Rhyming-poem search ad vs plain ad: CTR 1.1% to 2.13%, same CPA
Commentary byLiftArchiveOur analysis, not theirsPatrick McKenzie ran a deliberately campy rhyming search ad against a well-optimized plain ad for Bingo Card Creator, and click-through rose from about 1.1% to 2.13% at roughly the same cost per trial download.
Conversion per click dipped slightly, from 22% to 19%.
The test was published on the Kalzumeus blog in 2006.
- Click-through rate: about 1.1% for the plain ad vs 2.13% for the poem, roughly double the clicks.
- Conversion to trial download: 22% vs 19%. The poem ad converted slightly worse per click.
- Cost per trial download: about 32 cents vs about 31 cents, essentially flat. The win is more volume at the same cost.
- The full experiment, including a third ad, cost $4.61 in clicks over about a week.
Control
A plain AdWords ad, refined over a month, headed "Sight Words Bingo" with the text "Don't prep for an hour. Be ready in minutes with our software."
Variant
A rhyming ad headed "Dolch Sight Word Bingo" that read "Roses are red, violets are blue, Bingo makes students love you." It used the same keywords, ad group and landing page.
Why it likely won
Our reading of the result, not the practitioner's claim.
- Most ads on the page looked alike, so a playful poem stood out to people scanning results.
- The rhyme spoke to teachers' goals (students enjoying class) rather than listing features.
- A clever ad can attract curious clickers who are slightly less ready to act, which fits the small drop in conversion rate.
Caveats: Tiny spend and about one week of data, so sample sizes are very small, and this is a 2006 AdWords rotation rather than a page test. Bingo Card Creator was downloadable desktop software at the time, not SaaS.
Who ran it
Patrick McKenzie ran this test and published the results (Kalzumeus (patio11) blog). The numbers are theirs; the summary and commentary on this page are ours.
Read the original case study on Patrick McKenzie →
How to run this test on your site
- Hypothesis
- If we run a distinctive, playful ad alongside our standard ad on the same keywords and landing page, then click-through will rise without raising cost per sign-up because it stands out from similar-looking competitors.
- Primary metric
- Cost per trial or sign-up
- Guardrails
- Conversion rate per click
- Click-through rate
- Trial-to-paid rate from each ad
- Who should test it
- SaaS and lead-gen businesses on search ads where competitors' ads all look the same. A jokey tone may not suit high-trust or enterprise buyers, so judge on cost per qualified lead, not clicks.
Questions
Do creative or funny ads get more clicks?
They can. Here a rhyming ad roughly doubled click-through rate compared with a plain one, though on a very small sample from 2006.
Should I judge an ad test on click-through rate?
Not on its own. The poem ad converted slightly worse per click, so the real test was cost per trial download, which stayed about the same.