Keep It Simple, Stupid

Part 2: So You Decided to Buy a Social Media Company

(Click here to check out a pre-built model of this scenario right now.)

Start Simple & Add Complexity as Needed

As mentioned in Part 1, accurately stripping out the complexity from something is hard, and generally requires real talent. What’s a lot easier is to start with something that is a complete answer to the problem, but is too simple. Then, you just add necessarily complexity until your answer is just complex enough, and you’re done. 

What does this look like? Let’s go back to the Twitter story for some examples of where complexity can be (or not be) important to consider. 

As mentioned, while Twitter’s financial problem has a lot of variables in it, it’s one of those things that can be explained pretty simply at a high level.

  • Twitter is currently unprofitable. Over the last decade, it’s made money some years, and lost money other years, but it’s definitely lost money for the last two.
  • For a sense of scale, consider that it’s never made or lost more than $2 billion (profit/loss) since 2012. So a billion dollars is not an incomprehensible amount of money at Twitter,  but it’s a lot.
  • The business is definitely continuing to grow, even when it’s not profitable. In 2021, it reached $5 billion in revenue, steadily climbing (it was about $2 billion in 2015).
  • In order to buy it, Elon Musk borrowed a bunch of money against Twitter itself. So now that he’s in charge, it has to pay about $1 billion a year to service that debt — this is in addition to whatever its expenses are.

Those are the facts. So basically, Musk needs Twitter to find a billion dollars a year to continue owning Twitter, and as Twitter has been constituted historically, that’s a large amount of money. He could also continue to liquidate his assets and throw cash into this flaming pit, but even he can’t do that forever, and all likelihood he really, really doesn’t want to do that at all. That means he really does needs to find a billion recurring dollars from inside Twitter’s business. 

At the simplest level, you can simply sort any idea for doing that into two categories — ideas that will bring in additional profit, and ideas that will reduce expenses. 

This isn’t exactly an “actionable” plan, but it’s undeniably true. Maybe some external factor will lower the amount of money Twitter will have to find. More likely, something will increase that amount. Either way, the relationship stays the same — whatever that amount is needs to equal the combination of loss averted, and new gains realized. And, on the off chance that you do something that reduces cost, but in doing so reduces profit even more, you actually end up with a larger hole to fill.

But again, this isn’t actionable. We need to dig into some more complexity here, but now we can do it without losing sight of the fundamental issue.

Savin' Money

Let’s start with what’s already happened; Twitter got rid of a ton of employees. This project — laying off half of a company — is FULL of complexity, and while it all “matters”, not all of it is super relevant to this specific strategic conversation. For instance, when you’re calculating the financial impact of a huge layoff, severance is a pretty big factor. But does severance really matter in the context of trying to stabilize the company’s long-term financials? Honestly, no. We’re trying to get the strategic budget balanced; Twitter can debate exactly how much money to eat in the moment as a result of its tactics in a different meeting.

In short, severance is complexity we can/should ignore (it’s complicated and irrelevant to the larger math problem). However, the functions we cut, the money that saves, and the potential impact of those cuts to company performance are complexity we need to acknowledge. Otherwise, you could just say “fire everyone!” and be done with it, assuming you were a total sociopath.

If we add just that immediately useful set of complexity, things look like this:

There’s still a lot of hand-waving here — are these the right buckets? What’s “Operations”? — but you can already tell this is relevant complexity because it is highlighting some very real, and very impactful/dangerous consequences. But as someone who has been given a broad “cut the marketing budget” mandate before, let’s go down that particular rabbit hole a bit more.

Marketing — like all things — has a couple pieces to it. So it makes sense to start to identify some of those pieces, as well as potential impacts (aka, the reason we have those things).

Now, there’s a LOT here that is oversimplified to the point of being wrong. Sometimes that’s just because we don’t work at Twitter and have, for instance, no idea how much advertising they spend (if any!) to acquire a user. However, that is a really good number for someone who does work at Twitter to have, and I guarantee someone who either works there or used to work there has either a valid number for that, or a useful explanation about how we should think about this number (or whether we should ignore it). 

But that discussion is the point! You can draw a straight line (or a slightly squiggly one) between the thing you need, and the thing you have. In this case, breaking down the impact of a giant Marketing budget forces you to think about what that money is for. If the answer is really vague (“it helps Sales close deals”), sometimes that’s just how things work even in a data heavy world — people just don’t know or can’t directly connect a dollar out to a dollar (or a user, or client, or a cancellation, etc.) in. But it’s a useful exercise to start to get ballpark impacts and not just say that cutting a position means “we can’t do X anymore”. If you’re a wild-eyed cost-cutting billionaire on the debt clock, hearing “we spend money on marketing to get the users you want to monetize” or “if we don’t sponsor these two fancy events, we are going to lose 5 major brand advertisers” is going to give you pause even if “cutting half the department overnight is stupid and insane” does not.  

And of course, to our earlier point, the big question here is “how deep do we go?” You may find that breaking down exactly what constitutes a sponsorship, and what the different types are, is worth doing because without unpacking that you’re looking at either spending $55m or losing nearly 3 million new users. Conversely, you may decide that fifty million dollars (or those users) is/are a rounding error and you’re either fine with it, or willing to simply cut something else. But when you add complexity in context instead of just having random arguments about random things, it’s a lot easier to constructively say “this doesn’t matter” because the numbers are right there in front of you. 

Obviously you’d continue this exercise across all the other “cost reduction” areas outlined in our original model. In each scenario, you’d whittle down the logic to the simplest relationship between money and output that still makes sense. Maybe that simplest version is actually pretty complicated, or maybe it’s not. Either way, you’ll always know whether it’s worth the time, focus, and energy to drill down another level, something a lot of people in meetings and planning sessions aren’t aware of. In the context of cost-cutting, that complexity is likely to lead you to some pretty rough conversations, even if the numbers and rates remain debatable.

However, here’s a spoiler alert — Twitter isn’t going to survive just by cutting costs (although it might die). So let’s go to the other side of the ledger and work through some of these big new product ideas. 

Makin' Money

In addition to firing thousands of people who may or may not be absolutely essential to Twitter continuing to function, Musk has also promised to unlock the “incredible potential” of the platform and make a bunch of money from new things. One big (and very simple) ideas is simply to get regular users to pay $8 a month to use the service with additional benefits. Another was for Twitter to become some kind of payment platform, presumably with the company taking a cut of those payments like a credit card. These both sound very open-ended (because they are), but you can use some pretty basic, connected math to get your arms wrapped around them and see their realistic, near-term potential. 

If you’ve ever built a funnel before, you’ll recognize what we’re about to do here. The trick is that we’re going to go “backwards”, and trace this magic subscription and payment revenue all the way down to what Twitter actually has — Twitter users.

Step one… what causes subscription revenue? Subscribers! After that, you can work through all the other steps between paying subscribers, and today’s user base.

Steps? Well, basically there are a bunch of requirements for any given user here, today, to start making money for Twitter via a subscription.

  • they have to remain a user (conversely, Twitter might actually add new users just because they offer a subscription plan, but that seems counter-intuitive and ultimately pretty unlikely, especially since they have one already that nobody uses)
  • they have to be the kind of user who actually posts content, since the benefits of the subscription are targeted at people who post a lot 
  • they have to decide that, as someone who posts a lot, the benefits of Twitter Blue are worth actually paying for
  • As you go up through those requirements, more and more of Twitter’s current user base drop off. All of these numbers are speculation, but what’s not speculation is that by definition, these things will happen at some rate, and any plan assumes some number for those rates.

To add just a tiny bit of subject matter expertise, 2 to 10 percent conversion rates from free internet things to paid internet things are actually quite good. Obviously those rates could be a lot higher, or a lot lower, based on the particulars. But if you start with industry standard rates as your initial guesses, you can at least say things like “Twitter Blue would have to outperform basically every other freemium consumer internet service to make more than $100 million in revenue”, and that’s something.

You can do this same logic with different variables and work backwards from “magical payment platform revenue” to pretty quickly get something significantly more concrete.

It’s a totally different business, but a fairly similar process. In this case:

  • current users turn into users who give Twitter a credit card number
  • … those users turn into some amount of payments, which are worth some amount of money
  • Twitter takes a cut of those payments

The actual numbers plugged in to represent these factors can and should be debated. But the factors, and some of the base numbers… they’re hard to ignore. And just by building out this relatively basic funnel, you can tell that this idea is going to live or die based on getting users to connect cards, and getting them to make a lot of transactions through Twitter. One or two a month isn’t going to do it. 

Subscriptions + Payments + (lots of money) = Success!

Twitter is (at of this writing) a $5 billion dollar business with (again, at the moment) thousands of employees. It’s complicated, and I bet any discussion that broadly starts out with “how can we fix Twitter”, or “2023 Product Strategy”, or “does anyone have an extra billion dollars lying around” is destined to be broad and frustrating and go down a bunch of irrelevant tangents.

But to this large, specific, existential question — Twitter is not going to become a sustainable business simply because people start paying 8 dollars a month for it. The obvious counter-argument would be “but it won’t just be today’s Twitter! It will be better!” But with this logic in front of your face, “it will be better” suddenly isn’t good enough. You have to start making an argument that specific changes to Twitter are going to make people three or four times more likely to pay for it than they are to pay for other internet products.

"Doesn't anybody SEE this?"

Now, in the very visible case of Twitter, plenty of thoughtful people have put two and two together and explained the logic here. My favorite examples are people who started spitballing funnels like the ones mentioned earlier.

But for every clear-eyed analysis of the facts on the ground like this, we’re subject to endless updates and debates about whether various changes — some huge, some very small — are “good”, or will “work” or not. If you want to debate whether Twitter can become the most accurate source of information in the world, you’re going to have an extremely complicated debate with a lot of issues in it, and potential problems to solve. But none of that complexity matters, because even if you figure out that problem the numbers we just went through tell you that it’s almost impossible for that to solve your existential financial crisis.

Somebody who did a better job prioritizing complexity when looking at this challenge would have come to many of the following conclusions:

  • Advertising is a massive factor towards Twitter staying solvent, and you better not screw with that until you are printing money some other way 
  • Cost reductions are probably necessary but keeping Twitter sane is a labor intensive process so you better have some ideas to make that easier and more affordable 
  • Eight dollars a month is too high for casual users and probably not enough to take advantage of the very small group of Twitter people who actually use Twitter to make money themselves
  • If you want to make money off Twitter as a payment platform, you need to buy yourself some time 
  • Most importantly, if you are buying Twitter as a personal toy to fiddle with, doing so via a leveraged buyout with massive debt service is going to effectively eliminate your ability to play with it

Many things in life — landing rockets, building healthy social media platforms — are complicated. But many of the decisions around those things aren’t complicated at all. When you make them complicated, you might make yourself look, sound, or feel a little smarter, but odds are the only thing you’re really doing is increasing the chances of doing something dumb.

Let that sink in. 

(Cool, huh? Click here to check out a pre-built model of this scenario right now. Make a copy and build your own!)