Overview
An AI Q&A Set transforms your selected keyword into a set of intelligent questions and answers generated by different AI models.
This helps you uncover what users are asking, how AI engines like ChatGPT and Perplexity answer, and which competitors get mentioned most often.
Each Q&A Set gives you a market snapshot, showing your brand visibility, competitor mentions, and AI perspectives on your chosen topic.
How to Create an AI Q&A Set
You can create an AI Q&A Set from the Keywords page.
Step 1. Choose a Keyword
From your keyword list, click Generate AI Q&A Set next to the term you want to analyze.
This keyword becomes the main focus of the analysis and can’t be changed later.
If you need a different keyword, simply go back and start again.
Step 2. Define Your Target Market
Select your language and country.
These fields are pre-filled from your Brand settings, but you can adjust them if needed to test new markets or languages.
Step 3. Add Competitors
Enter at least one competitor brand or domain (for example, canva.com or adobe.com).
This allows Savannabay to compare how often your brand appears in AI answers relative to others.
💡 Tip: Add direct and indirect competitors for a richer comparison — even smaller niche players can reveal positioning gaps.
Step 4. Set Question and Answer Parameters
-
Number of Questions: Choose how many AI-generated questions you want per cluster (e.g., 2, 3, or 5).
-
Answers per LLM: Decide how many responses each AI model (GPT, Claude, Perplexity, etc.) should generate.
-
Credit Consumption: The panel below shows exactly how many credits your analysis will consume.
Step 5. Generate the Q&A Set
Click Generate AI Q&A Set.
Savannabay will run the analysis, sending your keyword and parameters to multiple AI models to gather real Q&A data.
Each response is stored and analyzed for brand and competitor mentions.
Viewing and Interpreting Results
Once generated, your Q&A Sets appear in the AI Q&A Sets tab.
Each set expands to show:
-
Generated Questions — real user-style questions derived from your keyword.
-
Competitor Mentions — visual bars showing how often each brand was cited by the AI.
-
AI Answers — full text responses from different AI models, displayed side by side for comparison.
-
Recommendations — actionable suggestions (available in the next step) to improve your visibility or positioning.
Running a New Analysis
Click Run New Analysis to refresh your existing AI Q&A Set with the latest data.
This uses the same keyword, language, competitors, and parameters as before.
Running a new analysis allows you to track how your brand visibility evolves over time and whether your AI Optimization (AIO) efforts are improving performance.
Your updated results will also be reflected on your Dashboard, giving you a visual trend of progress.
💡 Tip: You can run new analyses as often as you wish.
However, for most brands, running it once a month or once every 3 months provides the best balance — since it takes time for your optimizations to be reflected across AI models.
Understanding the Insights
-
Brand Share: Indicates your presence in AI responses compared to competitors.
-
Model Differences: Each AI (e.g., GPT-5, Perplexity) may cite different brands — helping you understand how algorithms perceive your niche.
-
Question Diversity: The more diverse the questions, the broader your visibility opportunities.
💡 Example:
For the keyword “how to create an infographic”, you might see:
Piktochart: 33%
Adobe Express: 34%
Canva: 33%
This means AI sees all three as equally relevant — a strong signal to optimize your brand mentions in AI-generated answers.
Best Practices
-
Always include at least one competitor (required for generation).
-
Start with 2–3 questions per cluster for faster insights, then scale up.
-
Compare results across different languages or regions to uncover new market opportunities.
- Make sure to always analyze the lower level of your brand, like your sub-brand or brand branch.
-
Re-run analyses periodically to monitor improvements in AI visibility.