Picked

Picked is an AI visibility and generative engine optimization (GEO) tool for marketing agencies and B2B SaaS teams.

When prospective buyers ask ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews, or Google AI Mode for software recommendations, Picked tracks whether your brand gets mentioned, in what position, and which competitor pages are cited as sources.

What it does:

1. Daily prompt tracking: It tests the prompts your buyers ask every single day across major AI models, giving you reliable trends instead of one-off screenshots.
2. Citation and source intelligence: It uncovers which specific URLs and directories AI models rely on when answering questions in your niche.
3. Content gap analysis: It identifies the exact prompts where competitors beat you and highlights the missing information the AI wanted.
4. AI crawler analytics: By connecting Cloudflare, Vercel, or your server via open-source middleware, it separates live search fetches (like ChatGPT-User when someone is searching right now) from background LLM training crawls.
5. Content generation to win lost answers: For prompts where you lose visibility, Picked researches the cited competition, writes a fact-checked brief and article draft designed to become the cited source, and can open it directly as a pull request on your repository.

Who should use it:

• SEO and marketing agencies: Agencies that manage client brands and need verifiable numbers to show in weekly reports. Features like dedicated workspaces per client and white-label PDF reports make client reporting straightforward.
• B2B SaaS marketing teams: Teams that want to know how often AI models recommend competitors, understand what the models read, and publish content that wins the recommendation.

Picked gives you the data to see where AI ignores your brand, and the workflow to do something about it.