Call to Action Examples That Actually Convert

The call to action examples worth copying all arrive with a number attached: the words, the place they sit, and what that combination converts at. Strip the number out and you are left with a list of pleasant phrases and no idea whether yours is doing anything.

A working one does four jobs at once. It names what the reader gets. It says who gets it. It sits where their attention already is. The click lands somewhere that keeps the promise the button just made.

Only the last one stays invisible until you count.

Placement alone moves the same four words from a 13.6% conversion rate to 0.9%. That spread sits inside a single study: First Page Sage's CTA conversion rate report, a meta-analysis of 71 client accounts, last updated 17 April 2025 (the report).

The offer didn't change. The copy didn't change. The result changed by a factor of fifteen.

So the unit worth testing is never the button by itself. It is the pair: the words, and the place they land.

Call to Action Examples by Goal, With the Numbers

Pick the goal before the wording. A button that hasn't decided what it is asking for can't be measured, which is how a page ends up carrying "Learn more" for three years while nobody notices that nobody clicks it.

Here are lines that fit each goal, the placement each rate was measured in, and the number to beat.

Worth knowing how those numbers were built, because it changes how you use them. The agency aggregated conversion data from 71 client accounts and segmented it by CTA style and on-page position, so each rate is an average for that style in that spot rather than a ceiling. Beat your row and the problem is somewhere else on the page.

Take a page selling a $40-a-month tool. The team wants trial sign-ups, newsletter subscribers and demo calls off the same page. That is three goals, which in practice means the page has none.

1. Keep the goal that pays: the trial.
2. Move the newsletter offer to the footer as a line of text.
3. Move the demo call to the pricing page, where the reader has already asked that question.

Now the table is useful, because you are measuring one row instead of guessing at three.

Do the same on your own highest-traffic page in the next thirty minutes:

1. Write the goal in five words.
2. Find that goal's row above and write the rate down. It is your benchmark for the next seven days.
3. Note the benchmark and today's date beside the page. That is the first fact about your CTA that isn't an opinion.
4. Demote or delete every other button competing with it.

By the end of the week you will know whether your page sits above or below its benchmark. Most owners can't say, which is why they keep rewriting the copy and calling it optimization.

The failure looks like three buttons of equal size. Each one gets a third of the attention, and none of them gets a decision.

The Four Words That Do the Work

Most weak buttons fail the same short checklist. There is a verb the reader performs, an object they receive, a person the sentence belongs to, and a number or a deadline attached.

One line. That is the whole rule.

Kissmetrics' collection of CTA button tests puts first-person framing, "Start my trial" over "Start your trial", roughly 90% ahead on click-through, and reports single-word swaps moving conversion between 10% and 30% (the tests). The collection doesn't date its entries, so treat the direction as the lesson and measure the size on your own page.

HubSpot's own CTA experiment found personalized buttons converting 202% better than generic ones (their research). Both figures come from vendors selling CTA tools. That doesn't make them wrong. It makes them prompts for your own test rather than promises.

Word count matters less than people think. High-converting buttons average about three and a half words, which is a consequence of being specific rather than a rule about being brief.

Watch a bad one turn into a good one:

1. The button currently says "Submit".
2. Name the goal in five words: get this reader to book a demo.
3. Move it to first person and lead with the verb: "Book my demo".
4. Attach the number: "Book my 15-minute demo".

Four words, one object, one duration. The reader knows what happens in the next fifteen minutes. "Submit" only tells them what they are doing to your database.

The usual miss is a button holding two ideas at once, normally an action plus a reassurance: "Sign up and stay in the loop". Slide the reassurance beside the button instead. It is the same discipline as writing a tagline for a small business: one promise, few words, no hedging.

The second miss is internal language. "Request a callback" is how your sales team describes the work. "Get a call today" is how the reader describes what they want. Same action, one of them in your house dialect.

Call to Action Examples Worth Copying, and the Rung They Sit On

The famous lines are worth reading as a set, because they are all doing one thing in different clothes. These are observed on brand pages catalogued by HubSpot and Wix. Neither collection publishes a tested result for any of them, which is the half of the story those lists leave out.

HubSpot's own "Get the CTAs" names the object rather than the action, so the reader knows what is coming. Evernote's "Get Evernote free" puts the brand and the price in the same breath. Netflix's "Get started" sits under "Cancel anytime", which answers the specific fear that stalls the click. SoundCloud's "Try it free for 30 days" states how long the commitment runs.

Orangetheory's "Claim your first month free" puts the offer on the button instead of hiding it in the paragraph above. Kolkata Chai Co.'s "RSVP for 3/7" puts the date on the button, which saves the reply that says they will try to make it. Chipotle's "Get my guac" is first person, possessive and tied to one day. Wool and the Gang's "Choose a kit" works because choosing is genuinely the first step.

What separates them is readiness. A cold reader needs "See how it works". A warm one will take "Get the checklist". A hot one is fine with "Start my trial". Show a hot button to a cold reader and you get a bounce.

That single mismatch kills more buttons than any wording problem. The words didn't change. The reader did.

Audit one page with these four checks:

1. Read the button out loud. If it doesn't say what arrives, it fails.
2. Decide the rung of the traffic reaching that page: cold, warm or hot.
3. Compare the two. A hot button on cold traffic is a bounce waiting to happen.
4. Write the fix, which is either a colder verb or a warmer offer. Rarely both at once.

Then sort your own lines this week:

1. Write your three best button lines on one page, whatever is live now.
2. Label each one cold, warm or hot, from the reader's point of view rather than yours.
3. Put each line on a page whose reader is at the same rung, and move the hot ones off the cold pages.

The Same Words, Fifteen Times the Result

Now the placement half, from that same April 2025 report.

Two patterns sit in that table. Interruption wins, and the bottom of the page loses. A download offer in a full page popup converts at 13.6%, and the same offer in a bottom banner converts at 0.9%.

Scale that once to see why it outweighs the copy. A page with 4,000 visitors a month, a download offer and a bottom banner converts 36 times. Move the same offer into a full page popup and the same traffic converts 544 times. You moved one element and never touched a word.

The catch is the goodwill the popup spends. It works because it interrupts a reader mid-sentence, and a reader interrupted often enough leaves and doesn't come back. The bottom banner is polite and cheap and almost nobody clicks it. That is the trade, and you make it on purpose.

Placement decides. Copy comes second.

Placement is also why we built the thing this blog belongs to. Every listing on The Board is a paid placement ranked strictly by what has been paid, nothing else, and rank decides how many impressions a listing receives. Any position starts at a dollar, and moving up buys the one thing no amount of button copy can buy: more people standing in front of it.

The two numbers worth understanding on any placement, paid or earned, are impressions and clicks, and the difference between them matters before you judge a button. It also helps to know what a single visitor costs once you are paying for the traffic that reaches it.

The other half of the pair is where the click lands. Promise mismatch is the most expensive bug in conversion and the cheapest to repair, because it needs no test to find.

1. Write the button's promise as a full sentence.
2. Read the first screen of the destination page and check that it says that sentence back.
3. Change whichever one is easier. That is usually the headline on the destination page.

When an ad says "first month free" and the page says "plans from $29", the click is already spent. The order to repair things in when your traffic is thin is in our piece on conversion rate optimization without much traffic, and the short version is this: fix the promise before you touch the design.

The Test That Gives You a Verdict in Seven Days

Most CTA tests end in a shrug because they were too small to answer anything. The arithmetic is worth doing before you write the second version.

At 400 visitors per variant and a 4% click rate, the 95% confidence interval is about plus or minus 1.9 points. Your honest result is somewhere between 2.1% and 5.9%, which is not a number you can build a decision on.

$$ 1.96 \times \sqrt{\frac{0.04 \times 0.96}{400}} = 0.019 $$

To call a one-point win, from 4% to 5%, you need roughly 3,000 visitors per variant. That figure decides whether a test is even worth starting on your traffic, and it is the number the example lists never give you.

It also explains the attrition. About one in eight tests produces a statistically significant improvement, according to the landing page statistics collected by involve.me. Run one test, get one answer, and expect most of them to come back with no difference at all.

Then run it this way:

1. Freeze the destination page. Nothing on it changes for the duration of the test.
2. Write two buttons that differ in exactly one thing. One word, one placement, or one rung.
3. Set the threshold before you look: "B wins if it leads by more than two points after 400 visitors each."
4. Split the traffic evenly and keep it even. A 70/30 split doubles the time to an answer.
5. Stop at 400 visitors per variant or seven days, whichever comes later.
6. Read one metric: clicks on the button.
7. Decide once. Keep it, kill it, or extend by one week. Then stop looking.

Worked example, using a page that gets 900 visitors in a week. Split evenly, that is 450 each. Variant A: 18 clicks, which is 4.0%. Variant B: 27 clicks, which is 6.0%. A two-point lead looks like a winner.

It isn't one yet. A's interval is 4.0% plus or minus 1.8 points. B's is 6.0% plus or minus 2.2 points. They overlap, so the honest call is to extend by a week rather than ship it.

Run the same split for another 900 visitors. If B still leads by two points, ship it, and put the next test on the destination headline. If the gap collapses to under a point, you have learned something more useful than a lift: your button copy was never the problem.

One more rule, and it is the one people break. Decide the stopping rule before you start and leave the dashboard alone until then. A test read every morning and stopped when it looks good is a coin flip with extra steps.

The Lines That Lose Money

These are the ones worth deleting before lunch. None of them needs a designer.

1. The button with no goal. "Learn more" as the only button on a page asks the reader to decide what they want, and they won't.
2. Two buttons arguing. A "Get in touch" beside your main offer lends your best position to your weakest ask. Contact CTAs on a mid-page banner convert at 0.9%, so demote that one to a line of text and give the button back to the offer.
3. Urgency that isn't true. A deadline that slides teaches readers that your deadlines mean nothing. Publish the date, put it on the button, and honour it even when it costs you a sale.
4. The offer hidden above the button. If the discount is a paragraph up, plenty of readers never connect it to the click. Put the offer in the line itself.
5. The button that asks for too much. A cold feed seeing "Order now" on a $999 product has no step to take yet. Give the first step only, and let the page do the second one.

Each of those costs money quietly, which is the only way conversion problems ever behave.

What to do with the page in front of you:

• If it gets under 400 visitors a week, don't split-test it. Pick the line whose goal and rung are right, change it, and watch the rate for a month.
• If it gets traffic and no clicks, the button is the problem.
• If it gets clicks and no conversions, the destination is the problem.

If your button lives in an email, the same rules hold and the placement question moves to where the button sits in the layout. The basics for small business mail cover where it goes and how many of them one send should carry.

Your CTA Sheet, and the One Test to Start Today

Copy this into a note and fill it in for one page. That is the whole exercise, and it takes twenty minutes with no tools.

• Goal, in five words:
• Benchmark rate, from the goal table:
• Placement, and why that one:
• Button line, verb first, first person, number attached:
• Destination headline, saying the same thing back:
• Test threshold, decided before the test:
• The date I read the result:

Filled in, it looks like this:

Goal: start a 14-day trial. Benchmark: 1.8%, slide-in. Placement: slide-in on the pricing page, where the reader has already chosen the category. Button: Start my 14-day trial. Destination headline: Start your 14-day trial, no card needed. Threshold: B wins by more than two points after 400 visitors each. Read the result on 4 November.

Then:

1. Fill in the sheet for your highest-traffic page.
2. Rewrite the button to verb, object, first person, number.
3. Move it to the placement your benchmark came from.
4. Set the threshold in writing and start the split.
5. Write down the date you will read the result, and don't look before it.

The first win is small and real. A week from now you will know a number about your page instead of a feeling about it, and the week after that you can change something else.

More on the rest of the funnel sits in Conversion.

Every button is a promise with a place attached, and the ones that pay for themselves are the ones where the reader, the words and the landing page all agree.