Easy A/B testing
Dead simple A/B testing tool. Developer friendly.
So I killed this product a couple years ago. The end user was probably a bit off (someone with technical + marketing role) so it was hard to find the correct people. But the final nail in the coffin was that Google Chrome (and other browsers) started to change their 3rd-party cookie policies, which meant Easy A/B Testing would cease to work correctly.
All in all this was very much a tech-first approach than a problem-first approach.
So some take-aways:
- Think who will be the person who pays the bills and how you're going to reach them
- Start with a problem, not with a cool technology
Got help from @lehtu for doing a simple product explainer video.
Used Keynote to compose the video, and MacOS' speech synthesis (voice called Kate) to narrate the video. Background music taken from first Google result of "free ukulele music".
Testing hypothesis that people are not willing to subscribe to a mailing list, but that they'd like to get hands-on with the service.
Driving some traffic to the website from Google Adwords to see how my funnel works.
Results:
- CTR 2.07%
- Average CPC 0.61€
- First registered user!
- Landing page conversion rate 5%
- Two best-performing ads were for "Optimizely alternative" and "Free A/B testing tool"
Learnings:
- Saved my day (and motivation) to get the first user to the service!
- Landing page has converted steadily around 5% the whole time (mailing list and free tier)
- With optimistic 0,60€ CPC, 5% conversion to user, 10% conversion to paid user, it would mean one paid user would cost 120€
- If billed annually, it would mean 10€/month to break even from ad costs (with 1 year churn)
A service needs a logo, right? Now it has one 🤩
Changed landing page from collecting mailing list to actually directing users to register to the service. Now anyone can join, not just the people from initial mailing list.
-
2
I think I understand the thinking process behind your logo (two unfinished letters: "a" and "b" referring to the ongoing process of testing?) and its design is clean but difficult to read and asymmetric: I am a graphic designer, let me know if/when you want to improve it. I am happy to help ;)
-
1
Nice, thank you! I'll be definitely improving it, thank you for your feedback. I'm not a designer myself, so any help is greatly appreciated! 😊
-
-
1
I will design something for you later this year... ;)
Sent email for mailing list that I got from earlier Indie Hackers post.
No sign-ups resulted.
Learnings:
Even though I'm really excited about my own product, people have other priorities. Perhaps ease my way in next time?
Users can now register, login, add pages and get API keys (so "full" flow). No payment options available yet.
Tried to setup A/B testing to one of my own websites, and made it in 3 minutes, as my landing page promises.
Advertised the landing page to a group of developers on Facebook that have recently created their own static website.
Results:
- 30 visits from ads
- Total visits: 84 (41)
- 0 conversions
- 0,28€/click
- 1.12% CTR
Learnings:
- Seems that "cold" leads do not convert to mailing list with same conversion rate
- Some organic sharing occurred because there was ~30 people bump in the very beginning
- Hypothesis here is, that people didn’t join the mailing list, but they would have signed up to try if signup was available
Comment