How Much Will the App Store Contribute to Apple’s Bottom Line?

Monday’s “Applevine” post on Metue summarized some of the current Apple (AAPL) reports and rumors circling the news world.  One of the elements included was a recap, and light review, of newly reported data on how the new iPhone supporting “App Store” is doing. The numbers were impressive and there is clearly great potential but the take here was cautionary; a bias toward pragmatism with predictions. One month seems too slight a sample to use for accurately forecasting revenue growth or impact on EPS.  A few raised flags of dissent. 

One comment speculatively said the store could be a 80 to 85 percent gross margin business. Another said that the store could add as much as ten or twenty cents to quarterly earnings per share. Those numbers weren’t supported. They were just “pie in the sky claims,” but still they are out there and they beg a question: what’s the App Store potentially worth – not qualitatively, not from a behavioral analysis, not from a zealous Apple fan, nor from a detractor - simply by the numbers. If we set aside the opinion that one month of data is too little to be meaningful and use it anyway, if we break out the Graham & Dodd, fire up the spreadsheets, how much of a contribution could the App Store make to Apple’s bottom line if the current levels are annualized? It’s got a great revenue story but how much for earnings?

 

How much might the App Store contribute to earnings per share if the store’s revenue grows to $500m, or passes $1 billion? 

 

 

What might it mean to Apple shareholders on a standalone basis that disregards the store’s greater contribution as a driver of iPhone (and iPod Touch) sales?

 

There are no easy answers, but in this post we’re going to try and set out a framework for looking at it – a way of adding numbers to wild speculation. 

SETTING UP THE MATH
To start, there a few certainties we can throw in the mix: we know Steve Jobs reported 60m downloads and about $30m in revenue for the first full month. We can annualize that easily enough, or project even higher to compensate for more countries getting iPhone “host nation” status in the coming weeks. Other facts: we also know Apple only keeps about 30% of the store’s revenue. From that 30%, Apple handles credit card processing, bandwidth, server management and other expenses related to the service. Last, we know Apple’s Q3 numbers: we know their revenue, gross margin, income, outstanding shares etc. (The complete earnings are here via the 10Q).

To apply it all, we need a model. Because Apple’s a global company and its sales results are influenced by currency fluctuations, cultural patterns, product supply limitations and more, and because the iPhone’s only available in select markets  (and waiting to come online in other countries), starting from scratch would be difficult. There are too many variables. Besides that, the App Store’s performance is going to be influenced by the quality of the products available and their pricing. 

Unlike iTunes, where prices are fixed and music and video are known quantities such that consumer choices are a matter of tastes, the App Store is something else.  Evidenced by among other things, the withdrawn “I Am Rich” application which was supposedly a joke, the App Store offers more variability in both the quality of its offerings (amateur and professionally developed applications), and the price ranges in which they are sold.

A comprehensive model with a chance at accuracy would need to account for many of these factors. There’d be a lot of “if” and “then” elements; a lot of variables. That’s not only a lot of work, it’s probably impossible to render with the existing public data. There’d be too many random guesses. 

Fortunately, there seems to be another way – we can compare iTunes to the App Store. Even though the two services are different, they share two important similarities: 

The first is the revenue sharing ratio with content creators:  
• Most estimates say Apple takes approximately 29-31 cents per song sold on iTunes. The music label and artist split the remaining piece of the 99cent fee. If, for the sake of calculation, we round that out and say Apple draws 30 cents, or 30 percent, that’s the same share they’re taking from the App Store. 

The second factor is costs:
• Both ventures are online commerce services that share some of the same expenses – credit processing, hosting (bandwidth) and server management. The numbers and scale are clearly different, but the relationships among them (cost per sale, margin etc.) may be similar enough to warrant comparison, at least as  place to start. (Some could even try and argue that on a “per unit” basis, the App Store might carry more expense because many of the applications are free which translates to a form of subsidy from Apple – e.g. these applications have costs (bandwidth, hosting) but generate no income to repay them.)

By comparing to iTunes, if shared cost structures and similar earnings to revenue relationships are accepted for the sake of argument, then if we can estimate what impact iTunes had on the current quarter at a per share level, we can also make some educated guesses on the App Store’s future contributions.

CALCULATING THE iTUNES EPS CONTRIBUTION FOR Q3
Some general numbers to start with: Using results from Apple’s Q3 10Q filing, Apple had net sales of $7.4b. The Cost of Sales was reported at $4.86b. Net Income was $1.072b. Gross Margin, which as a percentage is equal to Revenue minus the Cost of Sales, then divided by Revenue, was 34.8%.  

From the data we can also calculate Apple’s Net Margins. They are the reported Net Income (after taxes) divided by Net Sales (=$1.072b/$7.464b). For Q3, Net Margins were 14%. We’ll be able to use that number and equation in a second. 

We’ll also need to know Apple’s outstanding shares. They were 883,738,000 on a basic level or 903,167,000 on a fully diluted level.

The numbers: iTunes revenues are counted in a category Apple reports as “Other music related products and services.” This category includes iTunes Store sales, iPod services, and Apple-branded and third-party iPod accessories. Apple doesn’t break out the differentiation for the different components. That has to be estimated. For the purposes of the model we will by estimating iTunes accounts for almost the entire category, 97.7% to be exact. For the quarter ended June 28th, the category sales were $819million. If iTunes represented 97.7%, it yielded $800.2m in revenue.  

SIDEBAR: EXPLAINING THE ITUNES SHARE
Before moving forward, some will attack the 97.7% number as too high. It may be, but it’s not arbitrary. We derived that number by applying some known sales data to the mix. First, it’s reported that Apple sold approximately 1 billion iTunes songs between February 26, 2008, when 4 billion was reported by the company, and June 19, when the 5 billion threshold was reported. That’s a period of approximately 16 weeks - which means 62.5m songs per week.    (Note: The exact day the numbers were hit is unclear but the report days are being used as if exact for the sake of modeling). 

There were 13 weeks in Q3 (approximately March 30 to June 28) so if growth continued at the same rate, it would equal 812.5m songs sold (62.5 X 13). That is equal to $804.3m in revenue at 99cents a song for the quarter. Factually, that’s too high. The number doesn’t account for video sales and rentals, and adjusted to incorporate video, it would be too high a result for the June quarter’s actual revenue. For Q3, Apple reported $819m for “Other Music” revenue.   The song total and revenue number has to be lower. To guess how much so, we’ve compared the quarterly category grosses between Q2 and Q3.

For the second quarter, Apple reported “other music revenue” of $881m.  For the third quarter, it was $819m.  If I can do the math, that’s a sequential quarterly slowdown of about 7%. If the weekly song total also dropped off by the same 7% rate, it would yield 58.13m songs sold at iTunes per week (from the 62.5m song number used above). For the 3rd quarter, with its 13 weeks, that would mean total song unit sales of 755.6m. At 99cents per song, that’s $748m in revenue. Add to that an arbitrary $52m for video sales and rentals through iTunes and the iTunes store sold about $800m for the quarter. At $800.2m, that’s 97.7% of the reported “Other Music Revenue” category. The remaining revenue of near $19m accounts for Apple branded accessories and other iPod services.  

(Note: $52m for video sales is an arbitrary number but it tracks loosely to Apple’s June announcement that they were seeing 50k downloads a day. Specifically, at an average sales price of $12.50 to accommodate pricing for new and library titles (not rentals), 50k per day over 13 weeks would equal $56.8m. If that is knocked down a little to discount for the share taken by rentals, which are priced from $2.99 to $4.99, $52m would seem to be in a fair range.)

SO WHAT DID APPLE EARN PER SHARE FROM iTUNES?
As noted, it’s estimated Apple takes 29 to 31 cents a song at iTunes and the remainder of the 99 cent price is shared between the artists and labels. If we call it an even 30 cents, or approximately 30%, it matches Apple’s share of the App Store’s revenue. 

Even though that iTunes rate doesn’t necessarily reflect the distribution of income from video sales, it’s a starting point: 30% of the $800.2m we’re guessing iTunes grossed for Q3 is $240.05million. That’s Apple’s take before applying costs.

Nobody knows what those costs are. The variables include credit processing, which might be as much as 3% of each transaction plus a fixed minimum of a few cents, then there is bandwidth (hosting), server costs and more. 

It would be unusual to see a report of expenses at that level of detail from a company, let alone within operating categories/segments like “Other Music,” so we’re left to guesstimate again. Net Margins, which are calculated here as Net Income divided by Net Sales (using Q3 data), seems a reasonable spot.  Admittedly, that choice can be debated given how much of Apple’s income comes from hardware (the hardware businesses having a different cost structure than Internet services), but it’s something tangible to work with. 

With revenues and Net Margins, we have two of the three parts of the equation.  That’s workable to back into earnings. 

For Q3, Net Margins were about 14% (Gross Margins were 34.8%).

With about $800.2m attributed to iTunes, income for Apple’s thirty percent share calculated at a 14% Net Margin would be $33.6m for Q3 (e.g. =$800.2m * .30 * .14). Per share that’s 3.8cents at a basic level, or 3.7 cents fully diluted. (Based on 883.738k and 903.167k shares outstanding.)

Summed up: Using this model, $800.2m in revenue amounts, at iTunes, to a rounded-up 4 cents a share of Apple’s total earnings.

BACK TO THE APP STORE: POTENTIAL EARNINGS?
Monday’s report said Apple drew $30m in revenue from the App Store for the first full month. Annualized that would be $360m. Steve Jobs further predicted that the number could be as high as $500m or even $1 billion down the road. 

Using the exact model we’ve set for iTunes, that is, taking 30% of the draw as Apple’s share, then assuming net margins of 14% as a way of subtracting expenses, the picture of per share income becomes more clear. At the current float of outstanding shares listed above, the numbers lay out as follows:

Revenue	Apple’s 30%	Earnings (14% Net Margin)	EPS Basic	EPS Diluted
$30m	$9m	$1.26m	$.001	$.0014
$100m	$30m	$4.20m	$.005	$.0047
$200m	$60m	$8.40m	$.010	$.0093
$250m	$75m	$10.5m	$.012	$.0116
$360m	$108m	$15.12m	$.017	$.0167
$500m	$150m	$21m	$.024	$.0233
$750m	$225m	$31.5m	$.036	$.0349
$1b	$300m	$42m	$.048	$.0465

 So there it is. By this model, if the App Store’s current sales are annualized to $360m in revenue, it might add about $.017 cents a share. At half a billion in revenue it would be good for about two and half cents. At a billion, just under five cents a share.

And what if margins are higher? Since it’s all wild speculation, to provide a counterpoint, we looked at the result if we more than doubled Net Margins to a unrealistic 30%. (That is a number that would presume operating expenses are way below what seems realistic.) $360m in revenue would translate to about 3.6 cents a share come earnings. At $500m in revenue, it would yield about five cents a share. At a $1b revenue number, EPS would be boosted by about 10 cents.

CAVEAT EMPTOR – THE GUESS FACTOR
Financially, Apple only breaks out some data. As a result, revenue modeling, or cost modeling, at this level of detail invariably relies on guesses. The assumptions can always be challenged.  The basic arithmetic can go astray too (though hopefully it hasn’t). And even where assumptions are accepted and calculations are accurate, there are still lots of different ways the numbers can be applied.

Here, several assumptions have been made. Complex elements have been simplified. There is no doubt, some will be quick to point out shortcomings and challenge them. 

One glaring fact sure to get called out is the fact that iTunes sells more than music. The video business with sales and rentals is growing soundly. Nobody knows its share, nor how Apple splits the revenue with content owners.  Projections here could lean too heavily toward music and be off. 

Other issues that could be called out: using Apple’s reported 50k movies a day number doesn’t distinguish between rentals and sales. We have to guess.  Similarly, using the time lapse from Apple selling 4billion to 5billion songs is inexact too. We don’t know the exact day Apple hit those thresholds and even if we did, growth isn’t a constant.

Last, the applicability of the Net Margin rate to Apple’s content sales could easily be challenged. It surely may not fit. Apple makes its money from hardware and software. Each of those businesses has different expenses and profit margins, and they also differ from the Internet commerce component. Using a Net Margin percentage taken as an average from the entire business can’t possibly reflect the differences in the different divisions. 14% works for the sake of putting something together but it’s a guess that could be off in either direction.

For these shortcomings and more, modeling iTunes EPS contribution, or projecting what the App Store might contribute, is arguably a science project; an abstract hypothesis. That leads back to the original take – it seems with the App Store, pragmatism in predictions is warranted until more information is released.

But if breaking out the numbers is the goal, this at least creates a factual framework for looking at it.

Comments

[David Lentz:]

The Apps Store, like the iTunes Music/Media Store, is not significant in its direct revenue contribution, and is not intended as a profit center. The primary purpose of these "stores" is to give people more reasons to buy iPods, iPhones, and Macs.

How effective these "stores" are toward that end is difficult to say, but it seems to be working.

A better way to treat these "stores" financially is to model them as advertising that generates a small amount of direct revenue instead of a charge against earnings.

2008 Aug 13 09:48 AM

[deagol:]

First things first, your application of the Net Margin formula is flawed. It's 14% of the revenue number, not the gross 30% take. So, your estimates are off by more than a factor of 3.

Second, you did all those iTunes calculations for nothing. If you are treating the 14% Net Margin as a given, you don't need to know anything else. 99% of the article is running around in circles going nowhere. Here's a shorter version of it:

"The EPS contribution from Steve Jobs' estimates of the App Store sales (annualized) would be:"

$360m * .14 / 903m = $.056 (5.6 cents)
$500m * .14 / 903m = $.078 (7.8 cents)
$1b * .14 / 903m = $.155 (15.5 cents)

That's it. That's all this article is about. And it's still worthless, as you point out in the conclusion (applying the same company-wide Net Margin to any/all business segment is just inane). I completely disagree with your final statement that "this at least creates a factual framework for looking at it."

Sorry if this was harsh. Nothing personal, but if I were you I would want to delete this whole article, and maybe leave an apology to my readers, hoping that they understand that we all make mistakes once in a while.

2008 Aug 13 12:01 PM

[Seth:]

To the second comment – if you apply the formula as you’re suggesting it effectively assumes that the revenue sharing distributions (iTunes and App Store) are incorporated into the 14% Net Margin, that is, it reflects Apple’s earnings after sharing revenue and trimming off expenses. We can argue that is flawed:

14% is a number derived from company-wide results and the bulk of Apple’s revenue is generated from product sales that don’t have a comparable rev-share component. It’s not like they’re sharing 60% of MacBook or iPod product revenue with partners.

I chose to apply the margin rate to just Apple’s share (30% of Net Sales) because it yields a more conservative result which takes the revenue sharing component explicitly into account. Is it right? Maybe, maybe not. It’s acknowledged in the article that “using a Net Margin percentage taken as an average from the entire business can’t possibly reflect the differences in the different divisions. 14% works for the sake of putting something together but it’s a guess that could be off in either direction.”

As for the value of including the iTunes calculations - the original title of the article on my own site was “Modeling iTunes AND App Store EPS potential.”

2008 Aug 13 03:28 PM

[deagol:]

"It’s not like they’re sharing 60% of MacBook or iPod product revenue with partners."

In fact, this is exactly the case. That's what the Gross Margin is. Apple is indeed sharing (i.e. paying) an average of about 65% (obviously it varies by product) of their revenue with their suppliers and contract manufacturers. The costs of running the business are not included there yet, expenses like running the retail stores, the online store, customer support, R&D, advertising, administrative costs, etc. It's clear to me that hosting and credit card fees fit the definition of Operational Expenses, and the 70% they pay to content creators (music or apps) fit the definition of "cost of sales". So after taking out these costs and expenses, and after paying out taxes you get your Net Margin. Once again, this Net Margin of 14% is not only out of the portion of revenue they get to keep after paying off suppliers and contract manufacturers of MacBooks or iPods or music or apps (35% overall), it's 14% of the whole thing.

I will quote from your own article and use your own numbers:

"With about $800.2m [in revenue] attributed to iTunes [...] Net Margin would be $33.6m for Q3"

Well, Net Margin of 33.6m on 800.2m revenue would mean iTunes Net Margin is 33.6 / 800.2 = 4.2%. What happened to the assumption of 14% Net Margin? Furthermore, that means the Net Margin for the rest of the business is close to 16%. So, you're saying "for the sake of putting something together, let's use the same company-wide Net Margin for the iTunes store" but actually end up with iTunes enjoying only ONE FOURTH of the company-wide Net Margin. Insisting that "using a Net Margin percentage taken as an average from the entire business [...] could be off in either direction" when the actual number hidden in the math is considerably off in one direction (down) is clearly contradictory.

To me that seems quite a bit disingenuous, or a mistake. Why not just say "I will grant iTunes a much lower Net Margin than the rest of the business, about one fourth, just to be conservative"? Wouldn't that be much more frank and clear?

2008 Aug 13 08:40 PM

[steveballmer:]

Negative numbers!

2008 Aug 13 09:30 PM

[Neil Anderson:]

More apps at the App Store equal more iPhones and touches sold.

2008 Aug 13 10:55 PM

[deagol:]

"the original title of the article on my own site was “Modeling iTunes AND App Store EPS potential.”"

Yes, but it reads as if you needed the iTunes exercise to derive some numbers which you would then apply to the App Store model, which is not the case. Again, I will quote from the article:

"A comprehensive model with a chance at accuracy would need to account for many of these factors. There’d be a lot of “if” and “then” elements; a lot of variables. That’s not only a lot of work, it’s probably impossible to render with the existing public data. There’d be too many random guesses.

Fortunately, there seems to be another way – we can compare iTunes to the App Store."

The fact is that the iTunes model is not at all required to derive your App Store model. You just used the same model on both segments. What you ended up doing is exactly what you described as "probably impossible to render with the existing public data" and then applied it to both the iTunes Store and the App Store. Those "too many random guesses" are all still there, and on top of that, you piled on the additional “if-then” that both stores share "similar earnings to revenue relationships." You're adding an additional arbitrary constraint to what should be two different models (that they are similar) and argue that doing that somehow makes your theory more valid. It's actually the opposite.

The acrobatics you resorted to when coming up with a Net Margin guesstimate for iTunes in no way validate anything about the App Store model, and assuming the models are similar doesn't help either. They are both guesstimates, far from facts, and the assertion that this exercise "creates a factual framework" is wrong.

Had you said "in my opinion, even if we assume the App Store could become a one billion enterprise at some point, as Steve Jobs said, it still would not be a significant contributor to EPS" and then wrote an article about why you believed that (say you quoted Steve Jobs or Peter Op. when they've said they plan to run the App Store just like the iTunes Store "at a slight profit" or quoted other analysts who've talked about this as a "loss-leader" strategy, or anything BUT this whole "doing the maths" farce), then I would have thought your reasoned opinion was valuable, and in fact I would have agreed for the most part. If you then pegged the annual EPS contribution of the one billion dollar App Store at a 4.8 cents through some quick back-of-the-envelope calculation, I would probably disagree to the precision implied in that result (a range of 5 to 25 cents would be my guess), but I wouldn't have any problems seeing it as what it was: just a guesstimate, and not pretending to be something it's not.

But I guess when you throw a lot of numbers and math formulas and complicated models, a lot of people believe it must be true, or more accurate, or "a better factual framework." It really hides all the subjectivity behind all the math, doesn't it? And I don't hold this against you, it's just a fact of how people perceive things. I've also developed my own model for Apple, and it's very complicated and I don't expect many to understand much of it. But I still share it with people and I always insist that it's not one bit less subjective just because it uses a lot of math instead of pulling an EPS number out of thin air, and there's always some who think that it must be more factual because of the math or because its historic accuracy. I see you're acknowledging some of those disclaimers here, but there remain some inconsistencies and logical fallacies being passed as reasonable analysis.

Sorry for the awfully long comments and having wasted even more time on this.

2008 Aug 14 03:34 AM

[mollytjm:]

the app store will drive iphone sales sky high...not just here...but all over the world. and people who see the apps on phones will realize the potential for a lot more use than they thought. people are just starting to realize that 'it's not just a phone, folks!'.

this is a stock investment site, so i think that the growth potential of this company and profits is what's important. the numbers change every minute. what's important here is that Apple is going to soar, and with it's $$ reserve, no debt, innovation and multiple products and revenue streams, it's a great long term investment.

2008 Aug 14 10:49 AM