What I'd do differently
I could have never anticipated all the things I'd learn while building Timestamps. Product development, web development, digital marketing, growth, sales... the list goes on.
Here's what I'd do differently if I had to restart.
Not start with an HTML template.
I'm not a front end developer and wanted a quick way to get started with the web UI, so I started with a free HTML theme. It wasn't a theme meant for web applications and quickly became a pain to work with. The code quickly evolved into a mess with tons of custom css, which is what I wanted to avoid in the first place.
I ended up rewriting the entire site in bootstrap, which is what it currently uses. This was an expensive rewrite time-wise, but I was able to do it in 1-2 days because it was mostly a port of the old site. In retrospect, learning Bootstrap seemed daunting, but doing it from the start with a good foundation would have been a good idea.
If I were to restart, I'd potentially look into no-code solutions like carrd for even more rapidly prototyping and getting the UI out there. Learning to trade money for time & energy was one of the biggest lessons I learned.
Leverage no-code tools.
Developers vastly underestimate the value of their time. Just because you can code, doesn't mean you should. As a founder, your most valuable resource is time. Spending time developing when there are no code solutions that will get the job done in a fraction of the time is inefficient.
At first, it seems better to code it yourself, which you can do for free. But if you value your time for what it's worth, you'll find that the money you'll spend on no code will nearly always be worth it. Especially at the start.
Tools I would have looked into more deeply:
- carrd
- webflow
- servicebot
The straw that broke the camel's back for my HTML theme based UI was making a pricing page. It looked awful. Bootstrap had great looking pricing page components and I switched immediately. It still took a lot of work to make a pricing page in Bootstrap – services like Servicebot could have reduced that time even more radically.
Not tried to port to Google Cloud
I tried porting to google cloud so the site could run for "free" on their free tier. I've written at length before – this was an awful idea and wasted an immense amount of time and energy.
I did learn a ton about Google Cloud, but I ultimately would have saved time by never switching from Heroku. $7/month is not that much. Especially since I did start to make some money
Focused only on the most engaged customers
I spent many, many hours cold contacting everyone I could possibly think that might have a use for this. I wasted a lot of time reaching out to people who it wasn't a good fit for.
Identify the people that get the most value from it, and laser focus on them. How much do they need your thing? Will they use it frequently? Does it allow them to make more money?
Things I did well
I didn't mess up absolutely everything. There were things that were good:
- Porting back to Heroku from GCP
- Logging directly into Google Sheets
- Using Firebase to quickly get backend storage working
- Using Stripe instead of Paypal (better community support, better docs & developer experience = faster time to market)
- Deciding to make it free, while finding a way to keep getting paid enough by my enterprise customer to pay server costs
- Stopping putting effort into it when I realized it wasn't a good idea to build a business on.
