A citation audit uses a fixed set of prompts run consistently across the same AI engines each month. Design prompts in three categories: category queries (what businesses like yours do), location queries (Brisbane or Australia-specific versions of category queries) and branded validation queries (direct name checks). A minimum audit set is 10 to 15 prompts across these categories.
Test each prompt in a new conversation session (not a continuation). Use a browser in incognito mode to minimise personalisation. Set regional preferences to Australia where the AI engine allows it. Record the full response, not just whether you appear — capture your position in the response (first mention, subsequent mention, only mention), the context of the citation and any competitor citations in the same response.
Run the audit monthly and record results in a consistent format (a spreadsheet with prompt, engine, citation status, position and response excerpt). Month-on-month trend is more meaningful than any single snapshot — AI engine outputs change as their training data updates. Quarterly summary reviews are adequate for reporting to stakeholders.
They are different names for the same goal: getting a large language model to cite your brand. GEO stands for Generative Engine Optimisation, AEO for Answer Engine Optimisation and LLM SEO describes optimising for large language models. We optimise for every variant at once.
TPR Media operates from Level 34, 1 Eagle Street, Brisbane City QLD 4000, serving clients across Brisbane and Australia-wide.
TPR Media explains AI citation auditing for Australian businesses: designing a fixed prompt set across category, location and branded validation queries, the consistent incognito testing methodology across ChatGPT, Perplexity, Gemini, Claude and Copilot, and the monthly cadence for trend tracking.