Is a shorter funnel always better?

They say a shorter funnel is better for improving conversion rates, but...

When planning a screen, there's a common dilemma we face. When we have a lot of information to convey, we often wonder: "Should we cram it all into one screen, or add another layer of depth and split it across two?"

From an information architecture perspective, putting too much on one screen isn't ideal... But if my goal is to drive users toward a specific conversion event (like a subscription conversion), adding depth actually feels counterintuitive. The moment you add another step, unless the drop-off rate is 0%, you're bound to lose users, which inevitably leads to a lower conversion rate. Sure, better information structuring might deliver a sharper, more impactful message, which could help boost conversions. But for that to work, the positive effect would have to outweigh the sheer volume of drop-offs—and intuitively, that's a tough sell. Instead, it feels much safer to simplify the funnel and skip the extra depth, even if the message ends up a bit blunt or less impactful.

Within Alarmy's premium feature funnels, there's a spot where a bottom sheet pops up right before the purchase screen. The journey is designed to first notify users via the bottom sheet that the feature is premium, and then show the purchase screen. Technically, we could just let them know it's a premium feature on the purchase screen itself, but we intentionally added this extra structural layer to deliver a sharper message. As feared, the drop-off rate from the bottom sheet to the purchase screen hit a staggering 80% to 90%. I immediately wanted to run an experiment to kill the bottom sheet and simplify the funnel, but I wasn't entirely convinced that "funnel simplification = increased subscription conversion." The drop-off rate was so shockingly high that I started wondering, "Is it just because the message on the bottom sheet isn't appealing enough?" So, we decided to run two separate experiments: "Refining the bottom sheet message" and "Simplifying the funnel." Both shared the same goal: boosting the subscription conversion rate. And the results were pretty fascinating.

Doubling the subscription conversion rate with just the power of messaging..!

We tweaked the bottom sheet message to make it more appealing. Since we were only changing the copy, I didn't expect it to have a massive impact on drop-off or conversion rates—but the results proved me wrong.

• The experiment groups saw subscription conversion rates increase by roughly 1.16x in the US, 1.5x in China, and 2x in Japan.

• Notably, in Korea, all experiment groups except Group 4 showed more than a 2x increase in subscription conversions. (Group 2 hit 2.6x..)

• The variance in results across countries could come down to translation quality, since the "message" was the core variable. Or perhaps how users interact with bottom sheets differs slightly by region.

It was amazing to see that messaging alone could double the subscription conversion rate. The fluctuations across the different experiment groups just go to show how crucial messaging really is. I felt like the winning messages from this experiment would be incredibly useful for future tests. So, for the subsequent funnel simplification experiment, we baked those exact messages right into the purchase screen.

Simplifying the funnel doesn't always guarantee better conversions.

This time, we tried skipping the bottom sheet entirely and showing the purchase screen right away. In some variants, we also integrated the winning messages from the previous "message refinement experiment." Since we reduced the depth, I was absolutely certain the experiment group would win. But that wasn't exactly the case.

• Even though we showed the purchase screen immediately—improving the "purchase screen entry rate" by 45 to 50 times compared to the control group—some experiment groups didn't see a meaningful increase in subscription conversions. (US Group 2, China Group 1). Looking at the raw numbers, this is a pretty wild outcome. We eliminated a funnel step with severe drop-offs, and conversions still didn't improve...?!

• Meanwhile, the performance of the Korean experiment groups was impressive once again, with one group generating an 11x subscription conversion rate. (Korea Group 3). Even with the same simplified funnel, the variance between experiment groups was massive depending on the screen layout. (Groups 3 and 4 had a 2x difference between them.)

• In both experiments, the US and China didn't react very elastically to the variables, whereas Korea and Japan showed highly sensitive responses.

Cross-checking the results: Refining bottom sheet messages vs. Simplifying the funnel

In China, one group from the message refinement experiment drove a 1.5x improvement (Group 3), whereas the funnel simplification experiment had two groups that fell short of that mark. (Groups 1 and 2). Despite simplifying the funnel by a whole step, the improvement was essentially meaningless. If anything, it showed that having an extra layer of depth—and using that bottom sheet to deliver a highly impactful message—can actually lead to a higher subscription conversion rate.

We noticed a similar case in the US. The message refinement experiment yielded a group with a 1.16x improvement (Group 2), but the funnel simplification experiment had a group that couldn't even match that. (Group 2)

Q. If you were short on time and resources, would you add depth and refine your messaging? Or would you simplify the depth and focus on the layout of the merged screen?

A. I'd choose the latter, but I think I'd always have lingering regrets about the former. So, I'd probably implement the latter first, and whenever time and resources opened up, I'd definitely use it as a control to run an experiment on the former.