Search Challenges? AI Is Here to Help
Organizations have increasingly been turning to artificial intelligence (AI) technologies to help them make better decisions, improve customer experience and stay competitive. AI-powered solutions offer new opportunities for businesses of all sizes — and that includes enterprise search applications.
Since its launch in November 2022, we have been experiencing the biggest buzz around AI ever, thanks to ChatGPT. Every time I started writing this article, yet another story came out that forced me to review, delete and update my post. Today was no exception either. However, I decided to use today, Feb. 13 2023, as the cutoff and focus on the questions and considerations based on the available knowledge and insights — despite knowing that new questions will arise between writing and publication.
A Tipping Point for AI-Powered Search
Finding information has long been one of the biggest challenges in the enterprise. Improving findability on the organizational level requires knowledge and understanding, access to both structured and unstructured data, as well as an overview of the overall content architecture.
As a workshop attendee said years ago, "search is complex and beautiful." But search never has been as "trendy" as it has in the last few weeks. First Microsoft, then Google announced their takes on the next version of AI-powered search, raising the question if Bing, with the help of OpenAI, could overtake Google for the search crown.
While I can't predict the future, every day makes it clearer that we are approaching a tipping point when AI-powered search becomes a reality.
But what is AI-powered search, and what should we be watching for?
Related Article: AI and Enterprise Search: Who's in Control?
How Can AI Improve Search?
First we should understand where AI has the greatest potential to augment the search experience.
Relevance of search results: AI technologies can help organizations improve the relevance of their search results via a better understanding of user intent.
Personalization: AI technologies can help personalize search results, taking into consideration information on the user's profile, preferences, usage patterns, search history, etc.
Spell correction and auto-completion: AI-powered technologies can help search engines correct mispelled queries and suggest relevant autocompletion options.
Question answering: We already see this with ChatGPT and similar tools like Jasper AI. AI technologies can provide a more natural way of interacting with search by providing question-answering capabilities. As a result, users can express their queries in a more natural way while getting better (in theory) and faster results.
Multi-lingual support: Multi-lingual search is a big challenge for multinational, multi-lingual organizations. AI technologies make it possible for organizations to offer multi-lingual search capabilities, by understanding and translating the natural language of the query and then translating the results.
Trend analysis: Enterprise search managers have long analyzed search usage to identify usage patterns, and then used those insights to improve the future search. However, extracting the right reports, understanding the data therein, analyzing it and making the right decisions based on the insights is tricky. AI technologies can help detect patterns in the search data and even suggest appropriate actions as necessary.
Filtering out incorrect, outdated and irrelevant information: While we've all seen ChatGPT provide incorrect or even false responses, the same happens in most enterprise search applications all day, every day. You're based in Singapore and searching for the travel policies of your organization? The first results might be from the U.S. or UK, but the one you need is nowhere in sight. A search for the cafeteria menu offers menus from last year as the first page results. A search for your manager's name produces the latest Christmas party pictures as the top result .... Sound familiar?
AI can help search engines identify and remove these incorrect, outdated or irrelevant information from results, improving the accuracy and reliability of the information provided.
Voice search: Voice search has been on the rise. AI can improve the accuracy of voice search and make it easier for users to find what they're looking for through spoken queries.
All of these could have a significant impact on the search experience. When combined with an updated, enhanced search user interface — the future of search is here.
Related Article: Are Digital Assistants the New Face of Search?
Good Findability Starts With Good Content
Findability is much more than search: good findability starts with the content. The search engine is only exposing the content, so only improving the search engine technologies and features isn't enough. As the saying goes, "Garbage in, garbage out." If the content stored in our repositories is out-of-date, inaccurate and unreliable, even the best AI engine will fail to find a relevant response to a query.
Learning Opportunities
Think of the search engine and AI technologies as a magnifying tool: they can make it faster and easier to find what you need — but only if you have the right content in the first place. They can augment the traditional search experience, but cannot replace the human effort required in advance.
So where and how can AI help improve the quality of your content? Let's see a few examples:
Summarizing content: One of the most common problems in knowledge management is silent content. Given the vast stores of information existing in most businesses, users can't find or don't even know the information that's available in the various repositories. AI-powered summarization algorithms can automatically create short summaries of documents to display on search or discovery pages.
Generating meaningful document titles: How many times have you seen "Untitled" as the document title on a search results page? Or "Andrew's document 123"? And think of the irrelevant titles created every day when someone creates a document from a template or by copying another document.
AI can automatically generate meaningful titles for documents based on the document's content.
Generating tags and metadata: Adding tags to documents manually is a much needed, mostly underrated and completely underutilized task. AI can be used to create meaningful tags and metadata that accurately reflect the contents of each document.
Content lifecycle support: People create, edit and then publish content — and then in most cases, nothing happens. But unpublishing, updating and archiving are just as important as the publishing and approval processes. However, implementing these processes can be challenging — simple rules-based lifecycle stages might be not able to follow the organization's needs. This is where using AI models might be useful: for example, for automatic archiving of documents that have not been accessed in a long time, or asking the content owners to update/refresh their content as needed.
Above are just a few examples of how AI can help improve the quality of your content and findability in general. However, it's important to remember — AI is here to support humans, not replace them. It's not about replacing search either, rather augmenting the search experience — which will be an exciting journey for all.
Related Article: Has Microsoft 365 Been Clinically Tested?
Start With 'Why'
Business (and life) coaches always start by helping you find your "why" — this is where to start your journey with AI, too. It might be compelling to jump in and deploy all kinds of AI tools in your organization, but without knowing why, it will only add to your silos, increase the mess and confusion, and end up causing more headaches than improvements.
So why are you trying to implement AI and search technologies in the organization? What do you want to achieve? Is it about improving user experience, and adding value for them? Is it about operational efficiency, cost reduction or better decision-making?
Or is it just because it's what everyone else is doing?
Once you know your why, create a plan of how to get there. And be ready for change — this journey won't be linear and you'll need to adjust your plans on the go.
Good luck!
Learn how you can join our contributor community.
About the Author
Agnes Molnar is the CEO and managing consultant of Search Explained. Agnes is an internationally-recognized expert in the fields of modern search applications, information architecture, and Microsoft technologies.
Featured Research
Research Report
Read Now
eBook
Read Now
Guide
Read Now
Guide
Read Now
Research Report
Read Now
Buyer's Guide
Read Now
Infographic
Read Now
White Paper
Read Now
Related Stories
Smarter Experiences Win — Learn from Today's Leaders
