'Join the conversation' comment button: no difference in comments
Test run byMarketingExperimentsResearch publisher
Commentary byLiftArchiveOur analysis, not theirsChanging the MarketingSherpa blog comment button from 'Submit Comment' to 'Join the conversation' made no significant difference to comments, not even at 80% confidence.
MarketingExperiments (MECLABS) ran the test on both its own blog and the MarketingSherpa blog.
MECLABS rolled out the new copy anyway, because the test showed it did no harm.
- Blog comments: no significant difference on the MarketingSherpa blog, even at an 80% confidence level.
- Verdict: flat. The copy change neither helped nor hurt here.
- The same test on the MarketingExperiments blog showed a 34% lift, but only at 82.9% confidence, which is also short of a reliable result.
The change
Only the comment button text changed.
Simplified mockup drawn by LiftArchive from the shared screenshots.
Control
The comment form ended with a standard button reading 'Submit Comment'.
Variant
The same form with the button reading 'Join the conversation'.
Why it likely came out this way
Our reading of the result, not the practitioner's claim.
- Readers who decide to comment have already written their reply, so the button text is unlikely to change their mind.
- Comment volume on a blog post is small, so modest effects are hard to detect.
- The new copy fit the brand's message better, so a neutral result was enough to justify shipping it.
Caveats: Sample size, comment counts and duration were not shared. This is a low-volume action, and the test is from 2011.
Who ran it
MarketingExperiments ran this test and published the results (MarketingExperiments). The numbers are theirs; the summary and commentary on this page are ours.
Read the original case study on MarketingExperiments →
How to run this test on your site
- Hypothesis
- If we change our comment button from 'Submit Comment' to copy that states the value of commenting, then more readers will post because the action feels less like a chore.
- Primary metric
- Comments per post view
- Guardrails
- Share of comments removed as spam
- Comment length or quality
- Who should test it
- Content marketers and publishers whose blogs already get regular comments. On a low-traffic blog you probably won't get enough comments to detect a change.
Questions
Does button copy affect blog comments?
In this test it didn't, at least not measurably.
The decision to comment is probably made before the reader reaches the button.
Is a flat test result useful?
Yes. Here it showed MECLABS that the on-brand copy wouldn't hurt, which was enough to make the change.
Is 'Submit' bad button copy?
Many tests find generic 'Submit' buttons underperform copy that says what the visitor gets.
This test is one where the change made no measurable difference.
Similar tests
MarketingExperiments
Online DialogueMarketingExperiments