8 Tips on Preparing Your Content for AI
Since the arrival of AI–large language models (LLMs), chatbots, and voice-activated search and assistants–content publishers have been anxious about how to prepare their content for consumption by these tools.
The truth is, preparing content for AI, chatbots and voice assistants is no different from preparing it to be discovered on your website or through traditional search. Preparing content well should render it future proof and device- and channel-neutral.
What are the crucial steps in preparing content for a future that will include AI?
Know your audiences: Know what they care about and the language they use to describe their needs. Write succinctly and clearly on subjects that interest them. Use the terminology your audiences use. Use simple sentence structure. Keep the reading level to roughly eighth grade (middle school) for general content.
Test how it sounds: Pay attention to how the content sounds when read aloud (as it may be by AI assistants and AI-powered search). There’s no better way to learn how stilted your content is than to hear it read by Siri, Alexa, or Google search!
Tend your content: Have a content strategy and governance plan for producing and maintaining high-quality, relevant content. Continuously improve it by evaluating it against quality and performance metrics.
Follow standards: Make sure your content follows all the relevant digital standards. Structure it consistently, according to code standards and standard semantic markup language, so any system can interpret it–for instance, knowing what is a title, summary or abstract, a particular level of heading, body copy, task list, etc. Using the proper markup makes content readable by nonhuman consumers such as search engines, screen readers, and AI tools—all of which deliver your content to human users.
Establish a usable taxonomy: Organize your content the way your audiences think about it (follow their mental models). Use labels they recognize, and tag your content with audience-appropriate terms to establish relationships and aid discovery. Use audience-relevant keywords in body copy, to support search engine optimization (SEO). To boost your content’s shareability and help you wield more influence within your profession’s information ecosystem, use standard industry vocabularies when available. (Hint: Knowing how your audiences think and the terms they use means you need to do user research.)
Remember the metadata: Identify your textual, audio, and visual content items with helpful metadata (tooltips for hyperlinks, descriptive alt tags, a metadescription for each page), to support SEO, accessibility, usability, and AI consumption. For every visual-only or audio-only element, provide alternate text.
Break it down: Make sure instructions and tutorials are written as numbered steps. Use headings and unordered lists to break up large passages of text.
Test with humans and AI tools: Make sure your content, sites, mobile apps, and other digital products are created with your target audiences and known nonhuman consumers in mind—and then test periodically with these users to see if you need to change anything. New AI tools and LLMs are being developed all the time, and human users’ needs change, so your feedback loop with content consumers (both human and nonhuman) should never end.
So, take a deep breath. Despite the hype, AI is just another content consumer. When you prepare and maintain your content appropriately, using the steps above, you are preparing your content for consumption by an array of human and nonhuman users, including AI tools we haven’t even seen yet.
If you’ve already taken many, or all, of the steps above to prepare your content for distribution in desktop, mobile, and search, you are ahead of the game. Those steps will also help your content work well in AI applications.
If you’d like some help thinking through your organization’s content strategy, or how to use AI to advantage, get in touch!
