It may strike you as idiotic to start a fund without at thesis. We didn't. We started out with a narrow thesis and dropped it quite quickly. We were going to buy no-code tools. Anything that used to require a developer but now could be built without one would have been fair game. The only problem was that we couldn't find enough deals.
In retrospect, perhaps we could have, or perhaps we could today. We certainly didn't have enough patience, but learned our lesson after buying 3 products in a relatively short time period. As it stands, we have two dev tools and one bug reporting tool. One of these things is not like the others.
In a sense, we have an accidental thesis of acquiring dev tools. Now the question is still, "What is our thesis?".
On our last partners call, we discussed where to go from here. We have three, we're comfortable with them, and are ready to add a fourth. Is the fourth a $1M SBA deal? Or do we get really good at buying these tiny products (not even proper companies)?
On the one hand, buying something for $1M would feel cool. We'd finally have enough cash flow to hire a CEO and at least pay one more person full time. This sounds amazing. But it's also what everyone else is doing. We can't compete on deals against the big fish just yet, we may need to stay crafty for a while. We've also struggled a bit to find a lot of candidates at the $1M mark. There are a ton of cool products to pick up for under $100k. Small, but cool, with huge potential and some cashflow.
The second approach is slightly more nuanced and a shit ton more ops heavy. It would require having a lot of overhead at the fund and charge stuff to the companies. Broadly this is a bad idea, but I think it could work.
The idea would be to acquire a set of complimentary products and perhaps sell the portfolio as a whole with an increased multiple (i.e. we buy 3 products at 3x ARR and sell them for 4x ARR because we have all the operations in place). Short of that, the only alpha mechanism is growth.
We're still 'sleeping on it', or might just YOLO and do both. One extra layer is to add additional investors in the mix and start investing with other people's money alongside our own. The latter part, I'm keenly interested in, since as the fund operator it is another potential revenue source, and of course our own cash is limited. Now how we do that, or when we do that is all TBD.
I have three thoughts on our options from here:
Acquire more (tiny) dev tools
This is the most straightforward for us to execute on. In this scenario:
"What does XO do?" =>"We buy dev tools"
We have the cash to do a few more, and so far we've been able to grow all 3 of them. We'd just keep on keeping on. The downside is that we haven't taken any distributions yet given that there just not that much cash at the end, and we'd rather just re-invest for now. Also doing support sucks. I'm going to go out on a limb and say for all those eager operators saying "I want to take all the support tickets just to get a feel for the business," uhhhh... fuck that. No thanks. I'll read the summaries once a week. We're going to hire a support person stat. I want our devs to get beautiful tickets when something needs engineering's attention.
Our next acquisition would be another dev tool. We'd then cross promote the tool (where relevant) across our portfolio. Every marginal thing we acquire should bring us more developers. We'd be 'allowed' to acquire things like open source projects or newsletters even if they didn't have revenue because we would have a way to monetize that audience.
This is cool. It's straight forward. A part of me loves the idea of having an anti-fragile portfolio of products that all have a common customer. Yes that's slightly naive, but you get the point.
The execution on this would be difficult. We'd have to initially standardize on one, maybe two tech stacks (since we'd have to hire devs). These devs would sit at the fund level. This is where it gets a little funky. It's not that cool to charge the companies for serviecs. I would agree in a VC backed setting but these companies are not really companies, they're more like headless products.
In an ideal world, we could have a few support people servicing many products. We could also have a few devs that build many products. One thesis around where margin may come from is that when acquiring these small tools that really just do one thing, there really isn't this endless roadmap of feature requests. Take screenshot for example. The damn thing takes screenshots programatically. It's a narrow scope, we may not (and haven't had) to spend a shit ton of time (excluding re-building it from scratch of course) building new features. I wouldn't say it's complete but we don't have people banging on the door demanding the tool does more stuff before paying for it.
Another threat vector is the market. Dev tools may not be the right area. It feels right because we're slightly dev heavy on our team, but we may not be right. Existentially, there's pressure above from no-code and pressure below from open-source. I recently came across this list of open source tools. I installed a few, they're pretty good! I'm sure you could find holes in any thesis. Perhaps that's why everyone keeps them fluffy and generic like, 'We believe software will eat the world'. Or, even better, "Developers! Developers! Developers!".
Ultimately though, companies do want support and generally I've found some customers just want someone to call when shit hits the fan. I also do not believe that developers in some form won't be around in 10 years, even if it's only to maintain old systems. People are still employed to keep mainframes online (hello fintech). Poor guys are still writing Cobol. It apparently pays pretty well.
So now what to do with all these dev tools? I think we should make a brand play, as in XO exclusively acquires dev tools, and if you're a dev, you should know about XO. For each tool we acquire, we make XO known and throw our branding everywhere. That way, when we acquire a new tool, it is immediately more valuable because we bought it. This is the only way we get additional alpha besides growth.
This is similar to the FBA rollups blowing up right now (in a good way). Each product these guys acquire is more valuable simply because they bought it. Perhaps there are tools / languages we can standardize on to build and grow these things, shared resources, etc. I've tried (and failed) to execute on this once at a venture studio but I have some thoughts on how to do it better a second time around.
My rough prediction for this is that if we acquire 20-30 of these, there will be some that suck and some that crush it. Each year we can (tax optimally) sell the losers and keep the winners. It may even be possible one particular product warrants raising funds and building a team around. This would be wonderful. Initially i thought that perhaps we would not be subject to power laws give we buy stuff with revenue, but I think I'm wrong.
Roll up a particular kind of dev tool
I've realized I have a particular weakness around conviction. I like to see overwhelming data before committing. Rolling up a particular kind of dev tool is uncomfortable for memainly because not only are you investing in only one particular area (what if the wind changes direction), you're doubling down on it. One of the reasons I wanted to become an investor is to diversify my bets across multiple companies, as opposed to being an 'entrepreneur' and putting all my eggs in one basket.
Alternatively, if you strike lightening (i.e. choosing to rollup the 'right' kind of product), my guess is that you'd see outsized returns from this strategy compared to the diversified approach. Perhaps I'm more risk averse than I thought.
This is standard in real estate. "We're a new fund rolling up yoga studios." "We're a new fund rolling up storage units." This is common in real estate. I have a hard time thinking about this with software. We could, for instance, roll up all the screenshot tools. I don't know how big that market is. We could pick a different market, perhaps more generic, say AI, and buy anything that has an AI component. In some manifestation it would work. We may do it. But at the moment I don't have a clear strategy on what or how.
A variant on this might be to buy a flagship company or an anchor. You buy your anchor and then everything else you acquire should make the anchor company more valuable. I like this one too, but I think it's hard to find an anchor. We may need to stumble upon it in the portfolio approach.
Roll up a different thing / find a new customer segment
I can say these thing because I am one. Selling to devs is kind of a PITA. Or, a more recent comment on sales in general, "Sales suck monkey d*ck" - Elon Musk
Perhaps we'll continue to be flexible for a while, buying opportunistically and seeing what nets out. It is nice sometimes not dealing with developers, but the flip side is that devs are relatively good at telling you when stuff is broken. The average product user will just never convert or churn and never say a word.
I believe a thesis can be fluid. I don't see a need to be so strict. With our own money, this isn't an issue. I could see it harder to tell our story to outside investors though unless we post some outrageous returns, wherein we can start saying helpful things in meetings like "trust me, it's going to work", and then of course actually have it work.
I hope to look back on this post in a few years (you thought these were for you?!) and be right. lol.