For those using AI for funnels: Are you letting the AI write the copy too? And how's that converting compared to human-written copy?

Funnels & Conversions
1

I’ve started using AI for my funnels but still do the copy myself.

Thinking about whether to pass that off to AI as well. What differences have you noticed in conversion rates between AI-written copy and the traditional human touch?

2

We use AI for headlines but write CTAs ourselves. Saves time and converts about the same.

3

AI’s solid for most copy, but I always rewrite the emotional stuff myself. It completely misses the psychological triggers that actually convert. After a few hundred visitors, the difference is huge. AI might get you 2-3%, but when I write the emotional hooks, I’m seeing 5-6%. Let AI do the structure and feature lists, then you handle the parts that make people feel something.

4

AI copy works fine if you fix the calls to action.

5

I’ve been split testing this for months across different app campaigns. Pure AI copy converts around 2.8-3.2% for me, while human-written hits 4.1-4.6%.

Here’s the thing - AI gets way better when you feed it your winning copy first. I take my best headlines and CTAs, then ask AI to create variations in that style. Those hybrid versions convert almost as well as pure human copy.

The real advantage is speed. I can test 10 AI variations in the time it takes to write 2 human ones. More tests usually means finding better winners faster.

If you’re getting into AI copywriting, this breakdown covers some solid tools:

My process now: AI for volume, human editing for the winners that show promise.

6

AI does fine with basic funnel pages, but I tweak the headlines and main value props myself. Conversion rates stay pretty much the same if you give AI clear prompts about who you’re targeting. I treat it like a first draft - just needs some human touch on the important stuff.

7

I run AI copy for 48 hours straight, then test it against human-written versions. My rewrites usually beat the AI by 15-20%. But here’s the thing - AI’s crazy fast. I can pump out 5 AI variations while writing just one human version. With Web2Wave, I push new copy live instantly and keep tweaking whatever converts best.

8

Switched to AI copy 6 months ago after testing three apps.

Conversions dropped 0.3% at first, but the speed boost made up for it. Now I’m testing 15 copy variations weekly instead of barely getting 3 done.

Trick is feeding your best performers as examples. Don’t start from zero.