We use ZeroRank across multiple properties to track how our businesses appear in AI answers over time. We’re using defined prompt sets and recurring topics rather than treating a single AI answer as “truth,” and that makes the product much more useful for us.
The strongest part is the operational visibility: separate workspaces, prompt-level results, model/platform coverage, citations/sources, trend data, exports, and API/MCP options give us enough structure to use the data in our own Search + AI visibility workflow. It is especially useful for seeing where branded visibility is already strong and where non-branded discovery is weak.
What I’d most like to see improved is control and clarity around prompt creation and scheduling. We had to clean up extra vendor-generated prompts to keep one workspace limited to the exact prompt set we intended. More explicit guardrails around what gets auto-created, run frequency, and credit/capacity impact would make recurring use easier to manage.
I’d also like continued improvement in API/callback reliability and clearer explanations of how capacity is consumed across scheduled work.
Overall, ZeroRank has become a useful sensor in our stack. We do not treat its scores as universal truth, but as structured evidence to compare over time alongside Search Console and other sources. For that use case it has been genuinely valuable.