Writing a CTA That Converts: Copywriting, Placement, Psychology

In audits, the CTA is almost always the weakest link on the site. Not the architecture, not the design, not the colors. The button. Or more precisely: what's written on it, where it's placed and the context surrounding it.

The client's reflex is to ask the wrong question: "what color should our CTA be?". The real question is elsewhere. A CTA that converts isn't won on a shade of blue, it's won on three cumulative levers: copywriting, placement and the psychology of the moment it appears.

According to a HubSpot (opens in a new tab) study covering 330,000 CTAs observed over six months, personalized CTAs convert 202% better than generic ones. It's the variable with the highest return on effort you can work on for an existing site.

Today we break down what turns an ignored button into measurable action. Copywriting, placement, psychology, then a method to test without noise and the list of mistakes we see on almost every site in audit.

Most sites treat the CTA as a graphic component. A rectangle with text, placed at the end of a section because you have to put one somewhere. This approach is the main reason why so many sites have decent traffic and weak conversions, even when the other building blocks of a website that converts (opens in a new tab) are in place.

A CTA is a decision point. The moment the user sees it, they're silently asking three questions, often without realizing: "What happens if I click?", "Is this the right time to act?" and "What does it cost me?". A CTA that converts answers all three before being clicked.

Wording answers the first question. Placement answers the second. The psychological context around it answers the third. The three together. Optimizing one and neglecting the others produces marginal results. Aligning all three produces measurable conversion jumps.

The text on a CTA is the micro-promise your site makes to the user at the moment of action. A vague micro-promise produces a vague click. A precise micro-promise produces an engaged click.

The first-person verb rule

One of the most documented CTA A/B tests is the one run by Michael Aagaard for Unbounce (opens in a new tab). Changing a button from "Start your free trial" to "Start my free trial" increased click-through by 90%. The only difference: the shift from second to first person.

Why it works: when the user reads "my", they project themselves into the action. The CTA is no longer an instruction the site gives them, it's a decision they take for themselves. The psychology of engagement starts in the pronoun.

Specificity beats generality

"Learn more" is the most common CTA on the web and one of the least effective. It says nothing about what happens after the click. The user doesn't know whether they'll land on a product page, a contact form, a video or a 3,000-word article. Uncertainty creates friction, friction creates inaction.

Compare:

For each CTA, the specific version reduces uncertainty and answers the question "what happens if I click?"

The friction vocabulary to remove

Some words mechanically trigger resistance in the user because they evoke a burden: submit, send (in the context of a long form), sign up, create an account, buy. These words aren't forbidden, but they should be used consciously.

Replacing a verb that evokes effort with a verb that evokes reception changes the mental equation. "Get my quote" shifts attention from what the user gives to what they receive. Same action, phrased from the benefit rather than the constraint.

Value before the click

The best CTAs don't describe the action, they describe the result. "Calculate my annual savings" beats "Use the calculator". "See our healthcare case studies" beats "Portfolio". The CTA becomes a direct answer to the user's implicit question at this point in the journey.

A perfectly written CTA in the wrong place converts less than an average CTA in the right place. Placement is a question of timing: at what point in the journey is the user ready to act?

Above the fold, but not alone

A CTA in the hero section is essential. Research by Nielsen Norman Group (opens in a new tab) confirms that user attention peaks in the visible area above the fold, with about 57% of total time spent there. If your main CTA isn't there, you're showing it only to the half of visitors who scroll.

But the above-the-fold CTA serves already-convinced visitors (direct returns, qualified paid traffic, visitors referred by a trusted source). Others need proof before acting. Hence the rule: a CTA in the hero and a CTA after every block that delivers a proof of value.

After the proof, not before

Placing a CTA right after a section that demonstrates value is more effective than placing it before. After a quantified customer testimonial. After a video demonstration. After a reassuring key figure. The visitor just received a positive signal, their internal objection is reduced, the perceived cost of action drops. That's the moment.

The sticky CTA on mobile

On mobile, a sticky CTA (pinned at the bottom of the screen, always visible) typically increases conversions by 5% to 20% depending on the sector, according to data aggregated by CXL (opens in a new tab). The reason is anatomical: on a phone held in one hand, the thumb zone covers the bottom third of the screen. Anything placed outside demands additional effort.

Condition: the sticky must be discreet (reasonable height, controlled opacity), not cover content and let the user collapse it if they want. An aggressive sticky produces the opposite effect: it annoys and drives people away.

Hierarchy: one primary CTA, one ghost CTA

On the same page, having two CTAs with equal visual weight splits attention and dilutes click-through. The rule: one primary CTA (solid color, high contrast, assertive size) and one or several secondary CTAs called ghost (outlined, neutral background, modest size) for visitors not yet ready for the main action.

Example: primary CTA "Book a 30-min call", ghost CTA "See case studies". The ghost captures visitors in evaluation mode without cannibalizing the primary. The second path feeds the first.

This is the least exploited lever and often the most powerful. A CTA doesn't work in a vacuum: it works in a mental state you can construct around it.

Loss aversion

The work of Daniel Kahneman and Amos Tversky (opens in a new tab) established that a potential loss weighs about twice as much in decision-making as an equivalent gain. Framing a CTA from the loss avoided can outperform framing from the gain, depending on context.

Two approaches:

Test according to your audience. Loss aversion is particularly powerful in B2B where opportunity cost can be rationalized.

Hick's Law: fewer choices, more action

Hick's Law (opens in a new tab) states that the time needed to make a decision grows logarithmically with the number of options. Applied to CTAs: the more you offer, the less each one is chosen.

In audits, we regularly see pages with 5 CTAs of equivalent intensity above the fold: "Sign up", "Request a demo", "Download the guide", "Watch the video", "Contact us". Each cannibalizes the others. The right architecture: one visually dominant primary choice and, at most, one secondary alternative.

Progressive engagement (micro-commitments)

The consistency principle identified by Robert Cialdini (opens in a new tab) shows that a person who accepts a small commitment is far more likely to accept a larger one afterwards. Applied to CTAs: fragmenting the decision reduces friction.

Instead of a single CTA "Buy now" (strong commitment, low conversion rate), offer a journey: "See available sizes", then "Add to cart", then "Proceed to checkout". Each micro-action makes the next one more likely. It's the logic behind multi-step checkouts that outperform single forms despite requiring more clicks.

The CTA in a reassuring context

The immediate context of the CTA influences click-through more than the CTA itself. Reassuring elements nearby:

Optimizing a CTA without testing is intuition. Testing without method is noise. Here's the framework we use on engagements.

One hypothesis per test

An A/B test must isolate a single variable. Changing wording, color and size simultaneously makes it impossible to attribute the result to a specific cause. You'll know version B performs better without knowing why, and you won't be able to generalize the learning.

Formulate the hypothesis before the test: "Switching from 'Learn more' to 'Watch the 2-min demo' will increase click-through because format and duration reduce uncertainty." At the end you'll know whether the hypothesis holds or not.

Minimum sample size

An A/B test on too small a sample produces random results that get mistaken for truths. The rule: wait for at least 300 to 400 conversions per variant and a minimum of two full cycles (two complete weeks to capture weekday/weekend variations). Calculators from Optimizely (opens in a new tab) or VWO (opens in a new tab) give the exact size based on your baseline rate and the expected significance threshold.

Stopping a test as soon as it "seems" to be winning is the most common mistake. The first days are dominated by chance. A test that looks like a winner at day 3 can flip by day 14.

What you actually measure

The CTA click-through rate is not the final metric. A CTA can increase clicks and decrease downstream conversions if the wording over-promises or attracts the wrong visitor profile. Always measure two levels: the click on the CTA and the final conversion (purchase, qualified form, booked call).

The winning CTA isn't the one that generates the most clicks, it's the one that generates the most qualified conversions at the end of the chain.

You can have the best design, the best SEO, the best content and lose the conversion on the last meter because of a poorly thought-out button. Conversely, you can compensate for an imperfect page with a well-written CTA, placed at the right moment and backed by a reassuring context.

CTA optimization has the best effort/impact ratio of any conversion lever. A few hours of work on wording, placement and immediate context can produce double-digit gains in conversion rate, without a redesign, without budget, without technical dependencies.

It's the kind of work we systematically recommend in audits before any redesign. Because a site that converts poorly doesn't always need to be rebuilt.