Do CAPTCHAs hurt conversion rates? How to measure it and reduce friction
Updated:
Quick answer
A CAPTCHA can lower conversions when it makes real people do extra work, such as solving image puzzles or retrying after errors. A check that runs invisibly adds no step, so the loss should be small, but you should measure it instead of assuming. Compare the form completion rate before and after, split by device and by browser, and look at accessibility, since people who cannot solve a puzzle never convert.
Where does CAPTCHA friction come from?
- Image or text puzzles that real visitors must solve, which they may fail or abandon.
- Checkbox challenges that escalate to a puzzle for people on VPNs, shared networks or privacy browsers.
- Widgets that load slowly and delay the submit button.
- Consent banners that block the CAPTCHA until the visitor accepts, so the form cannot be sent.
- Inaccessible challenges for people using screen readers, keyboard-only navigation or voice control.
How do I measure the effect on conversions?
- Define the metric: form completion rate, meaning completed submissions divided by visitors who started the form.
- Record the rate for at least two weeks before the change, so weekdays and weekends are covered.
- Change one thing at a time, then record the rate for the same length of time afterwards.
- Split results by device, browser and country, because problems often hide in one segment.
- Watch blocked and failed verifications. A rise in failures from real-looking traffic points to friction.
- If you can, run an A/B test with the CAPTCHA enabled for one group and an alternative control for the other.
How to reduce friction without losing protection
| Choice | Why it helps |
|---|---|
| Invisible or proof-of-work checks | The visitor does nothing, so the form has no extra step. |
| Server-side token verification | Protection does not depend on visitor effort. |
| Place the check on abused forms only | Low-risk pages stay fast and clean. |
| No cookies or consent dependency | The check works for visitors who decline cookies. |
| Keyboard and screen-reader testing | Makes sure nobody is locked out, in line with WCAG 2.2 (3.3.8). |
What about accessibility?
WCAG 2.2 success criterion 3.3.8 (Accessible Authentication, Minimum) says an authentication step should not rely on a cognitive function test, such as solving a puzzle, unless an alternative exists. A CAPTCHA that locks out people with disabilities costs conversions and creates legal risk.
BurgCaptcha and conversion rate
BurgCaptcha shows no puzzle and sets no cookie, so there is no consent gate and no extra step for real visitors. Run it on one form on the free plan, which covers 1,000 requests a month, and compare your own completion rate.
Frequently asked questions
Do CAPTCHAs reduce conversions?
Visible puzzles can, because they add effort and some people give up. A check that runs in the background adds no step. Measure your own form completion rate before and after to know for certain.
Is it OK to remove the CAPTCHA to increase conversions?
Only if spam and fake sign-ups are not a problem for you. Otherwise replace a heavy challenge with a low-friction one instead of removing protection.
How long should a before/after test run?
At least two full weeks on each side, so that weekday and weekend behaviour are both covered, and long enough to collect a meaningful number of form starts.
Can a CAPTCHA hurt SEO?
A CAPTCHA on forms does not affect search rankings directly. Heavy scripts can slow a page, so keep the widget light and load it only where needed.