Launched a Project Twice with Two Horrible Names but Still Made Money!

I built a site for tracking AI answers about brands and instead of calling it trackAIanswers.com, trackyourbrand or something sensible I gave it the worst possible names - TWICE!

I'll get to why the names are so bad, but here's the story so far:

Early 2023 I was egosurfing the new AI-chatbots and asking Chat-GPT what it knows about me. After picking up my bruised ego since ChatGPT failed to list me as one of the co-founders of Teleport (a a16z portfolio startup I co-founded a decade ago), I started looking deeper into Large Language Models.

What started as an experiment in understanding how LLMs encode information turned into a project called HaveIBeenEncoded.com to help everyone track what AI-chatbots know about their name and get updates on those answers changing.

The signup rates were high and I thought I had something. I was even invited to talk about it on the radio in the US as the hallucination topic was making rounds, but I quickly learned that consumer willingness to pay for such a service was low and if I wanted to make the project sustainable, I had to make it economically viable.

Enter Enterprise!

The obvious answer here was to help companies track their brands on AIs as the chatbots were likely the place where their future clients would find them. The companies also have clear economic incentives to make sure they know what the AIs are saying about their brands and products. Hallucinations and errors can actually have an economic cost.

"AI for SEO" or "SEO for AI"?

The search landscape was changing. Search Engine Optimization or SEO as it's known - the art of driving your site up in search results is going to change forever. As everyone is scrambling to use "AI for SEO" to automatically generate more content to drive up their links, to me it seemed that the opposite was needed.

It was "SEO for AI" - the magic you could use to drive up your results in AI-chatbot answers that was needed, not "AI for SEO" - the magic of using AI to drive up the results in the old search world. In my engineering brain the name clicked and hence seofor.ai was born.

Oh but the name!

The project did better economically and soon started making more money than the consumer product. It was only when I looked at organic search when I realized that I have made yet another huge mistake. Turns out the name I picked like REALLY REALLY sucked!

The "SEO" keyword brings forth an army of people who are specialists in SEO, the kind that spend most of their waking hours making sure their results appear first in search results. Having "seo" as part of the name may seem logical, but …

it was more like summoning the most impressive army of competitors to fight against to win organic search.

To add insult to injury, the "AI" keyword is pretty much attached to everything these days. Again a hugely competitive keyword that together with "SEO" leaves little room to differentiate from the crowd that wants to use AI to generate more content.

All and all the name described more the technical aspects and logic in my head than anything going on in a person's head curious about what AI says about their company or name.

Wait, but it was profitable nevertheless?

If I would leave the project as it is, it would likely continue to turn a profit. Not because it's making tons of money with two horrible names, but because I just put it together in a frugal way wanting to showcase how to make sure your projects have a good chance of surviving even without outside capital. That’s how startups should be built, right…right? 😉

Here are the costs for the project to help you see how little is needed these days to test your ideas:

A shoutout here to deta.sh, ImprovMX and all the rest that build the wonderful tools.

Wait, what about development costs?

The project would remain profitable if my only goal was to cover the recurring costs. I think that's a viable strategy for many indie-devs and solopreneurs who know how to code and want to build side-projects. In reality we of course know that the biggest cost here was and will be the hours that go into building and marketing the product.

That one really depends on how you value your time and an easy hack to drive down those costs is to make sure you are having fun building. This way a project can pay off even without financial upside.

For ramen profitability it's still way off. You see I'm the kind of guy who wants to have ramen in Tokyo, so it’s like not even ME can afford me for this project! 😉

So you’re going to continue with two horrible names?

I picked two horrible names and still became "profitable" by way of covering the recurring costs. To find out if I keep the names or make another huge rebranding mistake before focusing more on improving the product you can follow me here on Hackernoon or on Twitter.

And you know you can always just google "track ai answers" to find out! 😁