Legalysis Review: strong product, fixable gaps (76/100)

Field	Value
Domain	legalysis.co
Category	AI-assisted contract review and legal document analysis
Pricing	Brief (Free), Counsel ($29/mo), and Chambers ($89/mo)
Pages crawled	38
Crawl date	2026-10-08

This 13-dimension review of the supplied 38-page crawl finds Legalysis presents a clear product for lawyers, founders, and small legal teams seeking a careful first-pass contract review. Its strongest dimensions are first impressions and positioning (84/100), usability (84/100), audience and messaging (82/100), design execution (78/100), decision-support surfaces (78/100), risk and stability (78/100), and technical SEO (78/100). The supplied summaries consistently describe a legible offer, a clause-by-clause mechanism, a human-in-the-loop stance, visible pricing surfaces, and a crawlable, server-rendered site.

The weaker part of the score is execution under pressure. Performance scores 67/100, with mobile LCP at 4.65 seconds on the homepage. Writing quality scores 68/100, while editorial QA scores 62/100, reflecting dense comparison content and a need for a stronger final pass. Accessibility is comparatively solid at 73/100, but it remains a fair-band dimension rather than a leading strength. Overall, within this supplied 38-page crawl, Legalysis has a credible product and a clear public proposition; its next gains come from reducing mobile friction and tightening editorial consistency.

Legalysis is immediately legible as AI-assisted contract analysis for lawyers and founders. The hero line, “Read every line. Miss nothing,” is concrete, and the supporting copy explains clause-by-clause analysis, graded risks, and one-page summaries. The boundary is equally clear: it is “an aid, not a verdict,” rather than a replacement for legal judgment. The strongest proof comes from the clause demonstrations and visible $0, $29, and $89 tiers. The homepage still gives equal billing to “the lawyer who values precision” and “the founder who has to sign anyway.” Make one buyer primary in the hero, then give the other a clearly labeled route. Keep the privacy promise, but add a concise retention explanation near upload and pricing.

A first-time visitor can identify Legalysis quickly: it reads contracts clause by clause, flags risks, and returns a plain-English summary. Features, Pricing, and the free first-contract offer are visible from the main navigation, while the pricing tiers provide concrete audience cues. The remaining friction is professional-fit detail. The supplied pages do not explain collaboration, integrations, upload formats, or how analysis enters a review or redline workflow. The free path is also under-specified: “Analyse your first contract free — no card, no setup” points to pricing, but the next product step is not shown. Add a short workflow section for founders and legal teams, list supported file types, and show the upload handoff and expected result.

Legalysis passes the ten-second comprehension task and makes the core journey easy to follow. The homepage explains upload, clause-level risk flags, and a plain-English summary; Features, Pricing, Contact, and Journal are reachable in one or two steps. The main usability break is at conversion: homepage “Start free” and “Analyse a contract” route to /pricing, while the pricing inventory records “Start free” as href # with no form or signup destination. A visitor reaches the commercial page but cannot verify where the product begins. Make that control resolve to a confirmed start flow or label the handoff plainly. Mobile performance also adds friction: homepage mobile LCP and TTI are both 4.7s.

The sampled pages establish a useful accessibility baseline: contact fields are labeled, core landmarks are present, language metadata is exposed, and no positive tabindex values were found. The verified barriers are structural and visual. Pricing follows h1 → h3 → h3 → h3 → h2 → h4 → h4 → h4, and the contrast audit fails on the homepage and sampled pricing and journal pages. No skip link is detected in the supplied samples. The journal and comparison samples also contain missing image-alt decisions. Repair shared heading templates, measure and correct the failing color pairs, add a focus-visible skip link, and complete informative alt text. A manual keyboard and screen-reader pass should follow.

Legalysis has a deliberate visual system: 24 inventoried colors with custom-property usage, IBM Plex Sans, IBM Plex Mono, and Spectral, 17px mobile body text, and a 44.2px homepage h1. At 390px, the homepage reports scrollWidth390: 390, so no horizontal overflow was observed there. Execution weakens on mobile interaction and accessibility. The homepage has 18 small targets and Features has 13, including an 18×18px header link, 43px clause buttons, and 26px footer links. Raise interactive hit areas to at least 44×44px with 8px separation. Contrast and heading-order audits also fail, while homepage mobile LCP is 4.7s. Fix touch targets and contrast before polishing the token system.

Legalysis is fast in the supplied desktop lab runs but needs a mobile critical-path pass. The homepage reaches 4,652.5 ms mobile LCP, against the 2.5-second good threshold; PageSpeed estimates 2,320 ms of render-blocking savings and 903 KiB of image-delivery savings. The hero is 473,201 bytes and 2048px wide for a 542px display slot. The sampled journal article adds CLS 0.187 and an unsized-image failure. Serve responsive image variants at the rendered size, defer non-critical blocking work, and reserve journal media dimensions. Re-run the same mobile lab trace after each change. These are directional lab results because no CrUX field metrics were supplied.

Legalysis writes its product promise clearly: “Upload a contract and Legalysis reads it clause by clause” is specific, audience-aware, and supported by concrete clause examples. The editorial weakness appears at scale. N1 reports one paragraph and 1,200 average paragraph words on the homepage, while the comparison sample reports 2,901 words and 29.3 average sentence words; rendered verification is still needed because extraction can flatten structure. That comparison page also reports 726px scroll width at a 390px viewport. Preserve the plain-English voice, but split dense copy into mobile-readable sections and responsive tables or cards. Canonicalize duplicate slash variants, including /pricing and /pricing/.

Legalysis gives visitors real commercial choices: Brief (Free), Counsel ($29/mo), and Chambers ($89/mo), with audience labels and a visible “MOST RETAINED” marker on Counsel. The compare index adds five dated comparison articles, and its mobile cards remain readable. This is guided decision support, not a bare feature list. The next gain is reasoning at the point of choice. Each tier needs a plain-language “Choose this if…” line and an explicit trade-off, supported by concrete limits, seats, usage boundaries, or service scope where those facts exist. The comparison template should also expose its methodology, source basis, and date before the verdict; sampled content contains “How we compared these agents” and “How we tested,” but consistency across the library is not established.

The supplied evidence points to signal inconsistency rather than a crawl shutdown. Robots permits crawling, the sitemap is present, HTTPS consolidates in one hop, and the nonexistent probe returns 404. The main loss-capable issue is /compare: it canonicalizes to /vs while the supplied inventory does not mark it in the sitemap. Choose one preferred URL, align internal links and sitemap treatment, and verify that the final URL returns 200 and self-canonicalizes. Separately, the comparison template reports 728px of scroll width at a 390px viewport, creating a mobile resilience problem. Fix URL policy first, then repair responsive containment. Traffic impact cannot be measured without GSC or analytics.

The sampled corpus is competent but not consistently finished for publication. The clearest QA signal is the comparison sample: N1 reports 2,901 words in one paragraph with 99 sentences averaging 29.3 words, while the crawler reports 726px scroll width at 390px. Because extraction can flatten paragraphs, verify the rendered DOM before rewriting; if confirmed, split the page into readable sections and make comparison content responsive. The crawler also reports duplicate titles and meta descriptions for /pricing and /pricing/, plus one journal image without alt text. Make one slash variant canonical, redirect the alternate, complete the missing image metadata, and add a 390px QA check to the publishing workflow.

Legalysis has a sound crawl and rendering baseline across the supplied 38-page sample. Robots allows crawling, the sitemap index is declared, HTTPS redirects in one hop, the nonexistent probe returns a true 404, and raw and rendered word counts match. The main technical risks are comparison-template overflow and URL inconsistency. /compare returns 200 but canonicalizes to /vs and is not in the supplied sitemap inventory; /features/ also returns 200 without redirecting to a single slash policy. The comparison sample reaches 728px scroll width at 390px, while oversized homepage images include hero.jpg at 462 KB and feature-redline.jpg at 720 KB. Normalize URL signals, repair mobile containment, and compress media at rendered dimensions.

Legalysis is a strong fit for lawyers, founders, and small legal teams that want a careful first pass over contracts rather than an opaque automated verdict. The product is easy to place, its clause-by-clause mechanism is specific, and its human-in-the-loop stance gives the proposition a credible boundary. Usability, design execution, decision-support surfaces, risk and stability, and technical SEO all sit in the strong band.

The fix list is focused. Improve mobile performance, where the homepage LCP is 4.65 seconds and the sampled journal article reaches 3.0 seconds. Give long comparison content a more consistent editorial pass, addressing the fair-band writing quality and editorial QA scores. Legalysis does not need a new story; it needs a faster and more finished expression of the story it already tells.

This 13-dimension review uses the supplied crawl of 38 pages, public dimension scores, and the one-line findings attached to those scores. The crawl date is 2026-10-08. Scores and quoted metrics are reproduced from the provided review inputs; this frame does not add owner-only findings or re-score the site.

Google Search Console enrichment is pending because GSC is not connected and access is unavailable. See How SiteList scores for the review method and dimension definitions.