ZeroRank review (2026): a $39 tracker that now ships fixes, and where it stops
We read ZeroRank's pricing, platform pages and changelog on 28 October 2026. Plans and the answer budget, tracking, Prompt Research, Google Analytics-based AI traffic, workflows, MCP and API, and the measurement layers it doesn't publish.
Quick verdict
ZeroRank gives more for $39 a month than most AI visibility tools give for three times that: 17 listed AI platforms, unlimited seats and prompts, an MCP server with write tools, and approval-gated workflows that can publish a fix to WordPress or Webflow. It is a good buy for small teams and agencies. It stops at measurement depth. There are no published server-side crawler logs, AI traffic comes through Google Analytics with limits the vendor states itself, and shopping, ads, personas and city-level tracking aren't on its pages. Teams that need those should look at Promptwatch.
At a glance
| What we checked | What ZeroRank publishes |
|---|---|
| Entry price | Starter $39/mo; a full report on your domain before you enter a card |
| Engines | 17 platforms on every plan; names ChatGPT, Perplexity, Gemini, AI Overviews, Claude, Copilot, Grok, DeepSeek |
| Seats | Unlimited on every plan |
| Crawler data | Readiness check (robots.txt, llms.txt, what each crawler got back); no server-side logs |
| AI traffic | Google Analytics connection; vendor calls it a lower bound |
| Fix layer | Recommendations, Copilot diffs, AI articles, agentic workflows on every paid plan |
| Data access | MCP server and REST API on every paid plan |
| Review score | No G2 score published |
Sources: zerorank.ai/pricing, the homepage, /about, three /platform pages and the changelog (latest entry 19 September 2026), all read 28 October 2026. zerorank.ai/llms.txt returns a 404.
Plans and the answer budget
ZeroRank repriced after August 2026, when Starter was $89. The current plans are sized by a monthly budget of AI answers.
| Plan | Price | AI answers a month | Prompt Research runs | Articles + optimizations | Workspaces |
|---|---|---|---|---|---|
| Starter | $39/mo | up to 2,400 | 10 | 2 + 2 | 1 |
| Growth (most popular) | $199/mo | up to 13,500 | 50 | 12 + 12 | up to 10 |
| Agency | $499/mo | up to 31,500 | 150 | 36 + 36 | unlimited |
| Enterprise | Custom | custom | custom | custom | custom |
The vendor's worked examples assume three standard models and the full budget. Starter covers 26 prompts daily, 160 weekly or 800 monthly. Growth covers 150, 900 or 4,500. Agency covers 350, 2,100 or 10,500. Scheduled and manual runs draw on one pool across workspaces, and higher-cost models use more of it. We saw no yearly prices in the page text.
The pricing text also lists white-label dashboards, emailed client reports, SSO and a dedicated account rep on every card, without marking exclusions. The June 2026 changelog tied white-label dashboards and emailed reports to the then Pro plan. Ask which plan includes them before you buy. Separately, ZeroRank has sold AppSumo lifetime deals; those are not the SaaS prices above.
Tracking
ZeroRank runs your prompts against live models and reads the answers back. Its line is "nothing is inferred from search rankings". Each prompt shows visibility, sentiment, average rank, mentions, AI volume, location and tags, and you can set cadence per prompt: daily, weekly, monthly or manual. The Chats view keeps every model's answer word for word with its sources, which is the view we'd use first when a client asks "what did ChatGPT actually say?"
Sources groups citations by domain and URL and labels each domain Corporate, Editorial, UGC, Reference, Institutional or Other. A UGC tab and gap analysis sit alongside.
Prompt Research
Shipped in July 2026. You give it a topic, a competitor domain or a question, and it returns prompt ideas with estimated monthly AI search volume, Google volume where measured, a 12-month trend and a country picker. Connect Search Console and it adds those signals. The AI Volume column then appears on the prompts table, so you can sort a tracked set by demand. Starter's 10 runs a month are enough to seed a project, not to research continuously.
AI Traffic
Also July 2026. Connect Google Analytics and ZeroRank reports sessions, visitors, engagement and conversions from ChatGPT, Perplexity, Gemini, Claude, Microsoft Copilot, Grok, DeepSeek, Mistral, Meta AI and You.com. A funnel runs from seen in AI answers to visited to converted, and landing pages show their citation counts. There's an API endpoint for it.
ZeroRank is candid about the limits. Google reports AI Overviews clicks as organic and Copilot referrals as bing.com, so the AI traffic figure is a lower bound. GA data can lag up to 48 hours.
Workflows, Copilot and content
This is the part that has grown fastest. Recommendations rank on-page and off-page actions by how many tracked prompts they touch. Copilot, a conversation since late August 2026, drafts a change as a diff against the live page, and nothing publishes until it's approved. With WordPress or Webflow connected, the approved change is pushed.
Agentic Workflows are trigger based. A typical flow starts when an answer drops you, runs a Citation Agent to pull the sources cited instead and a Gap Agent to score the gap, writes a brief, posts to Slack and re-checks on the next scan. Templates include an FAQ block for AI answers, a product page FAQ builder, a forum reply drafter, an article translator and a blog post from a YouTube video. Runs can be scheduled daily or weekly. The vendor says workflow automation is on every paid plan.
AI articles go through brief, draft and score, and the changelog says drafts are written by Claude Sonnet 5. Higher tiers can bring their own key through gateways such as OpenRouter.
MCP, API and agency features
The MCP server launched read-only and gained write tools on 22 July 2026, so an AI client can manage prompts and tags. Setup guides cover Claude Code, Claude Desktop and Cursor. The REST API arrived on 29 June 2026 with scoped zr_live_ keys. Both are on every paid plan.
For agencies, shareable client dashboards (June 2026) support a public read-only link, password protection and white-label branding with your own logo, colors and domain. Branded email reports go out through your own SMTP. The interface runs in more than a dozen languages.
Company and proof
Founded 2025, self-funded, nine people on the About page. Kelvin Çobanaj is CEO; the CMO previously worked at SurferSEO. About says "1k+ brands trust us" and the homepage says "loved by over 4,000 experts"; treat both as vendor marketing. The case studies are vendor-published: LanderLab (4.6x AI visibility), Clickflare (3x mentions in two months) and AdPlexity.
Where it stops
These layers are not on ZeroRank's published pages: server-side AI crawler logs and a crawl-to-citation path, a first-party visitor tracking script, ChatGPT Shopping tracking, sponsored-ad tracking, personas and state or city targeting. Its CMS publishing names WordPress and Webflow only.
Promptwatch publishes each of those. Agent Analytics logs real AI crawler visits from your CDN from Professional ($245). Visitor analytics use Promptwatch's own script or GTM template from Essential ($95). ChatGPT Shopping, Ads Radar and city tracking come with Business ($579), and Ads Radar and city tracking are on every agency plan.
Verdict
For a small team that wants wide engine coverage, everyone logged in, and a workflow that turns a lost answer into an approved page change, ZeroRank at $39 is excellent value. For a team that has to explain why an engine isn't citing a page, or what AI search earned last quarter, the measurement isn't there yet. That's when we'd move to Promptwatch. Start on its free Explore plan at promptwatch.com and check your crawler logs first.