Simple Choice: http://unbouncepages.com/simplechoice

This project is sponsored by my company which shall remain nameless to minimize bias and risk.

Simple Choice helps people who are overwhelmed with product choices when shopping. We cut through the noise. Here's our evolution:

Project name was "Best in Store" and the landing page variants A&B asking for email addresses got a pathetic overall conversion rate of 4.17% - one of them was myself; so in actuality we got 2.89% before we put it out of its misery. Below is variant A.

A. Best in Store

B. Our variant B had a brilliant important change of the tag line from "The best products, mapped for you." to "Expert recommendations, near you." Flows better, made no difference.

We hunkered down, refocused, pivoted a few hypotheses (ok all of them) and came up with 2 variants: G with just a button to learn more leading to a second page that asked for a product type and email address and E asking for email upfront. (We have a couple other variants that were either not launched or didn't show high promise so I won't clutter this up with those.)

G. Simple Choice, Only free trial button

E. Simple Choice, Email address ask up front, free trial button

We reached out to our social networks again - but this time we didn't get much traffic at all.

Question 1: how do you get past saturating your social networks with your brilliant product iteration of the day?

@acroll suggested we put up an AdWords campaign. I'd only done that once before so my partner and I stumbled through the phenomenally boring and text-filled process to create two ads - one that mentioned TVs (since our target token shopper is staring at a bunch of TVs), and one more generic abotu product choices in general. Then I sat there obssessively hitting refresh throughout the day to see no traffic at all. Wasn't til yesterday we realized perhaps Google only updates those analytics reports once a day (around 1am in my guesstimation).

BOOM Google sent us 69 clicks and our conversion rate on Variant G (free trial) went to 54% (it's now at 70.68% with no changes to the page) - the second page requesting product and email was 33.3% (it's now at 9.91%). Variant E topped out at 6.9% so we killed it.

Question 2: We really worked hard at changing only one thing at a time - but now it seems we have a basket full of different things. How do we compare Variant G with 2nd page requesting product+email and Variant E which was one page asking for email upfront? Or can't we?

Adam mentioned manually normalizing - OK, how do we do that?

Thanks,

PJ

Patricia Ju

Vater of Innos

1

3