Call To Action (CTA): what it is and how to write one that works

A Call to Action, abbreviated CTA, is the explicit invitation to do something. In its most visible form, it's a button with text: "Buy now", "Book a demo", "Download guide". In a less visible form, it's a link within a paragraph, a direct question at the end of an email, a QR code in a print ad. But the principle is the same: the CTA is the moment where the content—which until then has informed, persuaded, entertained—asks the recipient for a specific action.

Without a CTA, a page, an email, an ad, or a video is a conversation without a close. The audience consumes the content and leaves. With a well-designed CTA, a portion of that audience takes the next step. The difference between the two modes—a page with a thought-out CTA vs. one without—usually ranges between 0% and 5% conversion. On relevant traffic, that translates into real customers.

This article goes through what a CTA is, where it comes from as a concept, what operational rules work in 2026 (with data), and the mistakes still being made after three decades of digital marketing.

The origin: direct response advertising

The CTA as a formal concept wasn't born with the internet. It was born with direct response advertising, which developed in the U.S. from the 1920s-30s and consolidated mid-century.

John Caples, a copywriter at the Ruthrauff & Ryan agency and later BBDO, published "Tested Advertising Methods" in 1932, a book that remains essential reading. Caples was obsessive about measuring responses: each ad included a coupon to clip, a specific zip code, or a different phone number, so that response could be attributed directly to the ad. One of his central maxims: the ad must explicitly ask for the action. Without asking, you don't get it.

David Ogilvy—founder of Ogilvy & Mather—reinforced the principle in "Confessions of an Advertising Man" (1963) and "Ogilvy on Advertising" (1983). His own work on brands like Rolls-Royce, Schweppes, and American Express included explicit CTAs where applicable.

Catalogs and direct mail developed the written discipline of the CTA: each catalog page ended with an order code, each direct mail letter with an underlined PS asking for a response. Companies that depended on direct response—Sears, Reader's Digest, Time-Life, American Express—refined CTA copywriting over decades.

The digital era: buttons you ask for clicks on

The online era of the CTA began almost with the web. The first advertising banner in history was published on HotWired on October 27, 1994, an AT&T ad for its "You Will" campaign. The banner text was a provocative question with an implicit CTA: "Have you ever clicked your mouse right HERE? You will." Initial click-through rate: around 44%—a figure that looks like science fiction today. The first banner was a novelty in itself; that CTR dropped quickly with saturation.

In the late '90s and early 2000s, pioneering e-commerce sites (Amazon, eBay) iterated on "Buy Now," "Add to Cart," and "Place Order" buttons. The culture of A/B testing—brought to the mainstream by Optimizely (2010), VWO (2011), Google Optimize (2012-2023, discontinued)—turned CTA optimization into a rigorous discipline.

Famous studies circulated. HubSpot published around 2012 a test between a red and a green button where red won by more than 20% (a result that has since been contextualized: it depends on the contrast with the rest of the page). Microsoft Bing reported that a change in the blue color of its links produced more than 80 million dollars in additional annual revenue. Each specific test produced different results, and the soundest methodological conclusion was: testing is better than guessing.

The five decisions that define a CTA

Beyond theory, writing a CTA means deciding five things:

1. The verb

The CTA's verb is the main decision. And it's almost always chosen poorly.

Generic verbs that are seen a lot and perform little:

• Send.
• Accept.
• Click here.
• Submit.

Concrete verbs that orient the user about what's going to happen:

• Book (a demo, call, visit).
• Download (guide, template, example).
• Request (a quote, proposal, assessment).
• Try (free, for 14 days).
• Start.
• See (case studies, demo, example).
• Talk (to an expert, to sales).

The operational rule: the verb should describe what the user is going to get or what they're going to do, not what the system does. "Send" describes what the form does; "Book an assessment" describes what the user gets.

2. Specificity

A specific CTA converts better than a generic one. "Book a 30-minute assessment" is more concrete than "Contact". "Download the Excel template" is more concrete than "Download".

Specificity reduces cognitive friction: the user knows exactly what they'll receive, how much time they'll invest, what commitment they're taking on. Without that information, they hesitate. And hesitating means not clicking.

3. Friction

Every CTA asks something of the user. More friction = less conversion. Types of friction:

• Time: "Book 30 minutes" is more friction than "Download PDF" (5 seconds).
• Data: "Fill out a 12-field form" is more friction than "Fill out 3 fields."
• Commitment: "Request a quote" can trigger fear of "now they'll keep calling me a lot."
• Identity: "Create an account" is more friction than "Continue as a guest."

The operational rule: less-friction CTAs at the start of the journey, greater commitment as the user advances. A SaaS home page should rarely ask to "register a credit card"; it should ask to "see a demo" or "download a guide."

4. Visual hierarchy

A page has CTAs of varying importance. Visual hierarchy:

• Primary CTA: the page's main action. Large button, standout color, prominent position.
• Secondary CTA: an alternative for those not ready for the main action. Smaller button, "ghost" style (border without fill), or text link.
• Tertiary CTA: links within the body text.

Mixing hierarchies—three red buttons competing for attention—produces decision paralysis. The user, with no clue what's the priority, chooses nothing.

5. Placement

Classic research from the Nielsen Norman Group and more recent eye-tracking suggests that placement matters, but less than popular wisdom holds. The "above the fold" rule (CTA before the first scroll) is valid but not absolute:

• Short direct-sales pages: CTA at the top and repeated.
• Long sales pages: CTA at the start (for those already convinced) and at the end (for those who need to read everything).
• Blog articles: CTA within the text where it helps, at the end as the primary, in the sidebar as a secondary.
• Campaign landing page: CTA always visible (sticky on mobile), multiple conversion points.

Color, size, and design

The visual details that most affect a CTA's CTR:

Contrast: the CTA must stand out visually from the background. A green button on a green-background page can't be seen. The operational rule: the primary CTA should be the most contrasted element on the page.

Color: there's no universally optimal color. It depends on the environment. Contradictory studies on red vs. green vs. orange show that what matters is contrast with the rest of the design, not the color itself.

Size: a minimum of 44×44 px (Apple Human Interface Guidelines) or 48×48 dp (Material Design) for tap targets on mobile. Smaller penalizes accessibility and mobile conversion.

White space around it: a button cramped between other elements loses force. Breathing room = greater visibility.

Shape: rectangular with slightly rounded corners is the most universal. Purely round buttons or ones with strange shapes may not read as buttons.

Microanimation: hover on desktop, tactile feedback on mobile. Reinforces the perception of interactivity.

Microcopy around the button

The text near the button—not the button text itself—often decides the click more than the button does. Patterns that work:

Reduce uncertainty: "No credit card required", "14 days free, cancel anytime", "You'll get a reply within 24h". Removes reasons not to click.

Nearby social proof: "Join 5,000 teams already using Polimake" next to the CTA. Reduces the feeling of being a pioneer or a fool.

Urgency or scarcity (honestly): "Limited spots for May". Works when it's real; manipulative when it's false.

Guarantees: "30-day money-back guarantee", "100% free."

Mistakes still being made

Too many CTAs competing. Five buttons of equal weight = none wins.

A CTA without context. A "Buy" button before explaining what's being bought. It confuses, it doesn't convert.

Generic verb. "Submit" / "Send" where it could say "Book an assessment."

Not testing variants. Publishing the first CTA that came to mind without A/B testing even though the tool is available.

Ignoring mobile. A 30×20 px button that's easy to click on desktop but on mobile demands impossible precision.

CTA in purely brand colors. "The button must be corporate blue." Result: invisible if the page is also mostly blue.

Missing microcopy. Just the button, with no context about what happens next.

A CTA poorly aligned with the funnel stage. "Buy now" on an educational top-of-funnel page where the person doesn't even know they have a problem yet.

Not accessible. A button without a label for screen readers, without visible focus for keyboard, without sufficient contrast. WCAG 2.2 AA is frequently ignored.

Excessive use of intrusive modals and popups. The popup that appears on load and blocks content gets closed immediately and, on mobile, is a factor in the Mobile Friendly Update. It penalizes UX and SEO simultaneously.

Not measuring conversions per CTA. Having three CTAs on a page and not knowing which one converts is a waste of data.

Betting everything on a single primary CTA with no alternatives. Some users aren't ready to "Buy." A "See demo" or "Download guide" option captures those who would otherwise be left out.

Accessibility of CTAs

An accessible CTA:

• Descriptive text, not "Click here" (bad for screen readers out of context).
• Minimum contrast of 4.5:1 with the background (WCAG AA).
• Tap target of at least 44×44 px (iOS) or 48×48 dp (Android).
• Visible focus when navigating by keyboard.
• Clear state: hover, active, disabled—each with visual feedback.
• Correct label for assistive technologies (aria-label when the visible text isn't enough).

How to fit CTAs into creative operations

CTAs aren't a last-minute decision. They're part of the strategic message each piece needs.

Creative operations hold the coherence between message, page, and CTA. At Polimake, Studio defines conversion paths by audience and stage, CTA microcopy, visual hierarchy; Studio coordinates the A/B tests and learns what works; Media executes visual variants (backing video, image, motion).

This relates to persuasive communication, which frames how you arrive at the CTA, to above the fold, which defines where it appears, to the conversion funnel, which decides which CTA is appropriate at each stage, and to base communication, which holds the coherence of tone.

To wrap up

A CTA is the small sentence where it's decided whether all the content work translates into a result or not. Well written, with a concrete verb, appropriate friction, respected visual hierarchy, and supporting microcopy—it converts. Poorly written, with a generic verb, no context, competing with five others—it wastes the traffic that was so hard to get.

The practice that ages best: treating each CTA as a strategic decision with criteria (not as a design detail), testing variants with discipline, measuring conversion per point, and keeping coherence between the piece's message and the button's promise.

Quick reference

• Concrete verb, not generic. "Book an assessment" > "Send."
• Specific: tell the user what they get and how much time they invest.
• Friction appropriate to the stage: little at the start, more once there's trust.
• Clear visual hierarchy: one primary, one secondary, tertiaries as links.
• Strong contrast between button and background.
• Tap target ≥44×44 px (Apple) / 48×48 dp (Android).
• Supporting microcopy: social proof, guarantee, "no card."
• A/B testing when possible.
• WCAG 2.2 AA accessibility: contrast, focus, labels, size.
• Multiple CTAs aligned to the funnel: primary and an alternative for those not ready.
• Measure conversion per CTA, not aggregated.