, the metadata always mattered. With AI-enabled search, enterprise models need to be supported with labeling and metadata. The MIT Project NANDA shows when models aren’t supported, AI search tools fail in the workplace.
Search has, does, and always will rely heavily on metadata, consistent naming, and a taxonomy that defines how concepts relate to one another. When documents are untagged or tagged inconsistently, and when there's no taxonomy that connects synonyms, search has nothing to match against beyond raw keywords. The result is the classic search experience: the right document probably exists, but the terms the employee puts in doesn’t match the term that the writer chose, and the content can’t be matched.
Structure and search often don’t reinforce each other. In a 2022 article (10), Nielsen Norman Group found a number of issues that cause search problems, one of them being coherent metadata, page titles, and descriptions. Employees expect an intranet search to behave like Google. Because it doesn’t act like Google, employees lose confidence in the results displayed. Additionally, unclear or redundant titles, duplicate files or pages, and out of date information further degrades trust in search results.
What fixes intranet search
When an application or software isn't the problem, but content structure is, the approach to solving the problem changes.
You’ll need to work on Information architecture and taxonomy. IA defines how your content is organized and how concepts relate; taxonomy provides the controlled vocabulary and metadata that let systems recognize that different terms mean the same thing. In 2019, Nielsen Norman Group (5) already found that high findability is the result of well-defined information architecture paired with well-designed navigation. The right structure supports search, navigation, browse, and (increasingly) AI-driven retrieval.
Define the structure, then tag the content. Once you have the structure, it becomes the scaffolding for more content work. Content creators need to add in those relevant titles, descriptions, tag the content with taxonomy terms, and add other relevant metadata. You’ll also need to create content governance guidelines to continue to support search and findability.
Conclusion
Search is often a central complaint of intranets. Employees are dragged down in finding information at the same time they are being pressured to be more productive and efficient. To improve productivity, we have to support knowledge workers in finding the information they are looking for. Choosing a new search tool won’t fix the problem. We’ve looked at reports and studies from the last 15 years that show that poor search is still a main employee complaint. It’s time to take a different approach to solving the problem.
At Key Pointe, we help organizations diagnose the root cause of poor findability. We then build the information architecture and taxonomy to make search, navigation, and AI work. Reach out to Key Pointe.
References
The social economy: Unlocking value and productivity through social technologies (July 2012). Full report PDF https://www.mckinsey.com/~/media/mckinsey/industries/technology%20media%20and%20telecommunications/high%20tech/our%20insights/the%20social%20economy/mgi_the_social_economy_full_report.pdf
Overview page https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/the-social-economy
How digital organization impacts employees and the workplace (2023 survey of employed Americans) https://blog.adobe.com/en/publish/2023/09/29/how-digital-organization-impacts-employees-workplace
Top 10 Information Architecture (IA) Mistakes (Jakob Nielsen, 2009) https://www.nngroup.com/articles/top-10-ia-mistakes/
Low Findability and Discoverability: Four Testing Methods to Identify the Causes (Aurora Harley, 2019) https://www.nngroup.com/articles/navigation-ia-tests/
Information Architecture: Study Guide (2023) https://www.nngroup.com/articles/ia-study-guide/
Searching for gold: Harnessing the power of taxonomy and metadata to improve search (2025) https://www.earley.com/insights/enterprise-search-ai-era-retrieval-grounding
The Role of Information Foundations in Scaling a Business (2025) https://factorfirm.com/thinkhub/the-role-of-information-foundations-in-scaling-a-business/
The GenAI Divide: State of AI in Business 2025 https://mlq.ai/media/quarterly_decks/v0.1_State_of_AI_in_Business_2025_Report.pdf
Intranet-Search Essentials https://www.nngroup.com/articles/intranet-search/