Four-Leaf Review: strong product, slow load (83/100)

Four-Leaf is a paid SaaS product for technical professionals seeking jobs.

Fact	Value
Domain	four-leaf.ai
Category	AI-powered job search for technical professionals
Pricing	Paid; 5–20 USD
Pages crawled	40
Crawl date	2026-09-01

Four-Leaf makes a strong first impression because its audience and product role are clear: it is an AI job-search assistant for technical professionals. Positioning and audience messaging score 92/100 and 90/100, while design execution reaches 94/100. Usability and accessibility both score 85/100, giving the public site a solid, understandable structure.

Performance is the clear counterweight. The homepage records a 6.7-second mobile LCP against the cited 2.5-second recommendation, making the 58/100 performance score the lowest public dimension. Writing quality remains strong at 82/100, and technical SEO scores 88/100. Risk and stability scores 86/100, with public URLs served over HTTPS and the sitemap listing 1,506 URLs.

The remaining opportunity is decision support: comparison surfaces score 75/100, while review-content integrity scores 85/100 because comparison methodology and update timing are not disclosed.

Four-Leaf makes a specific, ownable promise: it is an AI job-search assistant built around the user. The claim is supported by seven named stages, including mock interviews, resume tailoring, cover letters, salary negotiation, email drafting, job matching, and LinkedIn optimization. The audience is equally clear, with roles such as Software engineer and Data scientist named directly. Netflix and Anthropic logos, plus specific Trustpilot testimonials, place proof beside the promise. The /compare surface and comparison article also make the bundled workflow legible against alternatives. Keep the core positioning; add more named customer evidence or outcome metrics if the product wants stronger proof of scale.

Four-Leaf speaks directly to technical professionals and answers the main buying questions with unusually concrete detail. The site names supported roles, explains the workflow, shows outcomes through Trustpilot testimonials, and gives two clear price points: a $5 five-day pass and $20 per month. Its language stays close to the audience, using phrases such as “built around you” and “find the roles that actually fit.” Comparison pages explain why the bundled workflow matters, while the free-trial CTA gives visitors a clear next step. The remaining opportunity is decision support for enterprise buyers: add a pricing calculator, ROI comparison, or more detailed case studies and outcome metrics.

The public site supports its core tasks well, but two labeling choices add avoidable friction. The homepage communicates the product purpose clearly, and users can reach pricing through navigation or homepage links. Support is visible through an email address. However, a homepage link labeled “See pricing” is less direct than “View Pricing” or “See Plans,” and the contact page offers email without a visible phone number or another direct method. Blog links also mix labels such as “Resources” and “AI interview copilots,” making content type less predictable. Rename the pricing CTA, broaden contact options where appropriate, and standardize article and guide labels.

Four-Leaf has a solid accessibility foundation, with nav, main, and footer landmarks and generally appropriate ARIA attributes. The audit still found several fixable gaps. One homepage image, the hero background, has no alt attribute, while 21 images use empty alt values that need to be distinguished between decorative and informative content. Some interface elements fail the reported contrast requirement. The contact page has no skip link, and the blog has an inconsistent heading sequence. Add descriptive text where images convey content, retain empty alt only for decoration, correct contrast, add a “Skip to main content” link, and repair the blog heading hierarchy.

Four-Leaf’s design execution is a major strength, combining coherent components with careful mobile behavior. At a 390px viewport, the audit found no horizontal scroll; 61 tap targets met the 44×44px size and 8px separation checks. Sampled text pairs had zero WCAG AA contrast failures. The system also stays disciplined: two button styles, one card-style cluster, consistent padding, and section spacing of at least 64px on desktop and 48px on mobile. Typography remains readable, with body text at least 16px on desktop and 14px on mobile. Maintain this system as new pages and components are added.

Mobile performance is Four-Leaf’s clearest weakness: the homepage LCP is 6.7 seconds against the cited 2.5-second recommendation, and INP is 480ms. The LCP is a 131KB hero image rendered at 1007px wide for a 750px viewport; the audit records no lazy-loading attribute and no preload. The page also has 28 blocking requests, including 19 JavaScript or CSS files and a 350KB main.js bundle. Ten third-party domains add more render pressure, while CLS is 0. Prioritize the hero resource, defer non-critical scripts, split the large bundle, and load third-party scripts asynchronously.

Four-Leaf’s writing has a distinct, practical voice, and the blog shows the strongest execution. “Step 1: extract the keywords (5 minutes)” gives advice a concrete timeframe, while “keyword Tetris” supplies a memorable product-specific metaphor. The homepage is harder to scan: its average sentence length is 30.6 words, which is costly on mobile. The About page opens with “Finding a job is like finding a Four-Leaf clover,” an analogy that delays the value statement. Three pages also share the homepage’s meta description. Shorten homepage sentences, replace the About H1 with direct product value, and write unique descriptions for login and signup.

Four-Leaf provides comparison surfaces, but they stop short of recommending a choice for a specific buyer. The pricing page links to “How we compare” and invites visitors to browse inclusions without a visible, segmented recommendation. Its plan tables also use all-checkmark rows, which makes trade-offs harder to see. On the compare page, “24 roles, plus custom” is not defined well enough to guide a decision. Add recommendations for segments such as solo job seekers and teams of 5–10, while keeping them overridable. Replace generic axes with criteria tied to the audience, including interview support, salary negotiation coaching, and cover-letter generation.

Four-Leaf’s comparison content is useful, but its trust signals are incomplete. The main compare page and individual versus pages do not show how comparisons were conducted, which criteria were used, or when the content was last updated. That omission makes otherwise clear claims harder to verify. The Four-Leaf versus Teal page also contains three affiliate links to Four-Leaf products without a visible disclosure. Add a concise methodology block with criteria, evidence basis, and update date to every comparison surface. Place plain-language affiliate disclosure above the first affiliate link. These changes would make the content easier to evaluate on its own terms.

Four-Leaf shows a stable public search surface. The homepage returns 200, public URLs use HTTPS, robots.txt allows the public surface, and the sitemap contains 1,506 URLs. Raw and rendered content remain closely aligned. The remaining risks are narrower: canonical inconsistency on a small set of legal, authentication, and demo templates, missing image alternative text, and a trailing-slash duplicate. Review canonical generation for those templates and choose one slash convention. Connected Search Console data would provide the next layer of validation for coverage and performance.

Four-Leaf’s content shows disciplined editorial choices, especially in the blog. The practical voice avoids generic markers, and the time-bound instruction “Step 1: extract the keywords (5 minutes)” gives readers a clear action. The About page fails the lede test because its H1, “Finding a job is like finding a Four-Leaf clover,” does not state the product’s value. Feature pages have a second structural gap: they list tools without 40–60-word self-contained answer paragraphs under each feature H2. Move the value proposition into the About opening and add concise answer paragraphs while preserving the blog’s specific tone.

Technical SEO is strong overall: sampled public pages return 200, HTTPS is enforced, robots.txt permits public paths, and raw and rendered content are close to parity. The sitemap lists 1,506 URLs. Four issues deserve targeted cleanup. Legal and authentication templates canonicalize to the homepage rather than following one deliberate policy. The trailing-slash variant of /how-it-works/ serves 200 without redirecting, and sampled titles run from 67 to 77 characters. Set intentional canonicals, normalize the slash convention, and shorten long titles. Image alternatives are also part of the cleanup: one image lacks an alt attribute and 21 images have empty alt attributes; distinguish informative and decorative images when adding appropriate alternatives.

Four-Leaf is a strong choice for technical professionals who want one clearly positioned job-search product spanning multiple workflow stages. Its 92/100 positioning, 90/100 audience messaging, and 94/100 design execution show a site that communicates its purpose with unusual clarity.

The priority is mobile performance: the 6.7-second LCP is too slow, and the review attributes it mainly to an oversized hero image and its loading treatment. Four-Leaf should also make comparison pages more useful with actionable guidance, and disclose how and when those comparisons were produced. The site explicitly names software engineers and data scientists among its supported roles.

This 13-dimension review combines a crawl of 40 pages on 2026-09-01 with the public dimension score table. It covers positioning, audience and messaging, usability, accessibility, design execution, performance, writing quality, decision-support surfaces, review-content integrity, risk and stability, editorial QA of content, and technical SEO.

The overall score is the rounded persisted SiteList score: 83/100. See How SiteList scores for the scoring method and data notes.