Acquisition #4 - SentimentInvestor.com

Sentiment Investor is XO Capital's 4th Acquisition!

Three weeks ago, we purchased sentimentinvestor.com. It was created by a team of university students in the UK who had already monetized it a bit. I've been keeping an eye on this space for some time and felt like this hatchling of a product was a good toe-hold into a market I want to be in and I love the practical machine learning aspect.

Effectively Sentiment Investor scrapes social media data for stock (and crypto!)  mentions and runs it through a series of machine learning models to determine potentially market moving trends as they're happening. It sells that data to hedge funds and other institutional investors who then use either our raw data or processed data to feed into yet another machine learning model to try and get some alpha (returns).

The thesis here is that more and more retail investors moving in a coordinated way (i.e. the meme stock phenomenon) is starting to have real sway over stock market prices.

Sentiment Investor Business Model

What went right

We bought the product at a great price. I have mixed feelings about that statement. If we were a venture backed startup I would say fuck price, we need volume. If the company "works" in venture land, the purchase prices won't matter. But we're not in venture land (some thoughts on that later perhaps), we're in bootstrapped land for now and just starting to figure out how to take in outside capital.

What went wrong

OMG. During diligence the dudes seemed great, super on top of it, super quick to respond. As soon as the money was wired, those kids disappeared (classic). Took FOREVER to get the domain, access to the damn email account as well as the code. We did get access to the google cloud instance which was great but I was starting to freak out after not hearing from them for 5 DAYS and spent an entire day ripping code out of docker containers and cloud functions (dumpster diving).  Not fun!

As it is now, I still have to create a new email address because the dude won't change the 2fa authentication on the primary account. stupid.

It was my fault. I just wired everything all up front without using escrow (I passionately hate escrow.com). To be fair this was our smallest acquisition ever. A toy. My toy! But I still should have followed best practices. Even as I write this we just wired for (sneak peak) acquisition #5, which is already going better than this one.

Oh and customer hand offs? Zero. Despite having an incentive in place to  effectively double the purchase price if they help us maintain the current MRR, we got absolutely no introductions to their existing customer. They also had a great pipeline (which we also have an incentive in place for). Did we get an intro to anyone in their pipeline? Nope! nada.

Felt like after we wired, they just DGAF anymore. Fine, but

Where we go from here.

This one was a complete rewrite. I knew it from diligence and accepted that. They were running everything through Firebase and it was costing $1400 a month! Damn! I think I'll get it under $300 all in each month.

From there it's a pretty cool concept. We're looking for signal in noise. We're using machine learning to get there. And for this product, the longer it's around, the more valuable it becomes. If you're a hedge fund trying to do practical machine learning on social media data, you can either spend months gathering and cleaning the data, or just buy it from an alt data provider like Sentiment Investor.

I think the re-write will be done before the new year.

Anyways I think we're over the hump. Hope you guys enjoy our trials and tribulations! We have the assets now (three weeks after the deal "closed") and are plugging away on the rebuild. Super pumped to turn this little biz around. I thought it might be fun to build an algo that trades automatically based on the data from the product. Just put like $5k in and trade a bunch of shit coins based purely on the data the product provides. Could be a horrible idea, but it will at least be entertaining.