> ## Documentation Index > Fetch the complete documentation index at: https://promptwatch.com/docs/llms.txt > Use this file to discover all available pages before exploring further. # Why is my visibility zero? > The likely reasons your Visibility Score shows zero, from data that hasn't arrived yet to missing brand aliases, and how to check each one in the product. Your dashboard shows a Visibility Score of zero, or close to it, and you want to know whether something is broken or the number is real. Both happen. Here are the causes, ordered by how often they turn out to be the answer, with a check for each. ### 1. Your responses haven't been collected or analyzed yet A response only counts toward your score after two steps finish: the AI model answers your prompt, and Promptwatch analyzes that answer to detect your brand in it. On a new project, the first collection starts right after your monitor is created and analysis follows continuously, but a fresh dashboard can show zero simply because nothing has completed both steps yet. **How to check**: open **Prompts → Responses** in the sidebar. If the list is empty or very short, collection or analysis is still in flight, not a real zero: the first run starts as soon as the monitor is created, and a row appears only after analysis finishes. See [how fresh is my data](/docs/academy/data-freshness) for the full timing. ### 2. The zero is real: models answer your prompts without naming you The Visibility Score averages across all analyzed responses, and a response that never mentions your brand counts as zero. If AI models answer "best project management tool for agencies" with five competitors and not you, your score for that prompt is genuinely zero. This is the score doing its job: it is showing you the gap, not malfunctioning. **How to check**: open a few responses and read them. If competitors appear where you don't, confirm it with the [Share of Voice](/docs/academy/share-of-voice) tab, and also check that your name, in any variant, really is absent from the answer text: if you can't find yourself in the answers either, detection is working and the models simply aren't recommending you yet. That is the starting point of GEO work, not a setup error. See [Visibility Score](/docs/academy/visibility-score) for how the score is computed. ### 3. Your brand appears, but under a name Promptwatch doesn't know Detection matches your brand name, your website domain, and your aliases against each response. The match is strict on purpose, so a generic word can't credit you with mentions you didn't earn. The flip side: if answers call you "Acme Incorporated" and your profile only says "Acme Corp", those mentions are missed and your visibility reads zero even though you're in the answers. **How to check**: read responses where you'd expect to appear. If you spot your brand under a variant name, abbreviation, or old name, add it as an alias in the **Brand Book**. Analysis uses the new alias from that point forward; already-analyzed responses are not recalculated. See [your brand profile](/docs/academy/brand-profile) for what makes a good alias, and check the website field too, since your domain anchors detection as well. The Promptwatch Brand Book alias field, showing Homestra B.V. as an alternative brand name. ### 4. You're only zero on some models Visibility differs sharply per platform. A brand can hold a strong score on ChatGPT and a flat zero on Perplexity, and the blended dashboard number can hide which is which, especially with few models tracked. **How to check**: use the model filter on the dashboard and look at each model separately. If a model has responses that name competitors but not you, it's a coverage problem on that platform. If it has no responses at all, that's a data problem: see [no data for a model](/docs/academy/no-data-for-model) and [choosing your models](/docs/academy/choosing-your-models). ### 5. Your prompts are too generic for anyone to win Prompts like "what is CRM software" produce explanatory answers that name few or no brands, so every brand scores near zero on them. A prompt set dominated by definitional questions produces a low score that says more about the prompts than about you. **How to check**: read the responses for your lowest-scoring prompts. If no brands appear at all, the prompt isn't a recommendation question. Replace it with the way a buyer actually asks, like "best CRM for a five-person sales team". See [finding the right prompts to track](/docs/academy/right-prompts-to-track). ### When zero is expected A new brand, a new market monitor, or a category where established players dominate every answer will legitimately score at or near zero at the start. The baseline is the point: you can't measure progress from a number you never recorded. If the checks above confirm the zero is real, your next stop is the [Content Gap analysis](/docs/academy/content-gap) to see what to build first.