We analyzed 18,648 popup campaigns and 469 million displays to see which elements actually convert. The gap between top and bottom is wild.
A few weeks ago, we pulled aggregate data from our platform and spent a significant amount of time going through it. I want to share what I found, because some of it surprised me, and I’m curious if it matches what others are seeing.
What’s the baseline conversion rate for a popup?
The average across all 18,648 campaigns (each with at least 100 displays) is 0.52%. That’s leads divided by displays. Not clicks, not hovers. Actual form completions or goal completions.
Some people will say 0.52% is terrible. Others will say it’s fine depending on traffic volume. It’s really just a starting point for comparison.
The dataset spans 3 years, 469.9 million total displays, and 2.43 million leads.
Which elements actually push conversions above that baseline?
A lot. But the degree varies wildly. Here’s the full breakdown:
| Element | CVR | Campaigns | Total Displays |
|---|---|---|---|
| lottery-ball | 7.40% | 54 | 217K |
| spin-to-win | 5.56% | 251 | 6M |
| multi-choice | 5.20% | 136 | 1.2M |
| dropdown-input | 3.36% | 395 | 8.8M |
| rating | 3.35% | 57 | 380K |
| radio-input | 2.74% | 126 | 4.4M |
| date-input | 1.93% | 159 | 5.3M |
| full-name-input | 1.85% | 1,214 | 25.2M |
| notice-and-consent | 1.53% | 1,705 | 55.3M |
| coupon-code | 1.49% | 1,703 | 36.6M |
| countdown | 1.46% | 906 | 20.2M |
| email-input | 1.39% | 3,976 | 107.8M |
| lottie animation | 1.31% | 584 | 20.6M |
| phone-input | 1.34% | 1,379 | 35.3M |
| headline | 0.72% | 13,059 | — |
| button-wrapper | 0.66% | 13,235 | — |
| text | 0.60% | 14,124 | — |
| image | 0.50% | 16,755 | — |
| button (no form) | 0.46% | 11,839 | — |
Note on reading this table: most popup campaigns combine multiple elements (image plus headline plus email field plus button, etc.), so a single campaign appears in every row that matches one of its elements, and the campaign counts overlap heavily. Each row’s CVR is the rate for campaigns containing that element, not for campaigns running it as the sole mechanic.
Why does lottery-ball have a 7.40% CVR?
Gamification. That’s the short answer.
The lottery-ball element is a spinning/draw mechanic where visitors can “win” a discount or prize. It only has 54 campaigns and 217K displays, so I’d be careful reading too much into that 7.40% number. Small sample, likely skewed toward stores that tested it specifically because they thought it would work. Classic survivorship possibility.
Spin-to-win has 251 campaigns and 6 million displays, which is a much more reliable number, and it’s still at 5.56%. That’s roughly 10x the platform baseline. That’s not noise.
The pattern is clear: give people something to do other than “fill in your email,” and they engage more. Multi-choice questions (5.20% CVR, 136 campaigns) follow the same logic. It’s an interaction, not a form.
What does the email-input number actually mean?
This is where it gets interesting. Email-input has the most campaigns by far (3,976), and the most displays (107.8 million), but its CVR is 1.39%.
That’s not bad. It’s 2.7x the baseline. But it’s miles behind gamification elements.
I think what’s happening is that email-input is the default choice. Stores add a popup, the template has an email field, and they ship it. There’s no optimization pressure. Meanwhile, someone who builds a spin-to-win popup almost certainly thought harder about what they were doing.
So the CVR gap might partly reflect intent and effort, not just the element itself.
Why is the image sitting right at the baseline?
Images are in 16,755 campaigns — about 90% of the dataset. So the “image” row is essentially measuring the platform average, because image-containing campaigns and the platform are nearly the same population. That’s why the 0.50% lands right where the 0.52% baseline is. The takeaway isn’t “images hurt” — it’s that an image alone doesn’t appear to lift CVR above average.
Buttons without forms sit at 0.46%, slightly below baseline. If your popup’s goal is a click-through to another page rather than a lead capture, it shows up here as a low CVR because the dataset measures conversions, not click-throughs. The number isn’t necessarily bad for those campaigns; it’s measuring the wrong thing for what they’re optimizing.
What about notice-and-consent at 1.53%?
This surprised me. That’s 1,705 campaigns and 55.3 million displays. People are apparently clicking “I accept” or completing GDPR/cookie consent flows at a meaningful rate.
My honest read: this probably includes popups where the consent action IS the conversion, like an age gate or a preference center. It’s not apples-to-apples with a lead generation popup. But I included it because it’s in the data.
Is gamification actually worth building?
Probably, if your audience tolerates it. Spin-to-win works really well for e-commerce.
The multi-choice data is what I’d actually act on. 5.20% CVR at 136 campaigns and 1.2 million displays is a strong signal. Asking a qualifying question (like “what’s your biggest challenge?”) before showing an offer converts better than just asking for an email. It’s a micro-commitment. You’re not asking for personal info first.
Dropdown-input at 3.36% (395 campaigns, 8.8 million displays) is the same idea. A selector question before the email field.
Personal take
I’ve been working in popup tools for a while, and I still find it counterintuitive that adding friction in the form of a question or a game improves conversions. You’d think asking people for more information would make them leave faster.
But the data consistently says otherwise. Engagement changes the psychology. Once someone answers one question, they’re invested. The email feels smaller.
The thing I keep coming back to is how many campaigns are still just “image + headline + email field + button.” That’s the 0.52% bucket. It works. But there’s clearly room to do more.
What to keep in mind
The lottery-ball sample is small (54 campaigns), so treat that specific number as directional rather than definitive. Those campaigns are likely run by people who specifically sought it out and set it up carefully, which skews the result.
Close rates, scroll depth before trigger, and mobile vs desktop splits aren’t available in this element-level data. All of that would change the picture significantly. Use this as a signal for where to test, not as a controlled experiment you can copy directly.
TL;DR: Looked at 18,648 popup campaigns across 469M displays. Gamification elements (spin-to-win, lottery-ball) hit 5-7% CVR vs a 0.52% platform baseline. Multi-choice questions are at 5.20%, which is the more actionable finding for most people. Email-only popups are the most common setup, but not the highest converting. Small caveats apply to the gamification numbers due to sample size.