Unpack the SAC

Richard Kramer: Welcome to Bubble Trouble, conversations between the economist and author Will Page and me, independent analyst Richard Kramer, where we lay out some inconvenient truths about how financial markets really work. Today we unpack the SAC, that's subscriber acquisition costs, getting to the heart of the issue, and the core of why we find ourselves reverting back to bubble trouble. What's the customer really worth? And how much did you spend to grab them and keep them. More in a moment.

Will Page: This week on Bubble Trouble, we're going to talk about customers. And here's the key point, they don't just hang off trees, they're acquired, but they cost something to capture too. And this week, what's been bugging my brain has been, is this concept of the acquisition cost. The subscriber acquisition cost. The SAC, the S-A-C. What if the latter costs exceed the revenues? What if this has been hidden accounting? Richard, on this show, we try and keep things simple. And in finance things should be kept simple, but it's this temptation to make them complex and the cynic in me says, by making it complicated, you can prevent people from asking the obvious questions. So we have an acronym this week. So break it down into its parts, the subscribers, the acquisitions and the costs.

Richard Kramer: Sure, Will. Well, the subscriber is something we're all familiar with because literally everything from our content streaming services to getting razor blades in the mail has been turned into a subscription. The idea that it's something you need regularly, so someone will provide it, so that you don't have to think about it. And by not thinking about it, you tend not to focus on what it really costs to you. Now the acquisition part is finding those subscribers and we talked before in previous Bubble Trouble episodes about this nebulous idea of the total addressable market.

Well, I suppose for one of those subscription services for shavers, the total addressable market is every post adolescent male in the world, so you could define that pretty widely. And the costs are finding those subscribers out of that huge total addressable market that people will look to. And it's critical to valuations, because it's assumed that subscribers times their growth rate times the churn, which is the number of people who fall out of being subscribers any given month, is a pretty simple way for analysts to calculate the future revenues of a business and subscribers tend to be sticky as a lot of people are lazy, they just don't tend to cancel subscriptions, even if they don't need them anymore.

Or there are certain subscriptions where we don't really have alternatives, like mobile service, you know, you have to pick one of three or four options, but you can't really avoid in the modern world living without a mobile or a broadband connection.

Will Page: Now, you mentioned the word there churn, which is a bit of a spanner in the works of us trying to keep it simple, because now we can go in two directions, you have the acquisition costs, and then you have the churn costs. So if I try and keep this really simple, the SAC, the SAC, the subscriber acquisition cost, sounds like it's your marketing cost, divided by your gross additions. But then you could counter argue and say, it's the net additions, our gross additions less churn. I brought in 100 customers this month, but I lost 10, my net is 90. Now, I might be uneasy with this because should the subscriber acquisition costs be a gross or a net calculation?

Richard Kramer: I think you want to think about it in terms of gross ads, because the things you do to keep subscribers paying are more downstream relating to the service, whereas the things you need to do to entice them to join are often a cost which you have to take up front.

Will Page: Right? And this is the cost of them getting into the shopping mall, once you're inside the shopping mall, let them do their shopping, but it's the subscriber acquisition cost is getting that person to come into your mall in the first place.

Richard Kramer: They can take the form of advertising, promotions, like free trial periods. And oftentimes those are designed to give you a free trial for three months, but they collect your credit card info and the user forgets when they're supposed to start paying. And typically, they're never reminded. There can be upfront costs, like if you get a mobile service package and they give you a subsidized or free handset, or the first one or three months free of a service and then it converts to payment. And there are costs alongside that to advertise installation of the App Store or other ads to entice you into the service. You know, there can be many factors behind churn that people can't afford the service anymore or they lose their jobs or there's an alternative service which is more attractive.

Will Page: So I'm thinking here about organizational structures, you could have two departments, essentially. You could have the subscriber acquisition cost department and the retention department and they're both optimized towards different metrics, which can be a good thing, but it can also create an element of friction, presumably.

Richard Kramer: Sure. And I think what you see in these bubble trouble times is that companies are spending like drunken sailors to acquire customers and acquire subscribers, but they don't pay enough attention to why subscribers leave.

Will Page: That's somebody else's problem.

Richard Kramer: It is indeed, and the upfront cost of acquiring those subscribers is seen as the entry ticket to becoming one of those multi-billion dollar businesses, but no one really pays attention to replenishing the value of the service on an ongoing basis, whether it's refreshing the content, or improving the speeds of your bandwidth that you're getting from your broadband provider, or making sure that you're, if you want to drive through a price increase of your, of your razor blade service that you have to add another mach four, mach five, mach six, blade to the, to the, to the razor so that it, it gets you always a better shave.

Will Page: Um, okay, this is why I'm feeling this podcast, this is a good one. SAC is relevant, because there's two questions before we wrap up part one, which is, when is the SAC too high and when is it unrealistically low? That's a relative question. But it doesn't stop there. Like you can't just simply look at that SAC figure and benchmark it against a similar company, but there's also the absolute context, which is, presumably you get cash burn, when the SAC is higher than the customer lifetime value, that's when the company is on an unsustainable trajectory.

So from a static to dynamic, Richard, unpack both in stages for me. When I, when I look at a simple SAC figure for let's say, a Telco or a Netflix style company, how do I know whether that's too high and too low? And when I look at it dynamically in terms of the company's financials, how do you start to begin to see whether they're actually burning too much cash, should I say, in order to get those customers through the door, get them shopping?

Richard Kramer: Certainly, there are some types of services, for example, utilities, where your likelihood to churn is very low. We tend not to change our high street bank, our electricity provider, our gas provider, all of these utilities. It may be that there's really only one option in our local area, it may be that there are relatively few options, but it's hard to differentiate why one electricity provider is any different than another. So there's some services where the customer lifetime value might be very high, because that propensity to churn is very low. If you look at the world of video content, you had Netflix standing alone five years ago, and now you've got Amazon Prime and Disney+, and Peacock, and Hulu, and dozens of others.

Will Page: Disney got 60 million subscribers in its first six months, I think.

Richard Kramer: Yep. And you have potential substitutability, which raises the likelihood of churn. The more you can find an alternative service and the easier that the service makes it so that you can cancel and restart, the more you have to factor churn into your equation. How do you know whether a subscriber acquisition cost is too high? You have to look at two things, I think, when you look into the future. The way to know whether a subscriber acquisition cost is too high, is to look at the duration of the likely subscription. Is it something you're going to have perpetually because Netflix keeps churning out terrific content, you're always going to want to see something on it? Or is it something where there's likely to be a substitute which crops up and that may diminish what you're willing to pay to acquire a subscriber?

And the second thing I think you need to look at, which is very difficult to calculate properly, is the propensity or willingness of a subscriber to absorb price increases. Now Netflix over its history has very successfully managed to slightly increase the price. There's lots of other subscription services where they struggle to increase the price, because maybe they're not providing something that's as unique or differentiated as some of the Netflix original content. And a lot of the high upfront subscriber acquisition costs can look a lot better down the road if you assume that there can be a 5% or a 10% price increase every year, or every few years to increase the value that you're bringing in against what you paid upfront to acquire that subscriber.

Will Page: You do inspire a thought in my head here, which is the New York Times, would presumably have a subscriber acquisition cost. We know it's hard for newspapers to get subscribers and music streaming services have hundreds of millions of subscribers. There's very few newspapers which have got more than one million subscribers globally, but the New York Times subscription service, you have all sorts of discounts, all sorts of promotions, all sorts of trials.

But if you want to unsubscribe, you have to phone them up and rationalize your reason for unsubscribing, it's not like you press a button and say, I'm out of here, you actually have to call them up and say, I would like to negotiate my own subscription, please. And they're gonna ask you for that phone call, it's an interesting way of managing the SAC and controlling the churn.

Richard Kramer: I'll throw in one other point there, which is the New York Times, uh, and other folks are a classic for offering you the, the easy on ramp. Hey, subscribe today, just 50 cents a week. Subscribe today, your first three months free, and it is an easy way to get you subscribed before they slowly raise the price. The ideal way for any subscription service to get off the ground is to be something that's viral, that everybody is clamoring to get their hands on, and has a relatively low entry price. And none of the services that are being thrown out there today by all of these young internet companies intend to have the same price in the future that they're offering you today, they all intend to exercise some form of pricing power, and raise the prices slowly.

Even though they may know that your mental benchmark is, as our producer Eric has raised in thinking about subscriptions to podcasts, it should be no more than a cup of coffee. So you have this entry price that's got to be very low and then you assume that as people get accustomed to using your product every day, it becomes a part of their daily routine. Then you slowly bring up the temperature, bring up the price in that boiled frog parable that, you know, you throw the frog in boiling water, it jumps out, you, uh, heat the water up around the frog and they don't notice.

Will Page: Okay, and I want to latch on to the viewer, you just said viral there and last week we discussed this quite a lot when we talked about companies that jump the shark, go up like the curve of a shark fin and then fall straight back down. Bringing part one to a conclusion here, is there anything you could say about the viral nature of how to think about SAC?

Richard Kramer: Any service that captures that lightning in a bottle virality has materially lower subscriber acquisition costs because people come to them, they don't have to go and find the people.

Will Page: Beautiful succinct summary of the virality of SA, a price maker or a price taker. We'll be back in part two, to unpack the SAC and put it into action. Back in a moment. We're back to unpack the SAC in the second part of this week's Bubble Trouble. What I'm taking from the first part, Richard, is attribution is stupid, to paraphrase Bill Clinton. What I've learned so far is that the main issue in calculating what caused the subscription, was it word of mouth, promotions, ads, social medias, recommendation, referral fees?

It's hard to get that correlation and prove causation, so somebody in the finance department has to make a decision of this was a subscriber acquisition cost. And guess what? This wasn't subscriber acquisition cost. There's discretion being applied and as an economist, this is where I get worried about accountancy. Accountancy is supposed to be rules, black and white rules. So presumably that personal in finance who draws the rules as to what is and what is not subscriber acquisition costs, might have a line of reasoning that's far removed from reality from what's happening down in the ad sales team or the promotional team, who are out there on the streets trying to bring in customers.

It reminds of a great quote in the book from somebody who was struggling with his tech startups, advertising your marketing team, declaring where costs should be attributed. And she said to me, "You know what? The people in advertising, they're not so bad, because at least when they talk to you, you know they're lying. They're scientific about lying." And it's never a more learned lesson in business told. Advertisers, they're paid to lie. Marketing, they're paid to look like they're not lying.

And I want to really start getting into this attribution thing. This is where it gets fuzzy. This is where it gets really tricky for me. So let's start with just a few sort of quickfire questions here. What can you shove into the SAC that shouldn't actually be there?

Richard Kramer: I would flip that question entirely by saying what companies want to do, especially with investors is to put their best foot forward and define subscriber acquisition costs as narrowly as possible. They want to say this costs us next to nothing to bring all these people onto our platform and by the way, to find customer lifetime value as high as possible by saying, they'll never leave.

Will Page: What you're saying, Richard, here is if those costs hit the gross margin, well, that's fine, because I can improve profitability over time, what I want to see from investors is, I don't have loads of customers, which allows me to improve my profitability over time. It's the sequencing that matters, correct?

Richard Kramer: No, I don't want you to misunderstand that. Typically, a companies cost of goods sold is the cost of providing whatever the service someone is subscribing to. In the case of a razor blade subscription, you've got to provide them razor blades. Uh, and if you get that your part of Dollar Shave Club, their cost of goods sold, their gross margin is the money you pay every month to get their razor blades minus the cost of sending your razor blades.

Now, the marketing department comes in and the sales and marketing costs are those upfront investments in encouraging more and more people to join the Dollar Shave Club. Marketing may get charged with the cost of giving you three free months of cost of goods sold, sending you three free razor packs. They may get charged with the cost of all the digital or out of home or physical advertising they're doing to entice you into the product. They may get charged for the influencers they pay money to tell you what a great service this is, but all of that set subscriber acquisition cost typically falls under the sales and marketing rubric, not in the cost of goods sold.

And that's where the marketing department, like you say, they've honed it to a fine art, to say to the top management, we'll go and find more subscribers, more dogs to eat the dog food, more people to, to come on to our platform, if you give us the budget to go out and seek them out.

Will Page: Yeah, nobody in marketing ever says they can do more of marketing with less budget.

Richard Kramer: No. And that's a natural thing that we see across companies of all sorts, which is you just have to look at what the incentives are. The incentive for the person in the marketing department is to spend all of their budget to see if they can find more subscribers before they go to the top management and then typically in those situations blame somebody else for their failures.

Will Page: Now, in the first part, you mentioned Netflix, I want to return to that as an example because that's a fixed cost content business with a huge subscriber base, but it's a good example of how discretionary the SAC figure is. New programming attracts new signups, House of Cards, without any doubt brought a whole bunch of subscriptions into their platform. So do you add the growth of their programming budget to their marketing budget to calculate the real acquisition costs? And how do you treat those costs over time?

Richard Kramer: No, I, I think a company like Netflix would argue that the programming costs builds a lasting catalog value. And for them, the subscriber acquisition cost are the ads that they do, the promos, deals with smartphone makers, like your Samsung Galaxy S20, get a, uh, a month or three months free of Netflix. Anything that eats into what the customer would otherwise pay, or as some sort of discount that you need to do to attract that customer to the service, that would be subscriber acquisition costs.

The programming costs are effectively their cost of goods sold, because that alongside the delivery of the, the video content via broadband networks and streaming, that is what they're selling a subscription to. And in the case of Netflix, the key metrics are to look at the amortization of those content costs, how many people are still watching Stranger Things, seasons one, two and three, versus what are the costs of coming up with the season four and, and planning future versions. And having to replenish that pool of content constantly, is what prevents the churn, what defends the business. And again, let's separate those two things from-

Will Page: Offensive and defensive.

Richard Kramer: Exactly. The upfront cost of grabbing more subscribers, bringing more people onto the platform is paying customers and the defense you play to make sure those people never leave. Now, Netflix is a funny example, because they make it very easy for you to turn your service off for a month. And by doing so, they've always had the view that, hey, what- if you're going to go away for a month or you don't want to use the service so much, or you don't like our content, we know that we can re-market to you and come back, and our cost of re-acquiring you as a customer is much lower because you're already acquainted with the value proposition we have of the service.

Will Page: And that's a great example there of a metric which a lot of investment analysts fail to understand. And I know this from my time at Spotify is, you can measure churn, you can look at how many people churned off in a period, but what you struggle to measure is, windbags, how many of them came back after two or three months.

So in Southeast Asia and low income markets where credit, debit card penetration was low, we had higher churn because Spotify had, you know, people who didn't have a stable way of paying for a recurring fee model, but we had high levels of windbags, but the investment community only saw the former, they had no way of calculating the latter.

Richard Kramer: I would throw the question back to the companies and I would say from my experience, very few companies give you a really full detailed picture of the gross ads, the net ads, the likely churn, the, the postpaid and prepaid for Telco services. And there are some for more mature industries like Telco that do give you a- appropriate levels of detail. But a lot of internet services are very, shall we say, parsimonious with that data. Sometimes they will argue that we're doing it for competitive reasons.

Other companies like Apple will tell you that, here's the huge number, I think it's over 600 million right now of subscriptions we support via the App Store, what they don't break down for you is whether that subscriptions to their own services, or subscriptions to third party services. So giving a giant number like that sounds impressive, until you realize the next three or four questions you really want to get answers to, to understand the value of that to Apple itself versus to other services they may be reselling on behalf of third parties.

Will Page: Fascinating, really following this one, which is great, because it is not easy, it does need to be broken down into stages. We're going to come to smoke signals now, stuff that, you know, the listener can watch out for when they're perhaps looking to invest in a company or reading a headline about a company that's booming, or a company that's busting.

I want to pass a lighter to you to start the first smoke signal, which is, I want you to explain to me and our listeners, what it means when you sort of rep stock compensation into the SAC figure. I, I, I get stuck with- uh, in researching for this podcast, I started to learn a lot about where you place stop comp, or just keep it simple for me, what do companies do with stock comp that could distort the SAC?

Richard Kramer: Well, again, if a company wants to present a view to the investment community, that it's relatively cheap for them to acquire subscribers, then they're not going to want to load a lot of additional stock comp, but one of the real smoke signals and the first smoke signal I would point to is that companies are spending like drunken sailors to add subscribers, but they're not paying attention to the constant updates to the service, that defense they need to play to prevent churn. And that's where you might see, for example, a correlation between stock compensation because they're trying to get the stock price up based on beating subscriber numbers versus thinking downstream how much they need to do to make sure those subscribers stay subscribed.

Will Page: Front loading the benefits, back ending the costs, essentially.

Richard Kramer: Yep. And, and there are many ways that even if stock compensation vests over a period of time, that people can try to trade against rising stock values, and use options and so forth to try to take some of the money off the table that they can get from inflating, uh, near term subscriber numbers, even if they know that those subscribers have a higher propensity to churn.

Will Page: Okay, give me a second smoke signal that helps us perhaps differentiate companies in this race for subscribers.

Richard Kramer: So ever since this direct to consumer movement has gathered pace, and we all somehow looked at our bank statements or credit card bills and realized we were subscribed to dozens and dozens of services, uh, I know, Will, you're such a music fanatic. I shudder to think how many different music services you're subscribed to.

Will Page: Uh, I love Vinyl services, I'm paying too much for Vinyl these days.

Richard Kramer: Right. And, um, so I think the second smoke signal is not being clear about how you go from the losses of SAC to the longer term payback of customer lifetime value. And what are the assumptions buried there? Is it that, that $10 a month service, you assume is going to rise to 12 and then to 15 over the next four years? Is it that there is no substitute for that, uh, the service you're providing, or are there many substitutes? In which case your ability to hold that $10 subscriber versus what someone else might do becomes more difficult.

So there is this very lazy assumption that once you get a subscriber, it's like the roach motel, you check in and you never check out. But I think that is not the case, in a world where you have lots and lots of options, and you have many new offers being presented to you. And the payback for that discount, the payback of that subscriber acquisition cost can easily have the stilts knocked out from under it by a rival service, or by your own failure to realize pricing power, or by the fact that the consumer gets fatigued of the service and wants something new, which is unfortunately, a, a, a factor for, for any of these services.

Will Page: So this is what I'm taking from part two as we draw this podcast to a close. It's just, again, to reiterate, in front ending the benefits, hey, look how cheap and easy it is for you to acquire all these millions of customers. And next period, I'm going to get billions and trillions of customers, but then back ending and hiding behind the curtain, the downstream costs. What is the cost of retention? What is the likelihood of churn?

You know, if I'm going to pay you two years revenue to give me a customer and you're telling me that that customer is going to stick for five years, well, what happens in year six? And who decided it was five years in the first place? There's discretion about defining the SAC, getting them in, and there's also discretion about how long will I measure before I say they are a lifetime customer, before they're awfully low, then I can start to bury the churn thing.

That brings to a close, uh, this week's Bubble Trouble, uh, where we've unpacked the SAC. In the first half, we are starting to understand the attribution question, what is it that goes into the calculation? How do you deal with correlation causation? Did that marketing budget actually achieve that subscriber or would they have subscribed anyway? And then the second half, we understood how to sequence to SAC, how you front loaded benefits and back ended costs.

And I think it's the combination of those two which stokes up trouble for another day. And if there's just one thing that we've learned on Bubble Troubles, is that stoking up problems for a future date is what causes the bubbles to become troubles. This has been Bubble Trouble with myself, Will Page and Richard Kramer. And we'll be back next week with more bubbles turning into troubles. Back soon.

If you're new to Bubble Trouble, we'd encourage you to follow the podcast wherever you listen. Bubble Trouble is produced by Eric Nuzum and Jesse Baker, at Magnificent Noise. You can learn more at bubbletroublepodcast.com. See you next time.