This is my story of trying to launch a startup. Told as it happens, from inception to (hopefully) launch.

This post is about how the idea was conceived.


How much did I spend on X last Y?

Lots of amazing ideas begin with a question, and that's what this story started with.

One day while I was traveling in Colombia, I got curious about how much money I had spent on Ubers over the past month. And for that matter, the past year. I had been traveling for three years with virtually no income, so it wasn't the first time I had wondered about my expenses. I figured that with all the apps and tools we have at our disposal, I should've easily been able to satisfy my curiosity within thirty seconds. And yet, I've tried all manner of financial tools over the years, and I couldn't remember a single one of them being able to do this.

So... I had my initial itch.

Idea #1: A smart search engine for my transaction data.

Here's what I wanted:

  1. I put in a search term. Let's say "Uber."
  2. I'm shown all my transaction data for that term.
  3. Crunch my data and give me some insights.

Insights didn't have to be complicated. Just some simple number crunching. I wanted to answer questions like:

  • How many Uber transactions did I have this week/month/year?
  • How much did I spend on them each week/month/year?
  • How am I trending?

And of course we could get even fancier with some graphs, more statistics, maybe some tracking functionality... but I'm getting ahead of myself.

I just wanted to answer: "How much did I spend on X last Y?"

At this point the idea was just something in the back of my mind. At the time I was focused on building a crypto trading bot and it wasn't until a couple of years later that I revisited the idea. I was back in Colombia in the middle of a full COVID lockdown, and I had basically thrown in the towel on TripleChili. TripleChili had been more of a fun project than a business idea, but I had decided that going forward, I was going to sell first before building. That's when I dusted the mental cobwebs off the idea and decided to look into it.

The first step was to make sure that I wasn't going to work on something that already existed. While I had tried every tool out there over the years, I figured there might be some new ones I hadn't seen yet.

Here's a short list of the best and most well-known ones I found:

Trying out the search functionality of all these tools yielded the same results: A list of all my transactions matching my query string. No crunching. No insights. So... maybe there's something there.

One more thing to verify was whether it was possible for me to even access any transaction data. I knew about Plaid but had never played with it. As it turns out, it's super easy to use their APIs to access financial data.

Alright, let's go!

So now I figured I had enough data to try to validate my idea. But I needed to sell first, then build.

Having read The Lean Startup several years ago and being tapped in to the startup community, I was confident I knew the formula:

Build a pre-launch landing page and take pre-orders.

I just needed enough pre-orders to prove that this was something other people wanted, and then I could start on the fun part of building a product.

I just had two problems:

  1. I had no idea how to sell a "smart search engine for transaction data."
  2. I had never sold anything before, so actually...  I just had no idea how to sell period.

Addressing problem #2 was straightforward: learn and practice. I had had some exposure to marketing over the years but now I needed to get serious. So I purchased a bunch of courses on Udemy and read The Adweek Copywriting Handbook. It was enough to take a stab.

Addressing problem #1 took a bit more creative thinking. I didn't think a "smart search engine" alone was going to be sellable. I needed another angle.

At the time, I had just read The Psychology of Money, and I became convinced that I needed to rethink the approach to finances. Instead of a purely analytical approach that most tools have, I needed to take a behavioral one. One quote that really stuck with me was:

"Money has a little to do with how smart you are and a lot to do with how you behave... financial success is not a hard science. It’s a soft skill, where how you behave is more important than what you know."  

Plenty of smart people make terrible choices with their money, despite knowing what choices they SHOULD be making. And one of the reasons for this is social comparison.

We naturally compare ourselves to others and when we compare ourselves to those that seem rich, we're making the comparison based on what we see. Money spent. But wealth is the exact opposite. It's money NOT spent. This also makes it NOT visible, which in turn makes the same comparison much harder.

But what if we COULD see wealth? What sort of comparisons would we make if we knew that our neighbor with the Tesla was in serious debt? Conversely, what if we knew that our other neighbor that still drives a fifteen-year-old Toyota also had a boatload of Tesla stock in his portfolio? How might this influence our behavior then?

This led to idea #2.

Idea #2: Use social comparison to bring awareness around our own financial habits.

This got me excited. If I had a bunch of transaction data from enough people, I'd be able to extrapolate some really interesting data. Stuff like:

  • How much do people spend on Uber?
  • What about Starbucks or Amazon?
  • How much debt to people really have?
  • What's the average savings rate or debt-to-income ratio?

Even MORE interesting would be the answers with respect yourself. So then the questions become more like:

  • Do I spend as much at Starbucks as other people in my neighborhood?
  • Does everyone else shop on Amazon as much as me?
  • Are other Software Engineers paying as much for rent?
  • Am I saving more than other people in their 30s?

There could be all sorts of ways to group the data, and there even seemed to be a strong case for individuals to create their own groups. Think about an accountability group, or even a friendly competitive group around portfolio management.

Having a way to concretely compare our financial habits against others could be a powerful way to bring awareness to the way we handle our money. Personal finance was always... personal. But once you started throwing other people into the mix, there were all sorts of new possibilities.

I also thought that the perfect delivery mechanism for this would be email. A big reason why I had abandoned so many tools in the past was because there was too much friction to use them. I didn't need a budget and I didn't need to categorize every transaction.

A weekly email to track financial habits and see how they compared to others seemed like a good starting place. But I had a classic chicken or the egg: In order to build the product, I needed the data, but to get the data I needed people using the product. This didn't seem like a problem I could really solve at this stage, so I decided to refine the idea a bit.

Idea #3: A personalized financial newsletter.

While I couldn't build a tool that relied on data that I would get after people started using the tool, a compilation of just MY data delivered to me in a weekly email could still be extremely useful. Not only would an email be frictionless, a regular weekly email would make it easy to gauge progress and reference history, which doesn't quite work so well in other tools or spreadsheets. On top of that, I figured it would be relatively easy to build out an MVP and once I had some users and data, I could start adding features from there.

So that's what I decided to start with: A personalized financial newsletter.

I brainstormed a few names based on which domain names were available and settled on Cabbage Stack.

Now I just had to validate the idea.

Episode #2 is about my first attempt. Spoiler alert: It didn't go quite as swimmingly as I had hoped.