CTA Button Copy: What Your Call to Action Should Say, and 7 Mistakes to Fix

You've got visitors. They scroll. Some of them even read. But almost nobody clicks the button.

Often the problem isn't the color, the size, or where the button sits. It's the two to five words on it. Those words are a promise about what happens next, and if the promise is vague, risky, or doesn't match what the visitor came for, they won't click.

This guide covers what a good call-to-action (CTA) button should say, the seven mistakes we see most often on founder-built pages, a simple formula for rewriting yours, and how to check that it's working.

Short answer

Good CTA button copy tells a visitor what they'll get or what will happen when they click, in a few concrete words, and the next screen delivers exactly that. Replace generic labels like "Submit", "Learn more", and "Get started" with a specific verb plus a specific outcome ("Start my free trial", "See pricing", "Get my free audit"). Then add one short line next to the button that answers the biggest worry at that moment, such as whether a credit card is needed. Keep one main action per page, and test wording changes on your own traffic rather than copying someone else's winning button.

Why button copy matters more than it looks

A button is a link with a promise attached. Nielsen Norman Group (NN/g), the usability research firm, puts it plainly in its article on vague calls to action: a call to action "is not a hypnotic spell". The value of the action has to be obvious to the user, and labels need high "information scent", meaning the words have to clearly signal what's on the other side.

In NN/g's microcopy guidelines, they note that calls to action often work by reducing uncertainty, giving users an easy next step. That's the job: make the next step feel clear, worthwhile, and safe.

And the words can matter a lot. Conversion consultant Michael Aagaard has published several button-copy A/B tests. In one he ran with Unbounce on a pay-per-click landing page, changing "your" to "my" in a free-trial button produced a 90% increase in click-through rate. In another, on a subscription payment page, changing "Create My Account" to a version with "Get Started" lifted sales by 31.03%. Those are single tests on specific sites, and as you'll see below, they even point in opposite directions. The lesson isn't "use this exact word". It's that a few words on a button can be worth testing seriously.

The 7 most common CTA copy mistakes

1. Generic labels that say nothing

"Submit", "Click here", "Learn more", "Continue". These describe the mechanics of clicking, not the reason to click. A visitor scanning your page sees the button before they read your paragraph, so the button has to make sense on its own.

Fix: Name the outcome. "Submit" on a newsletter form becomes "Send me the weekly tips". "Learn more" under a feature becomes "See how invoicing works".

2. "Get started" that leads somewhere unexpected

"Get started" is the default button on a huge number of startup sites. NN/g studied it directly in "Get Started" Stops Users. Their finding: the generic label attracts clicks, but users who expected information about the company were pushed into surveys, onboarding flows, and sales funnels instead. People lost trust and got annoyed, and some gave up or went to a competitor. NN/g's conclusion is that "a link is a promise", and a vague one may hit a shallow click goal without producing happy customers.

Fix: Say what "starting" actually means. If it opens a sign-up form, say "Create my free account". If it starts a quiz, say "Take the 2-minute quiz". If it books a call, say "Book a 20-minute demo".

3. Asking for too much, too soon

A button that says "Buy now" or "Talk to sales" in the hero section asks a first-time visitor for a big commitment before they understand what you sell. Colder visitors need a smaller first step.

Fix: Match the size of the ask to how much the visitor knows at that point on the page. Near the top, a low-risk action ("Try it free", "See a sample report") usually fits better. Lower down, after you've explained the product and shown proof, a bigger ask can make sense.

4. Hiding the cost, the catch, or the next step

Visitors hesitate right before they click because they're wondering about something: Will I be charged? Do I need a card? Will a salesperson call me? How long does this take? If the page doesn't answer, many people simply don't click.

Fix: Put one short, honest line of reassurance directly next to or under the button. Examples (illustrations, not a template to copy blindly):

• "No credit card required. Cancel anytime." (only if true)
• "Takes about 2 minutes. No sign-up."
• "30-minute call with a founder. No sales pitch."

Make sure it's true. A reassurance line the next screen contradicts costs you more trust than the click was worth.

5. Too many buttons fighting each other

"Start free trial", "Book a demo", "Watch video", "Read the docs", "Join our Discord", all in the hero. NN/g notes in the same information-scent article that multiple competing calls to action diminish each other's salience: each one makes the others harder to notice.

Fix: Pick one primary action for the page and style it as the main button. Keep one secondary option at most (often "See pricing" or "See how it works"), styled more quietly. Repeat the primary button further down the page, after the sections where a visitor might be ready to act, such as after your proof or pricing.

6. Button copy that doesn't match the headline or the ad

If your headline promises "a free website audit in 60 seconds" and the button says "Sign up", the visitor has to guess whether those are the same thing. If an ad promised a free trial and the landing page button says "Request a demo", it can look like a bait and switch.

Fix: Read your headline and then your button out loud, one after the other. The button should sound like the natural next sentence. "Find out why your landing page isn't converting" goes with "Audit my page free", not "Get started". If the headline itself is the weak link, start with these landing page headline mistakes.

7. Copying a "winning" button without testing it

It's tempting to read a case study, see that "my" beat "your", and change every button on your site. But look at how mixed the evidence is, even from one consultant. Michael Aagaard reported in an interview with MarketingExperiments that he had repeatedly seen lifts from switching "your" to "my" in button copy, while also noting that similar tests in Danish didn't seem to work. On a different site's payment page, he wrote that in his tests "your" had consistently outperformed "my", and he removed "My" entirely.

Fix: Treat published tests as ideas, not rules. What works depends on your audience, your offer, and where the button sits in the funnel. Change one thing at a time and measure it on your own traffic.

A simple formula for CTA button copy

Most strong buttons follow the same shape:

Action verb + what they get + (optional) a qualifier that removes doubt

• Verb: start, get, see, book, download, try, create, audit, compare
• What they get: the trial, the report, the template, the demo, the pricing, the quote
• Qualifier: free, in 2 minutes, for my team, no card

A few made-up before-and-after examples to show the pattern:

Page	Before	After
SaaS homepage	Get started	Start my 14-day free trial
Newsletter form	Submit	Send me the weekly tips
Agency services page	Contact us	Get a free quote in 24 hours
Template shop	Buy now	Get the Notion template ($19)
Feature section	Learn more	See how scheduling works
B2B product	Request info	Book a 20-minute demo

Then add the reassurance line from mistake 4 underneath, if there's a real worry to answer.

First person or second person?

"Start my free trial" or "Start your free trial"? As shown above, published tests disagree. A reasonable default is to write the button as the visitor's own words ("Get my free report") for low-risk, free actions, and to stay consistent with the rest of your page's voice. If you have enough traffic, it's an easy A/B test.

A 5-minute CTA check for your own page

1. The squint test. Look at your page for three seconds. Can you tell which button is the main one? If two or three look equally important, fix mistake 5.
2. The isolation test. Read only the button text, with no surrounding copy. Would a stranger know what happens when they click?
3. The click test. Click your own button in a private window. Does the next screen deliver exactly what the label promised, with no surprise forms or fees?
4. The worry test. Write down the one question a hesitant visitor would have at that button. Is it answered within a line or two of the button?
5. The match test. Read your headline, then your button. Do they promise the same thing?

If a button fails two or more of these, rewrite it before you touch colors or layout.

How to know if your new CTA is working

Measure the click-through rate on the button itself (clicks divided by visitors who saw it) and the conversion that follows (sign-ups, purchases, bookings). A button rewrite that gets more clicks but fewer completed sign-ups may be making a promise the next step doesn't keep, which is exactly NN/g's warning about "Get started".

If you don't have enough traffic for a proper A/B test, change one button at a time, give it a fair stretch of time, and compare against the period before. It's not rigorous, but it's better than changing five things at once and guessing which one mattered.

Get a second opinion on your CTA

It's hard to judge your own button, because you already know what happens when you click it. Page Sheriff, an AI landing page conversion audit by BizSaloon, lets you paste any public page URL and get a free preview with no sign-up: a conversion score out of 100, six sub-scores (Clarity of offer, Headline, Call to action, Trust & social proof, Objection handling, SEO basics), and the top fixes to start with. The Call to action sub-score lines up with mistakes 1 to 3 and 5 above, Objection handling with mistake 4, and Headline and Clarity of offer with mistake 6.

If you want more depth, there are optional one-time PDF reports (BASIC $4.99, ADVANCED $7.99, EXPERT $9.99 USD). No subscription.

Rewrite your main button, re-run the free Page Sheriff audit, and see whether the Call to action sub-score moves. Then let real visitors decide.

FAQ

What is the best CTA button text?

There isn't one universal best. The best text names a specific outcome the visitor wants ("Start my free trial", "Get my free audit") and matches what actually happens after the click. Generic labels like "Submit" and "Learn more" are usually the weakest choice.

Should my CTA button say "Get started"?

Only if it's very clear from the surrounding copy what "starting" means. NN/g's research found the label can mislead users and push them into flows they didn't expect. A specific label is usually safer.

How long should CTA button copy be?

Short enough to read at a glance, usually two to five words. Put extra detail, like "No credit card required", in a small line under the button rather than on the button itself.

How many CTAs should a landing page have?

One primary action, repeated where it makes sense down the page, plus at most one quieter secondary option. Competing buttons make each one less noticeable.

Does button color matter more than button text?

The button needs to stand out from the rest of the page, but no color fixes a label that doesn't tell people what they'll get. Fix the words first.