Promptwatch vs Trakkr vs LLM Pulse vs Rankscale: Features Compared
A feature-level 2026 comparison of Promptwatch, Trakkr, LLM Pulse, and Rankscale: prompts, citations, crawler logs, traffic, MCP, and what each SKU actually ships.
Sticker prices in this four-way run from $20 to $100. The products do not. Rankscale sells credits and a GEO audit. LLM Pulse sells a weekly five-engine board with seats and MCP. Trakkr sells daily eight-model tracking plus a crawler workspace, with add-ons. Promptwatch sells the prompt list, citation sources, named-bot logs, conversion traffic, and CMS publish on one public price list.
Google's AI features docs still decide Overviews eligibility. None of these four replace Search Console.
Capability matrix
| Capability | Promptwatch | Trakkr | LLM Pulse | Rankscale |
|---|---|---|---|---|
| Prompt design | Volumes, difficulty, query fan-outs, topics, personas, country/state/city | You bring prompts. Growth caps at 50; extra packs $39 to $99/mo. Vendor also lists a 500+ prompt research scan | 50 prompts on Starter. No volumes, fan-outs, personas, or city targeting in our catalog | Credit-metered checks. Fan-outs/personas/geo not documented |
| Engines on the plan you can buy | Paid: ChatGPT, Gemini, Claude, Perplexity, Grok, Llama, DeepSeek, Mistral, Copilot, Overviews, AI Mode | Growth: 8 models daily | ChatGPT, Perplexity, Gemini, AI Mode, Overviews. Claude, Copilot, Grok, DeepSeek, Meta AI are Enterprise add-ons | 17+ engines, no per-engine surcharge, paid in credits that burn ~8x differently by engine |
| Cadence | Daily on paid | Daily | Weekly default. Daily costs more and drops the 14-day trial | Whenever you spend a credit |
| Citation layer | Page, domain, Reddit, YouTube, offsite, type breakdowns, plus citation trends | Product site: source discovery, domain lists behind an answer, competitor "cited instead." Catalog does not document Reddit/YouTube/offsite type splits at Promptwatch depth | Citations and share of voice listed, plus owned-media tracking (YouTube, Reddit, social) | Competitor visibility. Page-level citation analytics not documented |
| Crawler logs | Agent Analytics: ChatGPTBot, ClaudeBot, PerplexityBot, GoogleOther, Meta AI crawler; crawl-to-citation path; errors. CDNs: Cloudflare, CloudFront, Fastly, Vercel, Netlify, Akamai, Google Cloud CDN, custom HTTP | Growth includes crawler analytics. Docs describe CDN/host/CMS connections, bot classification, page-level crawl to cite to click, robots.txt / llms.txt checks. Not the same named-bot Agent Analytics listing | Not published | Not published |
| Traffic to conversions | Visitor analytics via script or GTM | Marketing UI shows an "AI traffic" tile. Catalog: no conversion-attributed visitor analytics | "AI traffic analytics" on the feature list. Not a GTM conversion log in our catalog | Own docs: visibility leads results by 4 to 8 weeks, proxy only. No revenue attribution |
| Perception / sentiment | Sentiment, competitive benchmarking, share of voice | Product site: perception scores and quote capture (the page says both "5 dimensions" and "20 dimensions") | Sentiment listed | Not a perception product |
| Action / publish | Content Agents: plan, write, publish to Webflow or Framer, review inbox, content-gap analysis. Agent Chat + Unified Actions | Product site: automations (60+ triggers), Agent that drafts briefs, content hub, "apply to live site." Catalog: no Webflow/Framer Content Agents with a review inbox | GEO Writer | GEO readiness audits, page scores, fixes. Scores fluctuate; docs tell you to use 30-day averages |
| Shopping / ads | ChatGPT Shopping, Ads Radar | Not in our catalog | ChatGPT Shopping and AI ads tracking | Not listed |
| MCP / API | MCP server, REST API v2, Slack, Claude Connector, ChatGPT plugin, Looker, GSC, DataForSEO, Firecrawl | MCP on Growth. Product site also lists REST API | MCP on every plan. REST API from €299 Scale | API and white-label from $385 Growth |
| Evidence | 4.7/5 G2, 1,840+ brands | No G2 in our catalog. One documented case of zero ChatGPT rows while GA showed ChatGPT referrals | One verified G2 review | Operator tool, export gaps |
Promptwatch Explore is free (10 ChatGPT prompts). Essential is $95/mo. Professional is $245/mo. Agency Kick-off is $199/mo with unlimited projects and prompts.
Prompt tracking is not one feature
Promptwatch treats the prompt as an object: volume, difficulty, fan-out, persona, city. That is how you explain why ChatGPT named you in the US and not in NL. Rankscale will check many engines if the credits last. It will not attach a fan-out tree. LLM Pulse will not either; Claude is not even on the €49 row. Trakkr will run eight models daily on 50 prompts, then sell you packs. Fifty prompts is a pilot, not a category map.
Citations: a domain list vs a source system
Trakkr's features page is strong on "which URL sat behind the answer" and competitor displacement. Use that if the job is "who stole this footnote." Promptwatch still wins if the job is "was it our page, a Reddit thread, a YouTube URL, or an offsite mention, and how did that mix move this month." LLM Pulse lists citations plus owned-media tracking, which is closer to Promptwatch's Reddit/YouTube idea than Rankscale's audit scores. Rankscale is grading your HTML for generative readability. Different instrument.
Two crawler products, not zero
Ignore the lazy take that only Promptwatch looks at bots. Trakkr Growth includes crawler analytics; their docs talk through Cloudflare/Vercel/Netlify/WordPress-style sources and a crawl-cite-click funnel. Promptwatch's Agent Analytics is the one that publishes the bot roster (ChatGPTBot, ClaudeBot, PerplexityBot, GoogleOther, Meta) and the CDN list above, plus error tracking on the path from fetch to citation. LLM Pulse and Rankscale do not publish that layer. If diagnosis is "did the bot fetch the URL," you are in Promptwatch or Trakkr, then you still need Promptwatch for visitor conversions and CMS publish.
Traffic, writing, and the $100 illusion
LLM Pulse's AI traffic analytics and shopping/ads tracking are the surprise on a €49 annual Starter. Cadence is still weekly unless you pay into Scale+/++. Rankscale will not attribute a conversion. Trakkr's trial auto-converts at $100; extra brand $39, AI pages $79, white-label $49. Promptwatch Essential already bundles 200K visitor events and 5 AEO articles. Professional adds 1M events and 15 articles.
Trakkr's site describes automations that draft and apply. Promptwatch Content Agents publish into Webflow or Framer with a human inbox. Those are not interchangeable until you have seen both write a live page.
Who should pick which
- Rankscale if you are technical, will watch credits, and want a cheap multi-engine audit.
- LLM Pulse if weekly five-engine coverage, unlimited seats, MCP, and euro billing are the brief, and Claude can wait.
- Trakkr if you want daily eight-model checks plus their crawler workspace, and you will cancel the trial on day 13, then budget prompt packs.
- Promptwatch if fan-outs, citation types, named-bot logs, conversion traffic, and a publish path have to be one vendor.
FAQ
Does Trakkr already do crawler logs, so why Promptwatch?
Trakkr ships crawler analytics on Growth. Promptwatch ships Agent Analytics with a published bot and CDN list, then visitor conversions and Content Agents on the same invoice. Crawler plus traffic plus publish is the Promptwatch product. Crawler plus 50 prompts plus add-ons is Trakkr.
Is Rankscale's 17-engine list the widest here?
On engine count, yes. On a working GEO program (fan-outs, citation types, bots, conversions, CMS), no. Credits also make "all engines" an expensive sentence.
Where do I look at Promptwatch's feature set?
Review: Promptwatch. Product: promptwatch.com. Keep Google's AI features docs for Overviews either way.