"Submit" is the CTA on approximately 30% of web forms. It describes an engineering action, not a buyer outcome. Nobody wants to submit. They want to get something — an audit, a free trial, a quote. Ogilvy's rule from Confessions was to telegraph the benefit; the CTA button is the last place the benefit should appear, and the place where writers most often forget to. This is the long version of PageLint check FR-2: Primary CTA names action + benefit.
The CXL test that quantified the gap
Peep Laja documented a CTA copy test run by TextMagic, a business messaging company, that produced a +37.6% lift in click-through rate by changing two words. The original CTA read "Order information." The replacement read "Get free information."
The word "get" is a transfer verb — it implies the visitor receives something. "Order" is an initiation verb — it implies the visitor starts a process. In a commercial context, visitors are looking for what they receive, not what they initiate.
The mechanism is precise. "Get" names a buyer action: receiving. "Free" removes a monetary barrier before it becomes an objection. "Information" names the object of the transaction. "Order information" names a system action — ordering — with no benefit stated, no barrier addressed. It describes what the server does when the form submits, not what the user gets after it does.
A 37.6% lift from two words is not a rounding error. It is evidence that the specific language of CTAs has measurable, non-trivial effect on conversion — and that the default behavior (writing what the form does) is reliably worse than writing what the user receives.
Ogilvy's telegraph-the-benefit rule
Ogilvy's prescription in Confessions of an Advertising Man is about headlines, but the instruction applies with equal force to CTAs. The reader scans the headline first. The CTA is often the second or third element they read — especially on mobile, where layout collapses. If the CTA contains the benefit, the sale has been made twice.
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. If you haven't done some selling in your headline, you have wasted eighty cents.
The implication is not only that headlines must sell — it is that every high-scan element must carry the benefit. The CTA is a high-scan element. Visitors who skim a landing page will read: the headline, the subhead, and the button text. That sequence is the compressed pitch. If the button says "Submit," the compressed pitch ends with an engineering instruction. If it says "Get my free audit," the compressed pitch ends with the product's value.
The formula
Action verb (imperative or first person) + outcome or object. Applied examples:
- "Get my free audit" — verb (get) + qualifier (free) + outcome (audit)
- "Start my 14-day trial" — verb (start) + timeframe + object (trial)
- "See pricing" — verb + object (acceptable for secondary CTAs; too vague for primary)
- "Audit my page now" — verb + object + urgency
- "Show me the demo" — verb + object; conversational register signals the product is approachable
The first-person form ("Get my free audit" vs "Get a free audit") consistently outperforms in copy tests. Michael Aagaard at ContentVerve documented this pattern across multiple tests: first-person possessives create a psychological ownership effect before the conversion event occurs.
Nobody wants to submit. They want to get something.
CTAs to never use
Submit. Names what the server does. Not what the user receives.
Click here. Instructs a physical action. Says nothing about the outcome of the click.
Go. Directs movement without a destination.
Continue (without context). Acceptable as a multi-step form progression label, not as a primary CTA on a landing page.
Learn more (as primary CTA). Too vague. Signals indefinite commitment with no return specified. Acceptable as a secondary CTA when a primary CTA is present and visually dominant.
Buy now. Ogilvy noted in Confessions that "buy" is a resistance trigger when deployed prematurely — it puts the reader in the role of spender before they have been sold. "Get," "Start," and "Try" carry lower resistance load.
The common pattern across failed CTAs: they describe what the form does, what the system does, or what the visitor must do — rather than what the visitor receives. CTA copy that converts orients entirely around the buyer's outcome.
Secondary CTAs
The rules above apply strictly to the primary CTA — the button you most want the visitor to click, the one you'd lose conversion on if it underperformed. "Learn more" is legitimate as a secondary action when a dominant primary CTA is already present. "See demo," "Read case study," and "View pricing" are all acceptable secondary labels because the primary CTA has already established the page's core offer.
The failure mode on secondary CTAs is visual competition: making the secondary CTA the same size, weight, and color as the primary. The secondary should always recede — lighter color, smaller font, less prominent position — so the hierarchy is clear. A visitor who can't determine which CTA to click often clicks neither.
How PageLint detects this
LLM check. The engine evaluates whether the primary CTA — identified as the largest or most visually prominent button above the fold — contains two elements: (1) an action verb oriented toward the buyer, and (2) a buyer-oriented outcome, object, or barrier-removal qualifier. "Submit" fails both. "Get my free audit" passes both. "Start free" passes on verb; passes on barrier-removal; passes overall.
Confidence: High for single-CTA pages. Medium when multiple CTAs exist with unclear hierarchy — the engine flags this for human review rather than applying the check against an ambiguous primary.
The check does not evaluate color, size, or placement — those are separate FR checks. FR-2 evaluates language only.
Related principles
- FR-1 — CTA is visually dominant above the fold
- FR-3 — Form fields match minimum required
- DI-1 — Directional clarity toward primary action
Sources cited
Primary:
- CXL Institute. "How to Write a Call to Action That Converts." Peep Laja. CXL.com. (TextMagic +37.6% CTA test, button copy methodology.)
- Ogilvy, David. Confessions of an Advertising Man. Atheneum, 1963. Ch. VI. (Telegraph-the-benefit principle; headline investment rule.)
Secondary:
- Aagaard, Michael. "The Easy Way to Write CTAs That Don't Suck." ContentVerve. (First-person possessive CTA tests.)
- Ogilvy, David. Ogilvy on Advertising. Crown Publishers, 1983. Ch. 7. (Direct response CTA copy principles.)
- PageLint friction checks — FR-1 (CTA visibility), FR-3 (form field count), DI-1 (directional clarity).