Every Content Quality Audit check

Visitors form a first impression in under 5 seconds, and the hero section is the only content most of them will read before deciding whether to stay or leave. If the hero shows only a product name or an abstract tagline — no description of what the product does or who it helps — the visitor has no way to self-qualify. They cannot determine relevance, so they leave. This is the single highest-leverage piece of marketing content on the site: everything downstream depends on the visitor understanding the offer. An unclear hero turns every other marketing investment (ads, SEO, referrals) into wasted spend.

Why this severity: Critical because visitors who cannot understand the offer within 5 seconds leave before seeing any other content — an unclear value proposition above the fold disqualifies the hero from doing its one job.

Feature-based headlines describe what the product has. Outcome-based headlines describe what the visitor gets. The distinction matters because visitors are not buying a feature — they are buying a result: faster shipping, fewer support tickets, a design that doesn't embarrass them. When the headline names only a feature ("AI-powered analytics dashboard") or a technology attribute ("Real-time collaboration platform"), the visitor must do mental translation work to connect the feature to their own situation. Many of them won't bother. Headlines that name the outcome — "See what's actually driving revenue" — do that translation for the visitor, making the product immediately relevant without effort.

Why this severity: Critical because a feature-framed headline forces every visitor to translate the feature into a personal benefit before they can evaluate relevance — most won't, and will leave before the sub-headline can recover.

A features section that lists only capabilities without explaining their value puts the persuasion burden entirely on the visitor. "Automated backups" tells the visitor what exists; "Automated backups — your work is saved continuously so you can focus on building, not recovery" tells them why it matters. When fewer than 60% of feature descriptions include a benefit framing, the section reads like a spec sheet rather than a sales argument. Visitors who do not see themselves reflected in the copy — because benefits are absent — conclude the product wasn't designed for people like them. The feature-to-benefit translation is the marketer's job, not the visitor's.

Why this severity: High because a features section with mostly capability-only descriptions fails to communicate user value, leaving the visitor to do persuasion work that most won't complete before deciding to leave.

A clear CTA in the hero section is the mechanism that converts intent into action. Without one — or when the only option is a passive link like "Learn more" — a visitor who arrived ready to act has nowhere to go. They must navigate on their own to find what to do next, and most will not. The CTA's text signals what happens when the visitor clicks: "Start free trial" sets an expectation, "Learn more" does not. Primary CTA text that names the action and sets the expectation ("Get early access", "See a demo", "Start free — no credit card") converts at measurably higher rates than generic verbs. The hero CTA is the single highest-value conversion point on the page.

Why this severity: High because a missing or passive CTA above the fold means interested visitors have no clear next step — intent arrives but has no exit path, and the conversion opportunity evaporates.

Visitors who are genuinely interested in your product are also considering alternatives. If nothing on your home page answers "why you and not the other option?", the comparison happens entirely in the visitor's head, with no data you've provided. Generic positioning — "We're the best", "High-quality solution", "Built with passion" — is indistinguishable from your competitors' copy and provides no decision-making signal. Specific differentiation ("The only tool that works without a backend", "Built for indie developers, not enterprise teams", "Teams using this ship 2x more features per sprint on average") gives the visitor something concrete to evaluate. The specificity itself signals confidence and honesty — only a company that believes its claim would make it.

Why this severity: Medium because without specific differentiation, the home page leaves comparison-stage visitors with no signal to choose this product over alternatives — a gap that coexists with otherwise clear messaging but costs conversion among considered buyers.

Generic claims — "fast", "the best", "easy integration", "trusted by teams" — have been used on so many product pages that visitors filter them out on sight. They create no credibility because they cannot be verified and no competitor would dispute them. Specific claims — "average response time: 45ms", "2,400 SaaS teams", "connects to Slack in 2 clicks" — create credibility precisely because they can be verified and could be wrong. A visitor reading specific claims perceives that the company is confident enough to commit to a checkable statement. When more than 40% of a page's claims are generic, the page loses its signal-to-noise ratio: the specific claims get buried under the generic ones.

Why this severity: High because a majority of generic, unverifiable claims across the home page erodes credibility at the moment of evaluation — visitors who are comparing options rely on claims to differentiate, and generic claims provide no signal.

The hero headline and its immediately following supporting text are read as a unit — visitors process them together in their first pass. When the supporting text introduces a different topic than the headline, visitors receive two half-messages instead of one complete one. Neither registers clearly. For example, a headline about speed ("Ship faster") followed by sub-text about security ("Enterprise-grade security for your team") leaves the visitor unsure which of the two attributes is the actual value proposition. The dissonance also signals inconsistency in how the product thinks about itself — a subtle trust signal. The sub-headline's job is to answer the question the headline raises, not to pivot to a new claim.

Why this severity: Low because a mismatched headline and sub-text reduces hero clarity without preventing visitors from understanding the product entirely — the failure is inefficiency rather than a blocking confusion.

Web visitors skim before they read. Heading structure is how a skimmer navigates a page — they read only the headings to decide whether any section is worth their time. Headings that skip levels (`h1` → `h3` with no `h2`) break screen reader navigation and signal structural disorder. Vague headings — "Features", "Details", "Section 3" — provide no information to the skimmer, so every section looks equally irrelevant and nothing gets read. Descriptive headings form a readable summary of the page when taken in isolation: a visitor reading only the headings should be able to tell what the page is about. This is a prerequisite for accessibility, SEO, and basic readability.

Why this severity: High because vague or level-skipping headings prevent visitors from using heading structure to navigate and skim — a core reading pattern on the web that determines whether any below-the-fold content gets read.

Visitors follow a predictable mental sequence when evaluating a new product: first they need to understand what it is and whether it is relevant, then they decide whether it is credible, then they consider the price. When a page violates this sequence — opening with logo bars before explaining what the product does, or showing pricing before features — social proof becomes meaningless (proof of what?) and pricing feels premature (price for something I don't yet understand). The most common structural failure in AI-built sites is leading with trust signals (logos, testimonials) that have no referent yet, because the AI generates them early as credibility markers before the value proposition is established.

Why this severity: High because a section ordering that leads with trust signals or pricing before establishing the value proposition makes those signals meaningless — visitors see proof before they know what is being proved.

Dense blocks of prose are one of the leading causes of visitors leaving a page without reading the content. Eye-tracking research consistently shows that web readers scan in an F or Z pattern — they read the first line of each paragraph and skip the rest unless the first line hooks them. Paragraphs of 8-10 sentences are read by almost no one visiting a marketing site. Breaking copy into short paragraphs (2-4 sentences), bullet lists, and subheadings is not about dumbing content down — it is about matching how humans actually read on screens. A blog post with subheadings every 300 words allows readers to jump to the section they care about; one without subheadings requires sequential reading from the beginning, which most visitors won't do.

Why this severity: Medium because wall-of-text formatting suppresses the amount of content visitors actually read, reducing the effective reach of marketing copy even when the words themselves are well-written.

Content volume on a landing page has a floor and a ceiling. Below the floor (fewer than 300 words), the page doesn't provide enough information for visitors to understand the product and trust it — they leave because they can't evaluate the offer. Above the ceiling in the hero section (more than 50 words above the fold), the value proposition becomes exhausting to parse before the visitor has decided whether to engage. The hero's job is to hook and orient in under 10 seconds; all the explanatory detail belongs below the fold where visitors who are interested will read it. AI-generated landing pages frequently fail both conditions simultaneously: an overloaded hero that tries to explain everything, and a thin total page that adds nothing below the fold.

Why this severity: Low because incorrect content volume — a thin page or an overloaded hero — reduces conversion and trust signals, but is a fixable structure and copy issue rather than a blocking defect.

Placeholder content that reaches a production URL is one of the most immediate trust destroyers a visitor can encounter. Lorem ipsum text signals the site is unfinished — not a beta, not "simple", but abandoned or sloppy. Template strings like "Your headline here" or "Add your description" reveal that whoever built the site did not bother to replace them, which visitors interpret as evidence that the rest of the site (including the product itself) is similarly unfinished. Placeholder images from `via.placeholder.com` or `placehold.co` are common in AI-generated code because scaffolding tools insert them as defaults, and they survive into production when the product is rushed out without a final content review.

Why this severity: Critical because any placeholder content in a production-accessible path destroys visitor trust immediately and signals an unfinished product — the site is not ready to receive visitors.

Hard-to-read copy is invisible copy: visitors who encounter long sentences and passive voice constructions slow down, lose the thread, and stop reading. Research on web reading consistently shows that shorter sentences, active voice, and concrete language produce higher comprehension and lower bounce rates. The problem is compounded for AI-generated marketing copy, which tends to produce grammatically correct but unnecessarily complex prose: abstract noun stacking ("enterprise-grade workflow optimization"), passive constructions ("data is processed", "reports are generated"), and sentences that run to 30 words without a clear subject. These patterns make the copy technically accurate but practically unread.

Why this severity: High because marketing copy with more than 30% long sentences or majority passive voice constructions is significantly less likely to be read and understood — the persuasion argument never lands because the sentence-level experience is too effortful.

Voice inconsistency across pages is a subconscious trust signal. A home page that addresses the visitor as "you" in a warm, direct tone — and a pricing page that switches to "the user shall" in formal, passive constructions — makes the site feel like it was assembled from parts by different people. Visitors experience this as cognitive dissonance: the brand that felt approachable on the homepage suddenly feels bureaucratic. This is especially common in AI-built sites where each page was generated in a separate prompting session, each session producing slightly different copy characteristics. The inconsistency rarely prevents conversion on its own, but it creates friction at the moment visitors are evaluating trust — exactly when friction is most costly.

Why this severity: Medium because jarring voice inconsistency between pages creates cognitive dissonance at trust-evaluation moments, though it rarely prevents conversion entirely on its own.

Grammar and spelling errors on a marketing site signal lack of care — and in markets where visitors are evaluating professional services, SaaS tools, or anything with a business trust component, lack of care in the copy creates doubt about quality of the product itself. The mental model is unconscious but consistent: if they didn't proofread the sentence the visitor just read, what else didn't they check? AI-generated copy is particularly prone to subtle errors: subject-verb disagreement ("your team are"), incorrect apostrophe use ("it's" vs. "its"), and plausible-sounding misspellings ("Intergration") that pass a casual read but fail on close inspection.

Why this severity: Medium because grammar and spelling errors on key marketing pages — especially the hero — reduce perceived professionalism and create trust friction, particularly for business-to-business products where credibility is part of the evaluation.

Jargon clusters — three or more meaningless buzzwords in close proximity — produce the opposite of the intended effect. When a features section reads "Our cutting-edge, next-generation, enterprise-grade platform leverages revolutionary AI to deliver seamless, best-in-class synergistic results", visitors do not perceive quality; they perceive that the author had nothing specific to say and used buzzwords to fill the gap. The credibility signal of a marketing page decreases with each unsubstantiated superlative, because the reader knows from experience that "best-in-class" is what companies claim when they cannot name a specific advantage. Buzzword-heavy copy is also a strong signal of AI-generated content with insufficient review — a perception that undermines the product's quality story.

Why this severity: Low because jargon clusters reduce copy credibility and signal AI generation without review, but usually coexist with enough real information that visitors can still evaluate the product.

Factual contradictions between pages — different numbers for the same metric, pricing on the home page that doesn't match the pricing page, the product name spelled two different ways — damage credibility more than any individual error. The visitor who navigated from the home page to the pricing page and found different numbers now doesn't know which one is true, and the doubt that creates extends beyond the specific contradiction to the entire site. AI-built sites are particularly prone to cross-page inconsistency because each page was generated in a separate session: the home page might say "100+ audits" and the pricing page, generated three weeks later, says "50+ audits available." Neither is deliberately wrong — the information simply drifted.

Why this severity: Low because cross-page factual contradictions undermine credibility at the comparison stage — visitors who found the inconsistency lose trust in the accuracy of all the site's claims.

Pricing friction is one of the top reasons visitors leave SaaS and software product pages without converting. When a pricing page shows only "Contact us for pricing" for a product that is clearly not enterprise-sales (a solo developer tool, a consumer SaaS, an indie product), it signals one of three things: the price is high enough to be embarrassing, the product isn't ready, or the team is following a playbook that doesn't fit their product. Visitors who cannot find a price go find a competitor who shows theirs. Transparent pricing — even a range, even "starting at" — converts better than gated pricing for non-enterprise products. The tier contents matter too: a pricing page with three unnamed tiers and no description of what each includes requires a sales call to understand, which is an unacceptable friction for any self-serve product.

Why this severity: High because a missing or opaque pricing page removes the information visitors need to make a purchase decision — they leave to find a competitor whose pricing is visible.

Visitors who have understood your value proposition and are close to converting often have one or two blocking questions — "how does the trial work?", "can I cancel?", "is my data private?" A FAQ section or documentation link is the fastest way to resolve those questions without requiring a support interaction. Without it, interested visitors who hit an unanswered question leave to search for the answer elsewhere, and many don't come back. The friction of "I have a question and can't find the answer" is responsible for a measurable share of high-intent abandonment. A FAQ section with even 3-5 questions about common objections prevents this drop-off at minimal implementation cost.

Why this severity: Medium because missing FAQ or documentation means visitors with blocking questions have no self-service resolution path — a friction point that causes high-intent abandonment right at the conversion moment.

A blog or resources section without navigational structure is harder to trust and harder to use than no blog at all. Visitors who land on a blog index page with no post dates cannot assess content freshness — for any topic where recency matters (tech, compliance, product features), undated posts are untrustworthy. Blog posts without a visible path back to the index trap visitors: they read one post and cannot find the next one, so they leave the site entirely. For SEO, a blog listing page with accessible titles and dates signals a maintained content repository to crawlers; a blog with no listing page and posts only accessible via direct URL provides minimal SEO benefit and zero organic discovery.

Why this severity: Low because a structurally deficient blog is harder to use and trust than no blog, and provides diminished SEO benefit — but the impact is limited to visitors who actively seek out the blog.

Visitors use freshness signals as proxies for whether the product is actively maintained and whether the company is still operating. A footer copyright year that reads "© 2022" in 2026 is a four-year-old stale signal that the site has not been touched. For any product that is software — where bugs get fixed, features get added, and security matters — an outdated copyright year creates doubt about whether the company is still around. This is especially problematic for AI-built projects where the site may have been generated and deployed once, then not updated. The fix is one line of code. Blog post dates that are 18+ months old similarly signal abandonment to visitors who value recency of information.

Why this severity: Info because an outdated copyright year or stale content signals possible abandonment without causing direct product harm — a trust signal deficit rather than a functional defect.

Social proof addresses the most common implicit question every new visitor has: "Can I trust this?" A site with no social proof — no testimonials, no logos, no usage statistics, no review badges — asks visitors to trust purely on faith. For a new visitor who has never heard of the product, faith is in short supply. Even minimal, authentic social proof changes the calculus: one real testimonial with a name and company attached is worth more than a dozen anonymous quotes, because it is verifiable. Usage statistics ("2,400+ teams") signal scale even without naming customers. For AI-built projects, social proof is frequently the last thing added because it requires actual users — but launching without it means the site asks for trust it hasn't established.

Why this severity: Low because missing social proof leaves conversion-stage visitors without trust evidence, but the absence is recoverable once early users can provide it and does not block visitors from understanding the product.