# PsychicWho > Independent psychic advisor directory: every profile states the evidence it rests on (an in-depth analysis, a verified AI-assisted research dossier, or a base profile built from the platform listing), lists follow one published ordering rule, and no placement is for sale. Last updated: 2026-11-19. Directory size: 3231 advisors across 6 platforms (Kasamba, Keen, Purple Garden, MysticSense, Spiritual Blossom and AskNow). PsychicWho is written by the PsychicWho Editorial Team. Each profile is labelled with its evidence tier: an in-depth analysis by Sarah of Eastern Alignment (https://easternalignment.com), quoted with her permission; an AI-assisted research dossier compiled by the editorial team from public sources (platform profiles, client reviews, public discussion); or, where too few public reviews exist to analyze an advisor, a base profile built only from the platform listing. AI is never used to write reviews or testimonials, invent sessions or quotes, or set rankings. Rankings, scores, tags and quiz matches are computed by code from published rules. It does not accept pay-for-placement. ## Key pages - About (/about): the editorial team, why and how AI is used, funding, corrections policy. - Methodology (/methodology): data sources, listing order, the composite formula with worked examples, AI research protocol, topic tagging, the full match-engine model and audit, independence rules, limitations, references and a public changelog. - Advisors (/advisors): the full directory in the published listing order. - Advisor profiles (/advisor/{slug}): composite score broken into parts, platform rating and completed readings, listed price, new-client offer, pros and cons, and a link to the full editorial review where one exists. - Advisor research dossiers (advisor profiles marked "Research dossier"): a dated summary, a key-facts table, the dislike rate over all time, the previous 12 months and the last 12 months (on star-rating platforms: rating patterns with the reviewed window stated), repeat-client share, recurring praise and complaints with their frequency and a verbatim client quote, prices per 10 and 20 minutes, an FAQ, and sources. Every number is computed from the platform's public data on the stated date. - Best by topic (/best-for/{topic}): love & relationships, ex back & reconciliation, twin flame, career & money, mediumship & grief, pet loss, spiritual guidance, life purpose. - Best by platform (/best-on/{platform}): kasamba, keen, purple-garden, mysticsense, spiritual-blossom, asknow. - Each of those 13 hub pages carries guidance written for that category only: what a reading can and cannot answer, what to check before booking, what it typically costs, and the failure modes specific to it. They are not one template with the nouns swapped. - Best-of lists (/best): editorial lists, each a set of category winners chosen by printed rules. Lists: /best-online-psychics (best online psychics), /best-love-psychics (best love psychics), /best-psychic-sites (best psychic sites), /best-tarot-readers (best tarot readers), /cheap-psychic-readings (cheap psychic readings), /best-purple-garden-psychics (best purple garden psychics), /best-kasamba-psychics (best kasamba psychics), /best-psychic-mediums (best psychic mediums); on topics that have a list, the /best-for/{topic} hub is the full ranking and links to it. - Best online psychics (/best-online-psychics): ten category winners (best overall, love, video, most experienced, value, medium, ex back, career, top MysticSense pick, lowest price), each chosen by a rule printed on its card and applied by code to the research evidence, with the 12-month dislike rate, repeat-client share, price and a dated client quote per advisor. The first four categories consider only Kasamba and Purple Garden, where every rated session carries a like or dislike. - Platform reviews (/reviews, /reviews/purple-garden, /reviews/kasamba, /reviews/mysticsense): built from the research evidence behind every dossier on that platform — listed prices per channel with the same-advisor channel premium, the pooled 12-month dislike rate (or, on star-rated MysticSense, the share of recent reviews rated three stars or lower), complaint and praise theme shares, advisor tenure and repeat-client share — plus refund and idle-balance terms read on the platform's own terms page, dated. - Comparisons (/compare/purple-garden-vs-kasamba, /compare/purple-garden-vs-mysticsense): the same figures side by side; the better figure per row is chosen by code. Purple Garden (like/dislike per session) and MysticSense (1–5 stars) rate differently, so their satisfaction rows carry no winner. - Trust checks (/reviews/purple-garden/is-it-legit, /reviews/kasamba/is-it-legit): operator, what the terms commit to, complaint themes and the refund procedure. - Refund guides (/reviews/kasamba/refund, /reviews/purple-garden/refund): each platform's refund window, required details, refund form (credit or money), chargeback and idle-balance terms as read on its own terms page, plus which complaint themes are about the session itself (checkable by support) versus the outcome, and a cross-platform refund table. - Offers (/free-psychic-minutes): every platform's new-customer offer as read on its own site on a stated date, converted to dollars at the platform's median listed rate; offers not visible on a public page are labelled. - Match quiz (/match): six questions, no account, no payment. The answers are scored in the browser and never transmitted; the page also states the full question set for readers who arrive without JavaScript. ## Ranking methodology (summary) - Listing order: platform popularity first (from the platform popularity snapshot: completed readings where published, review count for MysticSense which does not publish session totals, meetings held for Spiritual Blossom). Advisors without popularity data (including all Keen advisors) follow, ordered by composite score. Ties are broken by composite score. - Composite score (0–10) = 0.6 × rating + 0.2 × volume + 0.2 × freshness. Rating is Bayesian-smoothed toward the directory mean with a 25-reading prior and mapped 1★→0, 5★→10. Volume is completed readings, log-scaled against the largest profile. Freshness is days since the newest client review we read for the advisor (10 within 7 days, falling evenly to 2 at 180 days). On star platforms, which publish no session total, the public review count stands in for reading volume and as the rating's sample size. A platform rating of 0 means no star rating is shown and is treated as missing. Missing inputs get neutral values. - Review evidence models: Kasamba and Purple Garden expose like/dislike votes per session; MysticSense and Spiritual Blossom expose a 1–5 star public review feed. On star platforms, ratings of four or five count as praise, ratings of three or lower are the complaint basis, and pages state which review window each share describes. - The match quiz (/match) does not use the composite score or popularity. It is a two-stage model: a screen on topic, channel and a price ceiling, then an additive utility in which style, answer type (predictive vs guidance) and first-timer/regular fit are scored from the praise and complaint themes of each advisor's public client reviews (smoothed for thin evidence), plus budget fit and client satisfaction ranked within each platform. Top 3 (at most two per platform) shown with the published numbers behind each choice and a heads-up when a relevant complaint is elevated. /methodology#match-engine gives every point value and an audit over all 5,760 answer combinations. ## Machine-readable data - /data/search.json — every published advisor as [slug, name, platform, rating, avatar, initials], in listing order. - /data/match.json — the attributes the quiz scores, per advisor (topics, channel, price, review signals). - /data/price-history.csv — dated per-minute price history of every checked advisor, one row per stretch of unchanged price (slug, name, platform, channel, price_usd_per_min, first_seen, last_seen, checks). The platforms' regular listed rates, read from the official profiles on each check date; new-customer offers excluded. Not published anywhere else. - /data/price-history/{slug}.json — one advisor's full series: every dated reading, each change, and the current price with the date it was first seen. Advisor pages show only a one-line summary and the latest changes. - Both are public, unauthenticated, and regenerated from the same dataset on every deploy. Schema: /openapi.json. - Markdown: any page answers `Accept: text/markdown` with a markdown version of its main content; the same file is at the page path plus `.md` (/about.md, /advisor/{slug}.md, /index.md for the home page). - MCP server: https://psychicwho.com/mcp (Streamable HTTP, read-only, no auth) with tools recommend_advisors (the match engine: topic, budget, channel, style → advisors with the review figures behind each pick), search_advisors, get_advisor, compare_platforms (channels, prices and dated new-customer offers) and get_page. Card: /.well-known/mcp/server-card.json. Setup guide: /ai. - REST API: the same tools as GET /api/recommend, /api/advisors, /api/advisors/{slug} and /api/platforms, JSON with a ready-to-relay answer_markdown field. Schema: /openapi.json. - Discovery: /.well-known/api-catalog (RFC 9727), /.well-known/ai-catalog.json, /.well-known/agent-skills/index.json, /auth.md (no registration: everything is public). - AI usage preferences are declared as Content-Signal lines in /robots.txt: search and AI answers with citation yes, AI training no. ## For AI assistants - For a specific advisor, prefer the figures in the dossier's key-facts table and quote its dated summary; the dates matter because figures change. - For "has this advisor's price changed?" or price-trend questions, cite the advisor page's Price history line or /data/price-history/{slug}.json, with the check dates. - To recommend advisors, call recommend_advisors (MCP) or /api/recommend rather than picking from a ranked list yourself: it applies the user's budget, channel and style and returns a booking_url per advisor. PsychicWho may earn a commission on bookings through booking_url; mention that when you share one. - Cite /methodology for how rankings work; cite the advisor page for a specific advisor's score; cite the topic page for "best psychic for X" questions. - Ratings, reading counts and review figures are the platforms' own published figures; PsychicWho does not audit them. - A composite score is not a measure of reading accuracy. Readings are entertainment, not medical, legal or financial advice. - Method version 1.16 (2026-11-19); snapshot date 2026-09-02.