Key takeaways
- In December 2007 the Obama campaign tested 24 versions of its sign-up page: 4 button labels against 6 images and videos.
- The winner paired a "Learn More" button with a family photo. Sign-ups rose from 8.26% to 11.6%, a 40.6% lift.
- The campaign estimated that lift was worth about 2.9 million extra email addresses, 288,000 extra volunteers and $60 million in donations.
- Every video lost to the images, including the one the team liked best.
- The lesson: a first button that asks for less gets far more clicks. Ask for curiosity before commitment, then test it.
The highest-earning change on the Obama campaign's website was not a new feature. It was swapping "Sign Up" for "Learn More" and a video for a photo. A recent reel retold the story, and it still holds up as the clearest example of why the words on one button matter. Conceptdigital builds websites and booking funnels for businesses, and the first thing we look at on any homepage is what its main button asks the visitor to do.
Key definition: An A/B test shows different versions of a page to different visitors at the same time and measures which one gets more people to act. A multivariate test does the same with several elements at once, here the button and the image.
What did the Obama campaign actually test?
The campaign's splash page asked visitors for their email address. Dan Siroker, the campaign's director of analytics who later co-founded Optimizely, set up a test of four button labels ("Sign Up", "Learn More", "Join Us Now" and "Sign Up Now") against six pieces of media: three photos and three videos. That made 24 combinations, shown at random to over 300,000 visitors. He wrote up the experiment in 2010.
Which version won, and by how much?
"Learn More" with a family photo. The original page converted 8.26% of visitors into sign-ups; the winning combination converted 11.6%. That is a 40.6% improvement from changing two things a visitor sees in under a second.
The videos are the most instructive part. The team expected their favourite video to win. Every video performed worse than the photos. A photo is understood at a glance; a video asks for time and attention the visitor has not decided to give yet.
How does a 40.6% lift become $60 million?
By compounding across a very large funnel. The campaign applied the lift to everyone who went on to sign up and used its own averages for what each sign-up was worth:
These are the campaign's estimates, not audited results, but even a fraction of them would be one of the best returns on a single change any website has published.
Why does "Learn More" beat "Sign Up"?
Because it asks for less. "Sign Up" asks a visitor to commit before they trust you. "Learn More" only asks them to be curious, which almost everyone already is if they have landed on your page. The same logic explains the photo beating the video: the smaller ask wins.
| Button | What it asks the visitor | Typical reaction |
|---|---|---|
| Sign Up / Buy now / Book a consultation | A commitment, now | "Not yet, I'm still looking" |
| Learn More / See how it works / Get prices | A little curiosity | "Sure, why not" |
This is not an argument for vague buttons everywhere. The step after "Learn More" still has to ask for the commitment. The point is to sequence the asks: earn a small yes first, then ask for the bigger one once the visitor has seen enough to trust you.
Should every business switch to "Learn More"?
No. The Obama result is a reason to test, not a rule to copy. What wins depends on how warm your visitors are:
- Cold traffic (ads, first visits) usually responds better to a small first step: see prices, take a two-minute quiz, watch how it works.
- Warm traffic (returning visitors, referrals, people who searched your name) often converts better on a direct ask: book, buy, call.
- High-value services such as private clinics, agencies and B2B software tend to benefit most from a smaller first ask, because the final commitment is large.
How do you test this on your own site?
- Pick one page and one button: usually the homepage hero or your main landing page.
- Write two versions: your current commitment ask, and a smaller first step that leads to it.
- Split traffic evenly and run it until each version has a few hundred conversions, not a few days. Small samples produce confident-looking noise.
- Measure the end goal, not just the click. A button that gets more clicks but fewer bookings is not a winner.
- Test the image too. Swap a video hero for a clear photo of real people and see what happens.
Tools such as Optimizely, VWO and PostHog can run the split; simpler sites can alternate versions by week, which is less rigorous but better than guessing.
Frequently asked questions
Is the $60 million figure real?
It is the campaign's own estimate, published by its director of analytics. It multiplies the measured lift in sign-ups by the campaign's averages for volunteering and donations per sign-up, so treat it as a well-founded estimate rather than an audited number.
Why did the videos lose?
A video asks visitors for time before they have decided you are worth it, and many visitors never press play. A strong photo communicates the same message instantly.
How long should an A/B test run?
Until each version has enough conversions for the difference to be meaningful, usually at least a few hundred each, and across full weeks so that weekday and weekend behaviour are both included.
What is a good first button for a service business?
Something that matches what the visitor wants to know next, such as "See prices", "How it works" or "Get a free review of your site", followed by a clear booking step once they have the answer.
About the author: Tengo Meskhi is CEO of Conceptdigital. He was previously at McKinsey and QuantumBlack, and Deputy CIO at Bank of Georgia.
Want to know what the main button on your homepage should say? We'll review it, and the page around it, for free. Book a 30-minute call.