Diploria AI, a new software platform from Sydney digital agency Reef Digital, has launched to answer a growing blind spot: what do AI assistants actually say when a potential customer asks for a recommendation?
What the platform actually measures
Diploria sends real prompts to up to 12 AI engines, including ChatGPT, Claude, Gemini, Perplexity, Grok, Microsoft Copilot, Google AI Overviews and Google AI Mode. It uses live APIs and stores the full response from each platform, rather than estimating visibility. Since launch it has captured more than 120,000 individual responses, including over 24,000 from Google AI Overviews and AI Mode.
The platform calculates an AI visibility score from 0 to 100, tracks share of voice against chosen competitors, and flags prompts where rivals appear and your brand does not. It also separates two answer types: “Memory” responses based on what a model already knows, and “Live” responses where the engine searches the web before answering.
Why marketers should care now
The underlying numbers are shifting fast. Roy Morgan found that 13.6 million Australians aged 14 and over, or 58%, used AI tools during an average four-week period in the March 2026 quarter; ChatGPT alone reached 10.5 million people. Telsyte put the change even more starkly: 12% of Australians now call an AI tool their primary way to look something up, more than double the 5% recorded a year earlier.
The commercial impact is following. Similarweb says AI platforms generated an average of 770 million referral visits to websites each month in the year to May 2026, more than twice the previous year. Adobe found that visitors arriving from AI assistants converted 42% better than other traffic in March 2026.
From dashboard to Monday-morning action
Diploria’s most useful layer is not the monitoring, but the prioritisation. It maps website pages and topics by readiness and opportunity, analyses individual URLs for AI readiness, and produces a recommended improvement list. A Brand Intelligence report rolls up issues, competitive insights and next actions. Teams can also connect Google Analytics 4 to see which sessions actually originated from AI assistants, and detect crawlers such as GPTBot and ClaudeBot in real time.
Founder Hadrien Brassens puts the difference bluntly: “Monitoring alone is quickly becoming table stakes. The more difficult question is what you do with that information on Monday morning.”
For teams starting from zero, the practical sequence is:
- Run the free 60-second checker across five AI engines for a baseline.
- Audit the five to ten purchase-intent prompts your buyers actually ask.
- Separate Memory from Live results; stale brand facts need different fixes than weak citations.
- Connect GA4 and confirm whether AI-referred sessions are landing and converting.
Built for a different buyer
Pricing is intentionally self-serve. The free plan includes monthly checks on ChatGPT, Claude and Gemini. Paid plans start at A$39 per month for all 12 engines, rise to A$169 for daily polling, and A$419 for an agency plan covering three brands. Annual billing cuts 20 per cent.
The launch also lands two months after New Zealand brand tracker Tracksuit acquired Sydney AI visibility startup Hall. Consolidation is clearly underway, but Diploria is positioning itself as a locally built alternative for APAC marketers who do not need an enterprise contract to know what customers are being told.
Source: Campaign Brief
