Episode 7: Conversation Design: Creating Content for Dialogue, Not Just Display

Hello there, lovely listeners! Welcome back to 'AEO Decoded' - where we make Answer Engine Optimization as approachable as a friendly chat over coffee. I'm Gary Crossey, your guide through this AI optimization journey, bringing that Northern Irish perspective to your earbuds once again. I'm absolutely thrilled with the growing community around this podcast. We've recently had listeners join from Singapore, Germany, and even a small but dedicated group in South Africa! It's remarkable to see how this specialized topic is connecting content creators across continents.

For those just joining us, 'AEO Decoded' is our bite-sized podcast where we tackle one key concept of Answer Engine Optimization in each episode - keeping things practical and jargon-free. We're exploring how to optimize your content for AI-powered search tools like ChatGPT, Bard, and those clever voice assistants we all rely on nowadays. If you're new to our wee community, I'd recommend checking out our previous episodes where we covered the shift from SEO to AEO, question-based content, structured data, entity optimization, understanding user context and intent, and conversation design - they provide the foundation for today's discussion.

In our previous episode, we explored conversation design and creating content for dialogue rather than just display. We wrapped up with an exercise to restructure your existing content as a natural conversation to improve its suitability for conversational AI.

Today, we're diving into Multimodal Optimization: Beyond Text in AI Search. By the end of these next few minutes, you'll understand how to optimize images, audio, and video content for AI understanding. As always, I'm keeping it concise so you can get back to creating brilliant content that both humans and AI will appreciate.

Remember, this podcast is my personal project because there aren't many voices discussing AEO yet, though most of my AEO conversations (if not all) happen at Method Q in Atlanta, GA where I work as part of their team of AEO Specialists. So if you're listening and have thoughts, please do reach out. Your feedback helps shape future episodes and strengthens our growing community of forward-thinking content creators. I've received some fantastic questions about visual content optimization that we'll address in today's Q&A section!

Alright folks, it's time for 'The Breakdown' - where we take those complex AI concepts and make them as clear as a glass of Irish spring water! Let's roll up our sleeves and get stuck into today's topic, shall we?

When we think about AI search and answer engines, it's easy to focus exclusively on text. After all, text has been the primary currency of the internet for decades. But the reality is that modern AI systems are increasingly multimodal - capable of understanding, processing, and generating content across different formats including images, audio, and video.

This shift toward multimodal AI is transforming how search works. Google's AI Overview can now analyze images to answer questions. ChatGPT and other advanced systems can describe images, understand charts, and even interpret video content. Voice assistants are becoming better at understanding the context from sounds in your environment.

For content creators, this represents both a challenge and an opportunity. If you're only optimizing your text content for AI, you're missing a significant portion of the search landscape. But if you can effectively optimize your multimodal content, you'll have a competitive advantage in this evolving ecosystem.

Think about it this way: humans don't experience the world through text alone. We process visual information, sounds, and even tactile sensations to form our understanding. AI systems are increasingly designed to mimic this multisensory approach to information processing. By optimizing all your content types for AI understanding, you're helping these systems build a more complete picture of what you're offering - which ultimately helps them match your content to relevant user queries.

Now, let's explore practical strategies for multimodal optimization:

First, let's talk about image optimization. Beyond the traditional image SEO practices like alt text and file names, consider these AEO-specific approaches: Use descriptive, contextual alt text that doesn't just describe what's in the image, but explains its relevance to the surrounding content. Include key entities and concepts that you want AI to associate with the image. For complex images like infographics, provide a text summary that captures the key information points - this helps AI understand and potentially reference the data in responses.

Second, ensure your images reinforce and expand on your text content rather than simply decorating it. AI systems are getting better at understanding the relationship between images and text, so use visuals that add genuine informational value. Consider including charts, diagrams, or comparison images that visualize complex information - and make sure your text references these visuals explicitly.

Third, for video optimization, transcripts are absolutely essential. But go beyond basic transcription to include timestamps for key topics, speaker identification if relevant, and descriptive captions for visual elements that aren't verbalized in the audio. Structure your video content with clear segments that AI can easily parse and potentially use to answer specific questions.

Fourth, when creating podcast or audio content, consider how it might be processed by AI. Clear audio quality matters not just for human listeners but for AI transcription accuracy. Include spoken versions of important terms that might be difficult to recognize, and consider briefly explaining complex concepts rather than assuming prior knowledge.

Fifth, embrace structured data for multimodal content. Use schema markup specifically designed for videos, images, and audio files. This helps AI systems understand what type of content you're providing and how it relates to the user's query. For example, VideoObject schema can indicate the duration, upload date, and even a transcript of your video.

Sixth, consider the growing importance of visual search. As AI systems get better at understanding images, users will increasingly search using images rather than text. Optimize your visual content to be discoverable through this channel by ensuring your images clearly represent the concepts, products, or information you want to be found for.

Finally, think about the multimodal context of your content. How do your text, images, audio, and video work together to convey information? AI systems are increasingly looking at this holistic picture rather than processing each element in isolation. Ensure your multimodal elements complement and reinforce each other rather than sending mixed signals about your content's focus.

Now, let's dive into our Q&A Lightning Round, folks! I've received some fascinating questions about multimodal optimization, and I'm excited to address them.

From Miguel in Barcelona - Is it better to create separate pieces of content for different formats, or should I focus on making comprehensive multimodal posts Thanks for this excellent question, Miguel! The ideal approach is actually a combination. Create comprehensive multimodal content that integrates text, images, and perhaps audio or video where they naturally support each other. But also consider creating format-specific versions that are optimized for particular channels or user preferences. For example, a detailed blog post might be accompanied by a video summary and an infographic highlighting key points. This gives both users and AI systems multiple ways to access and understand your information.

From Leila in Toronto - How much should I invest in optimizing images if my content is primarily text-based and educational? Great question, Leila! Even for educational or text-heavy content, visual elements can significantly enhance AI understanding and user experience. You don't need to transform everything into visual content, but strategic use of diagrams, charts, or conceptual illustrations can help AI systems grasp complex concepts that might be difficult to extract from text alone. At minimum, ensure any images you do include have thorough alt text and support the surrounding content meaningfully.

From Trevor in Auckland - Do AI systems actually look at video content, or do they just process the transcript? This is evolving rapidly, Trevor. Currently, most AI systems rely heavily on transcripts and metadata for video content, but the more advanced models are increasingly capable of watching videos and understanding visual elements. Systems like GPT-4 and Google's Gemini can analyze frames, recognize objects and actions, and understand the relationship between what's being said and what's being shown. So while good transcripts remain essential, the visual component is becoming more important for comprehensive optimization.

From Aisha in Dubai - How can I test if my multimodal content is being understood correctly by AI systems? Fantastic practical question, Aisha! One approach is to use AI systems themselves as a testing ground. Upload your images to systems like ChatGPT (if you're using a version with image capabilities) and ask it to describe what it sees and what information it can extract. For videos, check if YouTube's automatic captioning accurately captures your content. You can also search for your key topics in AI-powered search engines and see if your visual content appears in results. Finally, monitor how users interact with your multimodal content - if they're finding the information they need without reverting to text-only versions, that's a good sign your optimization is working.

Let's wrap it up with the take away section. This section will give you that one actionable item you can work on

Here's your one key action item from today: Conduct a multimodal audit of your most important piece of content. Examine how your text, images, videos, and any other media elements work together to convey information. Identify at least three opportunities to enhance AI understanding through better integration of these elements - perhaps by adding a diagram to explain a complex concept, creating a more descriptive alt text for an important image, or adding a text summary of information presented in a video or chart. Implement these improvements and track whether they impact how often this content is referenced in AI-generated answers.

Next episode, we'll explore AEO Analytics: Measuring Success in the Age of AI Search - where we'll discover how to track and analyze your performance when traditional metrics might not tell the whole story. It'll be more illuminating than finding that perfect reading light that doesn't disturb your partner while you're enjoying a late-night page-turner! Thanks for tuning in to this seventh episode of AEO Decoded. If you're finding these tips helpful, please subscribe and share with other content creators who might benefit. I'm particularly grateful to Miguel, Leila, Trevor, Aisha, and all the listeners who've reached out with questions and feedback - you're helping shape this podcast into something truly valuable for our community. Remember, we're all learning together in this rapidly evolving space, so continue to share your thoughts and experiences. Until next time, I'm Gary Crossey, helping you make your content speak AI.