Decision Path to My Landing Page and Concorde Effect

While I was developing my new product, I wanted to set up a simple landing page for a waiting list and offering early access. The email list was supposed to be very simple and satisfy the requirements below:

1. Should be very simple, save the email address and subscription date (if possible)
2. Minimum development effort
3. Sending welcome email to new subscribers

For similar tasks I normally choose Mailchimp, I can easily make an Ajax call, create automation for welcome email and it's done. Would take max 1 hour with a bit of design work. However, Mailchimp recently removed automation feature for free plan users. That was game breaker but no worries.

I wanted to try EmailOctopus for a while anyway, they have the welcome email automation, an easy to use API, should be a piece of cake. I created and verified my account, set up the list, api calls and automation details. I tested the flow with adding one or two emails only, after all this method was supposed to satisfy the requirement number 2.

I finished and deployed the page, started testing everything in production as a good indie hacker would do. There was something wrong, main feature which was to save the email addresses was not working, I kept getting errors. When I checked the logs, there it was:

The famous CORS error that left me helpless. That good API wasn't suitable for front end. EmailOctopus was a pretty good tool but it was not satisfying requirement number 3 anymore. So I had to come up with a better solution.

Following the sunk cost fallacy

At this point, I should have (probably) gone back to Mailchimp solution or saved the addresses in a Google Sheet. Something simple (requirement 1) and preventing me wasting more time on this (requirement 2). But I was in the Concorde syndrome spiral. I've spent considerable time already for that welcome email, I couldn't give up yet, I had to make it work.

I knew there were many ways to achieve requirement 3, I was thinking quickest way now, without writing much code, something like this:

1. Save addresses to Airtable
2. Create workflow with Zapier as whenever new record added to Airtable, add it to EmailOctopus list.
3. Create welcome email automation on EmailOctopus.

I followed the exact sequence and set everything up. After using 3 different platforms, I could finally see that shiny automated welcome email 🥹.

Realisation

After the site was deployed, I once again recognized that I had fallen victim to the sunk cost fallacy. I thought I had already invested too much time in this minor detail, so I had to make it work. In the end, it dawned on me that this minor detail wouldn't even matter much; the email would likely end up in the junk folder.

I believe speed is a critical factor in building a new product. From my perspective the ideal cycle should be: built MVP, measure demand, learn, and decide on next steps. Even though that's how I wanted to move, after jumping in the actual work, I forgot my focus and drifted away. I hope this little story help others to prevent similar mistake in their indie hacking journey.