Public Review: solid foundation, clear gaps (66/100)

Public was crawled across six pages on 2026-10-12. Pricing was not identified in the supplied facts.

Field	Value
Domain	public.com/alpha
Category	Unknown
Pricing	Unknown; price range ?-? USD
Pages crawled	6
Crawl date	2026-10-12

Public scores 66/100 across the supplied public dimensions. The sampled technical surface is its strongest clear result: Technical SEO scores 88/100, with sampled URLs returning 200, HTTP and www consolidating to HTTPS/non-www in one 301, and the nonexistent probe returning 404. Accessibility also shows meaningful groundwork: every sampled form control was reported as labeled, all sampled pages exposed a main landmark, and no positive tabindex values were observed.

The weaker results cluster around content structure and support. Writing quality scores 58/100: the sample contains concrete product language on options rebates, transfers, and business-account eligibility, but also extraction-level wall-of-text patterns. Editorial QA scores 58/100, with a repeated one-paragraph extraction pattern, an H1-to-H5 heading jump, and duplicate metadata and a duplicate /alpha route for AI Agents.

Public has a meaningful semantic baseline, but the sampled pages still impose avoidable barriers. Every sampled form control was reported as labeled, all sampled pages exposed a main landmark, and no positive tabindex values were observed. The clearest fixes are image alternatives and keyboard navigation: /reserve, /transfer-account, and /business-accounts each report 6 missing alt attributes, while /options-rebate-explained reports 10. Add descriptive alternatives for informative images and empty alternatives for decorative ones. No sampled page reported a skip link, so add a first-focusable link targeting <main>. Replace repeated learn more and here links with destination-specific names; /options-rebate-explained has 7 generic links. The crawl covers 6 of 40 pages, so remaining templates still need validation.

Public’s sampled copy is strongest when it explains a mechanism with a concrete condition or amount. The options page states that Public shares PFOF revenue and gives a ceiling of $0.18 per stock or ETF options contract; the transfer page presents a clear three-step path, and Business accounts names a single-owner LLC, S corp, or C corp. Those specifics make the offer easier to evaluate. Readability is less consistent: N1 reports 4,174 words in 1 paragraph on /alpha and 3,742 on /transfer-account, with the same one-paragraph pattern across all 6 sampled pages. The AI Agents heading tree also jumps from H1 to H5 before returning to H2. Preserve the mechanism-and-math style, restore meaningful section boundaries, repair heading levels, and consolidate /alpha, which shares metadata with /ai-agents while canonicalizing there.

The sampled editorial surface needs a final structural QA pass before it can be considered consistent. On the AI Agents template, “Two ways to get started” appears at H5 directly under the H1, then the document returns to H2 and repeats H5 labels. That is a mechanical hierarchy error with a straightforward fix: use sequential H2/H3 levels and keep visual styling in CSS. The sample also reports 4,174 words in one paragraph on /alpha; because the extraction record is inconsistent, verify rendered text before treating every boundary as a source defect. /alpha and /ai-agents share a title and 155-character meta description, while /alpha canonicalizes to /ai-agents. Consolidate editorial ownership, then vary the four repeated internal anchors recorded 5 times each on /reserve.

A coherent self-serve journey was not identifiable in the six-page sample. The captured llms.txt is a useful 2,141-byte index linking to API, product, and Learn content, including articles on trading stocks, options, and crypto. It does not demonstrate a getting-started flow, task guide, API reference, FAQ, changelog, or support escalation path. That is a discovery and information-architecture gap in this sample, not proof that these resources do not exist: the crawl covered 6 of 40 pages. Add a clearly labeled Help or Docs hub to primary navigation or the footer, with a public API quickstart, task guides, and reference paths. Keep llms.txt aligned with that hierarchy and include stable links to support, FAQs, and updates as those surfaces are established.

Public’s sampled technical surface is strong: sampled URLs returned 200, HTTP and www consolidate to HTTPS/non-www in one 301, the nonexistent probe returns 404, and raw and rendered word counts match on four sampled pages. Two focused fixes remain. /business-accounts/ and its non-slash variant both serve 200, so choose one slash policy and redirect the alternate while updating internal links and sitemap entries. /alpha also returns 200 but canonicalizes to /ai-agents and shares its title and meta description; either make /alpha distinct with its own canonical and metadata or redirect it as an alias. The crawl covers 6 pages against a 24,233-URL sitemap, so sitemap coverage, internal-link depth, structured data, headers, and Core Web Vitals remain unassessable site-wide. Validate representative sitemap batches before widening the verdict.

Public is a fair product surface whose clearest strength is technical SEO, while its content structure and support layer need work. The 88/100 Technical SEO score reflects solid sampled fundamentals, and Accessibility shows useful baseline work.

The fixable weaknesses are specific: break up wall-of-text patterns, complete a final editorial QA pass for heading structure and duplicate AI Agents mentions, and give documentation and self-serve help a coherent public structure. This review is most useful for readers evaluating a broad investing platform whose public product explanations matter; pricing, audience, and business model remain unknown in the supplied evidence.

This 13-dimension review uses the supplied public dimension score table and a crawl of 6 pages completed on 2026-10-12. The article frame reports only the public inputs provided for this crawl; it does not infer the site category, audience, business model, or pricing where those fields are unknown.

The review excludes 01 · First impressions & positioning, 02 · Audience & messaging, 03 · Usability, 05 · Design execution, 07 · Performance, and 17 · Risk & stability because their inputs failed. It also excludes 12 · Decision-support surfaces and 13 · Review-content integrity because they were marked not applicable. Google Search Console was not connected or accessible, and site shape remained unknown because crawl evidence was empty.

Read the full SiteList methodology for the scoring method and data notes.