Five times as many people read the headline as read the body copy. That figure has circulated on Twitter and in copywriting courses for sixty years, usually stripped of context, usually followed by advice to write more headlines, bigger headlines, headlines everywhere. The original argument is the opposite. David Ogilvy was making a case for concentration — for putting most of your effort into one line, not multiplying lines. This is the long version of PageLint check CL-1: Single dominant H1 in hero. It explains what the source claim actually says, why modern landing pages consistently misread it, and what the check specifically tests.
What Ogilvy actually wrote
Ogilvy states the arithmetic twice in Confessions of an Advertising Man (1963), and then again in "How to Create Advertising That Sells," the Ogilvy & Mather trade ad from 1972. The primary formulation is in Chapter VI, "How to Write Potent Copy":
On the average, five times as many people read the headline as read the body copy. When you have written your headline, you have spent eighty cents out of your dollar.
The corroborating version, a few paragraphs later in the same chapter: "If you haven't done some selling in your headline, you have wasted 80 per cent of your client's money."
Neither passage contains the instruction to write multiple headlines. The instruction is to write one — and to treat it as the primary investment. The 80% figure is a constraint on how much budget (attention, design, copy effort) a headline consumes. Spending it on several competing headlines is, by Ogilvy's own arithmetic, spending the same 80 cents twice.
The 1972 Ogilvy & Mather trade ad makes this explicit: "That is why most Ogilvy & Mather headlines include the brand name and the promise." The goal is efficiency — one line doing the most work. Ogilvy's practice was to write a single dominant headline for each ad, then test variants. He reportedly considered 26 different headlines for the 1958 Rolls-Royce campaign before settling on "At 60 miles an hour the loudest noise in this new Rolls-Royce comes from the electric clock." One headline went to press.
The reading pattern that makes this matter
Ogilvy's evidence base was print split-tests and reader-survey data, not eye-tracking. But the underlying reading pattern he observed maps cleanly to what CXL Institute has measured on digital pages: visitors arrive, scan the dominant text element, make a stay-or-go decision in a few seconds, and only then read further. If two text elements compete for dominance, the eye flickers between them. Fixation splits. Neither element gets the full attention budget.
If you haven't done some selling in your headline, you have wasted 80 per cent of your client's money.
The five-second test — developed at CXL Institute as a formalized version of this reading pattern — asks: can a participant answer "what does this product do?" from the hero alone, after five seconds of exposure? The test consistently fails when the page has two competing H1-level elements. The participant cannot tell which one to anchor on. Their answer reflects neither.
Why this check exists
The misreading of Ogilvy's claim produces a specific failure mode on modern SaaS landing pages.
A founder, having absorbed the "headlines are 80% of the work" idea, over-indexes on headline volume. The hero gets an eyebrow, an H1, a sub-head styled at H2 weight, a second statement in the same size, and a CTA that itself contains a sentence. Each element is individually reasonable. Collectively, they eliminate the visual hierarchy that makes any one of them dominant.
The programmatic check that CL-1 runs counts the <h1> elements in the parsed DOM and passes only when there is exactly one. This is a strict test. The reasoning: the H1 tag is a semantic signal that tells both browsers and screen readers what the primary heading of the page is. If a page has two H1s, it is claiming to have two primary headings. That claim is incoherent — and it's the HTML-level symptom of the reading-hierarchy problem Ogilvy was describing.
Two additional failure modes are common enough to flag separately:
The H1 lost in an image. Some pages render their primary headline as text baked into a hero image. The visual H1 exists; the semantic H1 does not. The parser sees no H1. The audit fails at the programmatic level, but the underlying problem is that the headline is invisible to accessibility tools, search engines, and any context where the image doesn't render.
The logo-as-H1. Some page templates mark up the site logo as an H1 for SEO reasons. The semantic H1 exists; its content is the brand name, not the product description. A visitor scanning the page has already seen the brand name in the nav. The H1 is wasting its position.
What this looks like in practice
The check fails in three recurring patterns. Here are before-and-after pairs for each.
Pattern 1: Two equal-weight statements
Pattern 2: Eyebrow + H1 + styled-H2 at the same size
Pattern 3: The H1 that needs the sub-head to make sense
Blind headlines show up on well-known homepages too. In our ranking of 24 SaaS value proposition examples, Webflow and Asana score lowest on Clarity because their H1s put a metaphor or an abstract noun where the product category should be.
Common mistakes when CL-1 fails
Three competing H1s in hero. The most literal failure: multiple elements with the <h1> tag. Usually caused by a template that marks up section headers as H1s, combined with a hero that also has an H1. Fix: audit the DOM, not the visual design. What looks like a sub-head may be semantically an H1.
H1 hidden in an image. Common on older Webflow and Squarespace templates. The visual design shows a strong headline; the HTML shows no H1 at all, or an H1 with aria-hidden="true". The visual hierarchy is fine; the semantic structure is broken. Fix: render the primary headline as real text, even if CSS is required to match the visual treatment.
Brand-name-only H1 when the brand is unknown. "Acme." as an H1 works if Acme is Salesforce. It does not work if Acme is a 12-month-old SaaS startup. The brand name alone passes the one-H1 test but fails the five-second test that CL-3 runs. The two checks are related: CL-1 verifies structural dominance, CL-3 verifies comprehension. A brand-name-only H1 typically passes CL-1 and fails CL-3.
Multiple eyebrow-H1-sub-head stacks. Some landing pages have two or three "hero" sections stacked vertically — each with its own eyebrow, H1, and sub-head. The first fold passes CL-1. But the overall page has six H1s and no clear primary. CL-1 catches it because it counts every H1 on the page, not just the ones in the first viewport.
Once the page has one dominant H1, test what it says. The free headline analyzer scores a single headline on the five-second test, jargon, specificity and three more checks.
How PageLint detects this
CL-1 is a programmatic (PROG) check. It parses the page DOM and counts every <h1> element on the page, not only those in the hero. H1s with identical text count once, because many sites render the same headline twice for mobile and desktop. Exactly one H1 passes. No H1 is flagged high. Two or more are flagged medium.
Two edge cases to know about:
-
Logo-as-H1: a short brand name in the H1 passes CL-1, because the check counts tags and does not read them. CL-3 then tests whether the hero still says what the product is and who it is for.
-
H1 inside a hidden element: the parser reads HTML, not CSS, so an H1 inside a
display:nonecontainer still counts. A page that hides an H1 for SEO while showing a different visual headline can pass CL-1, or fail it as a multiple-H1 page.
Confidence for this check is High because the test is binary and structural. DOM parsing either finds one H1 or it doesn't. The blind spot is CSS: the check sees every H1 in the HTML, including ones a visitor never sees.
Related checks
- CL-2: Heading hierarchy is logical — checks that H2s and H3s nest under the H1 without skipped levels.
- CL-3: Hero passes the 5-second test (what + who) — runs the CXL five-second test on the full hero block.
- CL-4: Headline specificity — once CL-3 passes, tests whether the headline makes a concrete claim instead of one that fits any product.
Related principles
- CL-2 — Heading hierarchy is logical
- CL-3 — Hero passes the 5-second test (what + who)
- CL-5 — Jargon and insider terminology
Sources cited
Primary:
- Ogilvy, David. Confessions of an Advertising Man. Atheneum, 1963. Ch. VI: "How to Write Potent Copy," p. 106.
- Ogilvy, David. Ogilvy on Advertising. Crown, 1983. Ch. 7: "Wanted: a renaissance in print advertising."
- Ogilvy & Mather. "How to Create Advertising That Sells." Trade advertisement, 1972.
Secondary and corroborating:
- Roman, Kenneth. The King of Madison Avenue: David Ogilvy and the Making of Modern Advertising. Palgrave Macmillan, 2009. (Biographical corroboration of Ogilvy's headline-testing practice.)
- CXL Institute. Five-Second Test methodology. Referenced via Schottmuller's four-question framework; PageLint uses CXL's operationalization of Ogilvy's reading-pattern claim.
Further reading:
- Hopkins, Claude. Scientific Advertising. Lord & Thomas, 1923. Ch. 6: "Being Specific." (Required reading per Ogilvy; foundational to the specificity checks in CL-4 and CL-5.)
- The Clarity lens: 8 checks for landing pages that convert — situates CL-1 within the full Clarity methodology.
- PageLint principles — complete framework reference.