VisiScan vs HubSpot AI Search Grader 2026
Two free AI snapshots: VisiScan's four-engine local scan versus HubSpot's three-engine AEO Grader. Sample counts, $49 report, $50 AEO upsell, and a third column for a real tracker.
VisiScan and HubSpot AI Search Grader are both free first clicks. They are not the same snapshot. VisiScan (visiscan.app, Mike Holp, pricing.md 24 August 2026) asks 5 buyer questions across ChatGPT, Claude, Perplexity, and Gemini, two samples each, up to 40 live answers. HubSpot's tool, rebranded AEO Grader in 2026 after the Xfunnel acquisition, runs ChatGPT, Perplexity, and Gemini only. No Claude. No published question count like VisiScan's 5 / 2 math.
The two free scans are two different snapshots. VisiScan runs four engines including Claude, with a published question count of 5 questions and 2 samples each, up to 40 live answers. HubSpot's tool, rebranded AEO Grader in 2026 after the Xfunnel acquisition, runs three engines and skips Claude, with no published question count like VisiScan's 5 over 2 math. The difference matters if Claude is in your brief, because only one of the two free scans covers it.
If you wanted a conversation starter, either works. If you wanted a meter, neither does. Promptwatch is the column for standing prompts, crawler logs, and conversions.
The use-case split is the first decision. A conversation starter is a one-off grade that gets a meeting on the calendar. Either free scan does that. A meter is a standing measurement that stores prompts over time and joins them to crawler logs and conversions. Neither free scan does that. Promptwatch is the column for the meter, because it stores the standing prompts, the crawler logs, and the conversions that a one-off grade cannot.
Snapshot
| VisiScan | HubSpot AEO Grader | Promptwatch | |
|---|---|---|---|
| Free | 5 questions, 4 engines, up to 40 answers, 1 scan per URL | Instant 0-100, no account, no run cap on the listing | Explore: 10 ChatGPT prompts |
| Engines | ChatGPT, Claude, Perplexity, Gemini | ChatGPT, Perplexity, Gemini | Longer list including Grok, Overviews, AI Mode |
| Paid next | $49 report (144 answers) or $29/mo weekly Monitor | HubSpot AEO $50/mo, ~25 prompts, 28-day trial | Essential $95/mo |
| Methodology | Public page; live vs replay labeled | Unpublished; scores "not guaranteed" | Fact sheet, prompt/response caps |
| Files / publish | Schema and llms.txt on $49 | Shareable grade | Content Agents on paid |
| Crawler logs | Readiness signal | None on this listing | Agent Analytics on Professional |
Read the table by column. The free row shows the snapshot shape: VisiScan is 5 questions across 4 engines up to 40 answers with 1 scan per URL, HubSpot is an instant 0 to 100 with no account and no run cap on the listing, Promptwatch Explore is 10 ChatGPT prompts. The engines row shows the coverage: VisiScan includes Claude, HubSpot skips it, Promptwatch carries a longer list including Grok, Overviews, and AI Mode. The paid-next row shows the upsell: VisiScan is a $49 report with 144 answers or a $29 per month weekly Monitor, HubSpot AEO is $50 per month with about 25 prompts and a 28-day trial, Promptwatch Essential is $95 per month. The methodology row shows the transparency: VisiScan has a public page with live versus replay labeled, HubSpot is unpublished with scores "not guaranteed," Promptwatch has a fact sheet with prompt and response caps. The files row shows the output: VisiScan ships schema and llms.txt on the $49 pack, HubSpot ships a shareable grade, Promptwatch ships Content Agents on paid. The crawler-logs row shows the proof layer: VisiScan has a readiness signal, HubSpot has none on this listing, Promptwatch has Agent Analytics on Professional.
HubSpot weights sentiment up to 40 points of the grade, plus presence quality, brand recognition, share of voice, market position. VisiScan splits visibility (named in the answer) from readiness (can AI find you). llms.txt: trust and comparison intents at 1.5x. Different score, different job.
The scoring methods are different jobs. HubSpot weights sentiment up to 40 points of the grade, plus presence quality, brand recognition, share of voice, and market position, which produces a single 0 to 100 that is heavy on how you are described. VisiScan splits visibility (named in the answer) from readiness (can AI find you), and weights llms.txt trust and comparison intents at 1.5x, which produces two scores that separate being mentioned from being fetchable. Different score, different job, so do not compare the two numbers as if they measure the same thing.
HubSpot mention counts look aggregated from indexed content, not query volume, on our listing. VisiScan claims live answers with sample counts. Treat both as dated observations. LLM output moves.
The provenance caveat is the time caveat. HubSpot mention counts look aggregated from indexed content, not from query volume, on our listing. VisiScan claims live answers with sample counts. Treat both as dated observations, because LLM output moves between checks, and a grade from last month is not a grade from this month.
VisiScan's case
Claude on the free pass. Question math you can audit. Localization from the site. $49 pack with quotes and paste-in files. 7-day refund if five new things are missing. Monitor at $29/mo if weekly 12 questions are enough. Harborlight demo is illustrative. 328 scans / average 25 are vendor. Email copy versus $49: screenshot.
The VisiScan case is the auditability case. Claude is on the free pass, which is the differentiator against HubSpot. The question math is published, so you can audit how many questions and samples produced the grade. Localization comes from the site. The $49 pack ships quotes and paste-in files. The 7-day refund applies if five new things are missing. Monitor at $29 per month is the weekly row if 12 questions are enough. The Harborlight demo is illustrative, not a result. The 328 scans and average 25 are vendor numbers. The email copy versus $49 comparison is a screenshot, not a side-by-side in the article.
HubSpot's case
Zero setup, no 1-scan-per-URL cap on the listing, shareable exec artifact. Rebrand to AEO Grader, old URL redirects. The $50 AEO upsell is closed-loop at about 25 prompts. One documented test on our listing showed 0% visibility for a domain other tools still showed as present. Auto competitor and persona setup needed manual correction in that testing. Mixed knowledge cutoffs (GPT variant Aug 2025, Gemini Jan 2025, Perplexity live on that write-up) mean sub-scores are not the same clock.
The HubSpot case is the artifact case. Zero setup and no 1-scan-per-URL cap on the listing make it the lowest-friction scan. The shareable grade is an exec artifact. The rebrand to AEO Grader redirects the old URL. The $50 AEO upsell is closed-loop at about 25 prompts. The documented test on our listing showed 0 percent visibility for a domain other tools still showed as present, which is the undercount risk. Auto competitor and persona setup needed manual correction in that testing. Mixed knowledge cutoffs, with a GPT variant from August 2025, Gemini from January 2025, and Perplexity live on that write-up, mean the sub-scores are not on the same clock, so a single grade blends different recencies.
The third column
Promptwatch Explore is free without HubSpot's unpublished method and without VisiScan's one-scan cap. Essential $95: 50 prompts, 6,000 responses, citations, visitor events, five AEO articles. Professional $245: 25M crawler logs. No $29 Promptwatch plan. VisiScan Monitor at $29 is weekly four-engine re-grades, not that stack.
The third column is the standing column. Promptwatch Explore is free without HubSpot's unpublished method and without VisiScan's one-scan cap. Essential at $95 buys 50 prompts, 6,000 responses, citations, visitor events, and five AEO articles. Professional at $245 adds 25 million crawler logs. There is no $29 Promptwatch plan, and VisiScan Monitor at $29 is weekly four-engine re-grades, not the standing prompt plus crawler log plus conversion stack.
Pick
Both free scans if you like two mirrors. VisiScan if you need Claude and a written sample recipe. HubSpot if you only wanted a logo-friendly grade for a slide. Promptwatch if the next invoice is supposed to measure anything next week.
The pick is the use-case pick. Run both free scans if you like two mirrors on the same domain. Pick VisiScan if you need Claude and a written sample recipe, because it is the only free scan that covers Claude and ships paste-in files. Pick HubSpot if you only wanted a logo-friendly grade for a slide, because the shareable 0 to 100 is built for that. Pick Promptwatch if the next invoice is supposed to measure anything next week, because the standing prompts, crawler logs, and conversions are the objects that measure over time, not once.