Conversion copywriting is writing with one job: get a specific visitor to take a specific action, such as buying, signing up or booking a call.
It is judged by an A/B test, not by taste.
In the 230 copy tests in our database, the changes that win most often make the offer clearer.
They name the product, lead with the result the buyer wants and use plain action words.
Clever wording and extra explanation lose more often than people expect.
The job title "conversion copywriter" is usually credited to Joanna Wiebe, founder of Copyhackers.
The Copywriter Club describes Joanna Wiebe as having "invented the title."
The idea behind it is simple. Brand copy tries to make people feel something about a company.
Conversion copy tries to move one reader one step closer to a decision, and it is rewritten until the numbers say it works.
This page draws on that evidence. Every claim below links to a test you can open, with the original numbers and the team that ran it.
For the full set, browse all headline and copy tests or the whole test library.
Copywriting for conversions: what tested copy rewards
Copywriting for conversions is less about finding magic words and more about removing doubt.
A visitor arrives with a question ("Is this for me?
What do I get? What happens if I click?").
Copy that answers those questions fast tends to win.
Copy that makes the visitor work tends to lose.
Before the patterns, one caution. Of the 230 copy tests we hold, 209 are recorded as wins, 12 as losses and 9 as flat.
That is not the real win rate of copy changes.
Agencies and vendors publish their winners and rarely publish their failures, so treat every number here as "this worked once, for this store" and not as a forecast for yours.
1. Say what it is before you say why it is great
The cheapest copy fix is often a missing noun.
When Source Patrol added the single word "supplement" to a Movexa landing page headline that never said what the product was, purchases rose 89.97% at a reported 98% significance (no sample size was given).
CityCliq tested four homepage taglines. The vague "Businesses grow faster online!" sent 25.3% of visitors to pricing.
The literal "Create a webpage for your business" sent 47.8%, a 90% lift.
The advertising-themed "Online advertising that works!" did worse than the control, at 20.2%.
Convertibles saw the same thing on a VR accessories store: a hero headline that simply names the product added $19,441 a month in revenue.
If a stranger cannot tell what you sell from your headline alone, fix that before you test anything else.
Our guide to writing headlines covers this in more depth.
2. Lead with the outcome the buyer wants
Once the product is clear, the next most repeated win is moving the benefit to the front.
Dave Diederen switched a product title from product-first to benefit-first wording and recorded +6.96% conversion rate and +11.67% revenue per visitor.
A related Dave Diederen test using an "Achieve [outcome] with [product]" title format reported +16.92% revenue per visitor.
Kasper Dolk changed a product hero headline to the desired outcome (the example given is "Wake Up Feeling Like You Actually Slept") and reported $823,115.54 in extra revenue per year from that one test.
Oliver Kenyon rewrote a product page from general claims to copy about soreness, fatigue and recovery, with no layout or price changes, and reported +$159,125 a month and +$0.604 revenue per session.
Small versions of this work too. Chêne Gear added one benefit word to each product name, and every variant beat the original, the best by 27.1% conversion rate.
That one comes from vendor marketing with no sample size, so read it as direction, not proof.
For how to turn a feature list into outcome copy, see features vs. benefits.
3. Write for the reader you actually have
Generic copy speaks to everyone and convinces no one in particular.
SplitBase found that BestSelf visitors were not getting past or understanding the top of the Self Journal page.
A headline aimed at "professionals," naming the goal-setting benefit and adding a user count, lifted Self Journal sales 27% across roughly 20,000 visitors per version.
Context from the traffic source matters as well.
Dave Diederen matched the social proof line under the add-to-cart button to the ad angle each visitor clicked, which produced +12.70% conversion rate and +9.74% revenue per visitor (average order value fell 2.63%).
Oliver Kenyon tested a pain-relief angle against a longevity angle for the same dog supplement and reported +28% CTR and +$116k a month.
The lesson is that the "right" message depends on who is reading.
Your best source for it is the words customers already use in reviews, support tickets and surveys, not a brainstorm.
A clear unique selling proposition gives every page the same core claim to adapt.
4. Make buttons and links obvious, not clever
Calls to action are where creativity hurts most.
Dave Diederen tested "Explore the Collection," "Find Your Fit" and "Discover Your Style" on a homepage button against plain "Shop Now."
"Shop Now" won, with +4.99% conversion rate and +8.15% revenue per visitor.
On product pages, Dave Diederen also tested "Try Risk Free Today," "Get Yours Now" and "Start Your Journey" against "Add to Cart," which won every time.
The best button copy tells people what they will get.
Veeam Software changed a link from "Request a quote" to "Request pricing" because surveys showed visitors wanted prices, and click-through rose from 0.54% to 1.40%.
Note that this measured clicks to the inquiry page, not finished leads.
Wording that raises the commitment can backfire. Encyclopaedia Britannica tested "Become a subscriber" and "Become a member" against "Start your free trial" on a subscription popup, and both variants lost, with CTR down as much as 20%.
Going's "Trial for free" button beat "Sign up for free" by 104% in premium trial starts, but that change also sent visitors to a different offer, so it is an offer test as much as a copy test.
More examples live in our CTA copy tests and the guide to microcopy.
5. Cut words, unless the reader still has questions
Research on reading behavior supports shorter copy. In a 1997 Nielsen Norman Group study with 51 users, a concise rewrite of a promotional page scored 58% better on usability and a version that was concise, scannable and objective scored 124% better.
A later analysis by Jakob Nielsen estimated that users read about 20% of the words on an average page.
That is usability research, not sales tests, but the A/B tests point the same way.
Dave Diederen shortened a guarantee from "or get 100% of your money back" to "or your money back" and recorded +13.08% conversion rate and +11.88% revenue per visitor.
Oliver Kenyon removed a single line of copy from a cart page and reported +$40,413 a month.
FreestyleXtreme found that long descriptions placed above product images pushed the images and button down, and the page without them converted 31.38% better at 96% confidence.
But "shorter wins" is not a rule. MECLABS Institute tested a short HealthSpire Medicare page against a longer one that explained the products and introduced the phone agents as advisors, and the long page produced 638% more leads.
M7 Corporation's long-copy white paper email drew about 3x the response of the short version.
The way to reconcile these: length should match how much the visitor still needs to know.
A complex, high-stakes or unfamiliar purchase needs answers.
A visitor who has already decided needs a clear path.
NextAfter added value copy and endorsements before a seminary's donation ask and donations fell 28%, because people on that page had already decided to give.
See landing page copy and product descriptions for page-specific advice.
6. Rewrites can lose, so test them
Copy that looks better to the team can still lose.
Clean Commit rewrote the copy on four product pages around value, features and outcomes, along with layout and trust changes, and conversion rate fell 28.3%.
That test had only 1,765 visitors over 41 days, so the result is weak, but it shows why you test before you ship.
Another Clean Commit test of shorter, benefit-led descriptions came out at +1.4% conversion, not significant: following the playbook does not guarantee a lift.
How to test your own copy
- Start from research, not opinions. Collect the exact phrases customers use about their problem and their doubts. That is where outcome headlines and objection-handling lines come from.
- Write one clear hypothesis. For example: "Visitors don't know this is a supplement, so naming it in the headline will raise purchases." Our guide to an A/B test hypothesis shows the format.
- Change one message at a time when you want to learn. Bundled rewrites can win, but you won't know which line did it.
- Measure the action you care about. Patrick McKenzie tested a search ad offering free word lists that raised CTR from about 1.1% to 5.8%, yet conversion dropped from 22% to 14% and cost per trial rose from about 32 to 43 cents. The ad lost on the metric that mattered. For stores, revenue per visitor catches cases where clicks rise and sales don't.
- Run it long enough. Copy effects are often small, so plan the sample size first and don't stop at the first good day. See how long to run an A/B test.
If you need a starting structure for a first draft, copywriting frameworks such as problem-agitate-solve can help.
Treat the framework as a draft, and let the test decide.
Conversion copywriting examples from real A/B tests
The examples below show the before and after wording where the source gave it.
Each result is as reported by the team that ran the test.
Open any test for the sample size, metric and caveats.
| Where | Before | After | Result | Ran by |
|---|---|---|---|---|
| Homepage tagline | "Businesses grow faster online!" | "Create a webpage for your business" | 25.3% to 47.8% click-through to pricing | CityCliq |
| Landing page headline | Headline without the product type | Same headline plus "supplement" | +89.97% purchases | Source Patrol |
| Cart message | "Free shipping will be applied at checkout" | "Congrats! You unlocked FREE shipping!" | +11.08% conversion rate, +13.93% revenue per visitor | Dave Diederen |
| Delivery labels | "Delivery above $150", "30 Days Return" | "Free Delivery above $150", "Free 30 Days Return" | +12.3% checkouts | Wingify Services (Zalora) |
| Guarantee line | "or get 100% of your money back" | "or your money back" | +13.08% conversion rate | Dave Diederen |
| Homepage button | "Explore the Collection" and other creative labels | "Shop Now" | +8.15% revenue per visitor | Dave Diederen |
| Product page link | "Request a quote" | "Request pricing" | 0.54% to 1.40% click-through | Veeam Software |
| Homepage button | "Get a Demo" | "Free Demo" | +99.42% clicks | Greenlane Search Marketing |
| Lead-gen button | "Join Us" | "Make Money Flipping Websites" | +33.1% CTA clicks | Empire Flippers |
| Subscription popup | "Start your free trial" | "Become a subscriber" / "Become a member" | Both lost, CTR down as much as 20% | Encyclopaedia Britannica |
| Donation page | Straight to the ask | Value proposition and endorsements first | -28% donations | NextAfter |
@DaveDiederenExample: putting the good news first
Two of the examples change almost nothing except word order or tone.
Zalora's perks were already free, but the labels didn't say so up front.
Moving "Free" to the start lifted checkouts 12.3%.
Dave Diederen's cart message offered the same free shipping in both versions, but "Congrats!
You unlocked FREE shipping!" framed it as something the shopper had just earned, and revenue per visitor rose 13.93%.
If your offer has a benefit people care about, check whether your copy states it plainly and early, or hides it behind a condition.
Example: telling people what happens next
"Request pricing" and "Free Demo" both answer the question a hesitant visitor has before clicking.
Veeam's visitors wanted a price, not a sales conversation, and the new label said they would get one.
Greenlane Search Marketing's "Free Demo" on Corcentric's homepage told visitors the demo would cost nothing, and clicks rose 99.42%.
That test ran for two weeks on about 800 visitors and counted clicks, not booked demos, so it needs a repeat before anyone relies on it.
Empire Flippers swapped "Join Us" for "Make Money Flipping Websites," which names the reason to click, and CTA clicks rose from 2.88% to 3.84%, or 33.1%, at 96% confidence.
Example: an outcome headline on a landing page
Vasta replaced an Infinite Age landing page headline built on broad wellness slogans with two shorter headlines, each under 12 words, that led with the supplement's main outcome.
Only the headline changed. Completed purchases rose 35% over 10 days at 95% confidence, though the source does not say which of the two new headlines won.
This is the same pattern as the benefit-first product titles above, applied to a page built for ad traffic.
Visitors from an ad have a few seconds to decide whether the page is about their problem.
A slogan makes them guess. An outcome tells them.
Example: borrowing the customer's words
Laura Roeder Social Media replaced a newsletter headline that asked "Could You Use a Free Plan For Promoting With Social Media?" with a line taken from customer feedback: "Yours is the only newsletter that I actually read." Sign-ups rose 24.31% over 12 days.
The winning headline does two jobs. It answers the visitor's real worry (another newsletter I'll ignore) and it does so in a reader's voice rather than the seller's.
This is why research into customer language is the first step in the testing process above: the best lines are often already sitting in your reviews and emails.
Our page on social proof covers other ways to put customer voices on the page.
Example: when more copy hurt
The Britannica and NextAfter tests are useful because they lost.
"Become a subscriber" asks for a bigger commitment than "Start your free trial," and fewer people clicked.
NextAfter's extra persuasion slowed down visitors who had already decided to give.
Both point the same way: persuasive copy only helps when the reader still needs persuading.
Where to start on your own site
If you only have time for a few copy tests, the database suggests an order.
Start with the main headline on your highest-traffic page, since that is where the "say what it is" and "lead with the outcome" wins come from.
Next, check your main buttons for creative labels that could be replaced with the plain action.
Then look at the small lines near the decision point: shipping, returns, guarantees and pricing labels.
Leave long-form rewrites for last. They take the most work, they bundle many changes together, and two of the clearest losses on this page, from Clean Commit, were full rewrites.
What the examples have in common
- Clarity beats cleverness. The winning versions use plainer words than the ones they replaced.
- The benefit comes first. "Free," the outcome or the product name moves to the front of the line.
- The copy answers the visitor's next question. What is it, what does it cost, what happens when I click.
- Small changes can matter. Many wins come from a word or a line, not a full rewrite. That makes copy one of the cheapest things to test.
- Losses are real. Extra explanation, higher-commitment wording and full rewrites all appear on the losing side.
To judge whether a result like these would be worth chasing on your own site, look at the lift together with the sample size and metric on each test page, not the headline number alone.