Operational Leverage

Fellow Collegiant Ramster doubts Baruch, and writes:

So if the iPhone is such a gamble for Apple, what’s the alternative? Stay out of the phone business entirely?

Ramster also questions whether Apple really needs to go mass market. There’ll be enough fanboys out there, he opines, to create

 . . . a pretty decent sized niche…which leads to the question of scale that you mention. It strikes me that when a few hundred million handsets are being sold per year, a niche market can still hit numbers that would be considered mass market in any other space (e.g. 10s of millions of units/yr).

Well, Ramster, for your first question, there IS no alternative to iPhone for Apple, is there? Other than sitting on the iPod franchise and milking it for cash and managing its decline. But of course that would be a bit of a multiple shrinker.

In that sense I guess is not strictly speaking a gamble, or at least one with any optionality. It just proves that Apple may not have been in quite as strong a position as we thought.

But your second question goes to the heart of the Apple thesis I propound, and is why answering your question is deserving of a post by itself. I fear there is much misunderstanding here.

Let’s do a thought experiment. Let’s say Apple flogs “10s of millions of units/year”: 20m iPhones at USD500 a pop wholesale, which is USD 2 bn in revenues. Say Apple makes a 50% GM (I’m being generous), but has to pay 500m in R&D and 300m in SG&A (these totals may not be enough, they are just illustrative). That means Apple makes USD 200m in EBIT, a 10% margin for Apple. Oh dear. Wailing and gnashing of teeth.

But just sell 5m more units and you make 450m in EBIT, a 22.5% margin. Got that? 20% more units gives you over 100% your profit. Once you cover your cost it drops straight through to the bottom line. Or you can keep the same margin and spend more in R&D or have a cooler ad campaign or build a distribution network in Mongolia or wherever. This is what we call operational leverage. It comes from scale.

Without scale, other people with scale can come in and eat your lunch. Let’s ignore SG&A, even though it’s equally important, and just concentrate on R&D instead. Nokia spends 12 BILLION fricking dollars on R&D per year. And still has a 20% margin. If its R&D was only half as efficient they could outspend Apple’s illustrative business model above by diverting just 10% of that. Not straight away, but over 6 months, over 12 months, you’d see Nokia start to have better features than Apple, in iPhone-like models that are released before Apple’s. Moreover, Nokia would start to offer more variants to fit different tastes and budgets, and extend the business into new customer segments.

Soon Nokia will be selling more units and making more money, free EBIT to spend more on R&D. The advantage thus becomes self supporting, a virtuous circle for Nokia, a more vicious one for Apple. Eventually they have to drop out of the handset market. This is not a fanciful example. This is exactly what Nokia did to Panasonic, Ericsson, Sony, BenQ, Siemens, and Motorola. It’s why they have a 45%+ share, 3x more than anyone else.

And that’s why Apple needs to go mass market.


20 thoughts on “Operational Leverage”

  1. Or, you just release an SDK and get everybody else to do your R&D spending for you…

    More to the point, how much R&D money do you think went into the iPhone? How much R&D money has Nokia spent on trying to do web browsing on a phone? You’re confusing inputs with outputs…

  2. Many thanks for honoring my comment with an elevation to post-hood. However, I have much disagreement. It felt like you’re overstating both R&D and SG&A numbers so I decided to take a look at their Q1 report. Nokia is actually two companies: a handset company and a network infrastructure company (Nokia Siemens Networks). Helpfully, their report breaks out the financials for these two groups. When comparing Nokia to Apple, we should consider their handset division (which their report calls Devices and Services).

    Out of their D&S Q1 revenue of 9263 million Euros, their R&D and SG&A costs are 766 and 806 million Euros respectively (BTW, Nokia’s total R&D was 1.375 billion in Q1, which is about 2.1 billion USD…which is a fair bit less than the 3 billion that your 12 billion annual figure would give).

    I agree that your premise that the R&D represents a fixed expense that gets reduced in its impact on EBIT as sales increase (the argument for scale). However, I think that Apple can pull off the iPhone with much less R&D than you suggest. The reason is that Apple is not a wireless company, despite the fact that it makes a wireless handset (RIM’s the same). Many people make this mistake that RIMs a wireless technology company. It sure looks that way but it’s actually a consumer electronics, networking and (most importantly) a services company. The fact that their handsets happen to have a wireless chipset is an enabler for their entire business strategy but actually has modest impact on their R&D. This is because RIM and Apple both buy their wireless chipsets from real wireless companies (ones that do real R&D). They don’t do any wireless R&D when picking a 3G (or WiFi, Edge, GSM, 1xEvDO or whatever) chipset vendor. They just do a vendor evaluation…which is much cheaper than real R&D.

    Nokia, OTOH, is a real, hardcore, wireless company. They do pure research, participate in standards bodies, and design and develop handset and infrastructure equipment. They have a huge handset range, both in terms of device type (smartphones to ultra-cheapies) and air interface technology (GSM, CDMA, 3G). They’re at the forefront of the two competing 4G technologies, LTE and WiMAX. That’s expensive stuff and that’s where a big chunk of the group’s R&D spending goes. If you look at Nokia Siemens Networks, they spend almost as much on R&D as the Devices group (607 million Euros) while getting a bit over a third of the revenue. That’s brutal and that’s why it sucks being a telecom infrastructure company. Their Devices division’s R&D also reflects their hardcore wireless nature. They almost certainly make their own handset modem chips. And as I said, they make a ridiculously diverse range of handsets. So what you really need to compare is Nokia’s smartphone R&D spending with Apple’s. My guess is that it’s in the range of 1 or 2 billion USD annually).

    At the end of the day, Nokia’s trying to do a LOT more than Apple. Apple is focusing on small line of smartphones where it can bring laserlike focus. The other factor that affects your EBIT number is the price premium that Apple will be able to command relative to Nokia (or RIM, HTC, etc). The world of Macs and iPods shows that Apple can command a decent price premium compared to the alternatives (huge in laptops, modest in iPods). A 10% price premium relative to competing smartphones is modest enough to minimize sticker shock while still boosting EBIT. In your own example, $550 iPhones (instead of $500) doubles EBIT to 20%.

  3. Felix, don’t confuse apps with device R&D. Some guy providing Donkey Kong for iPhone is a very different type of spend than HW R&D (e.g. integrating a 5mpxl camera with flash into an OS), and overall OS R&D. These are huge, collaborative projects. Both RIMM and Symbian have released SDKs ages ago, by the way, and loads of 3rd party apps have been written for these platforms already. That no-one knows this may not be good news for AAPL.

    Ramster, you are right. Nokia has to spread its R&D over lots of projects, wireless equipment among them. You are right this is not quite half the overall budget. I also got the FX rate wrong, getting confused between GBP and EUR.

    Sadly Q1 sees the weakest R&D spend in the year at Nokia, so extrapolating an annual figure from that no work. I reckon they’ll spend about EUR5.5bn in R&D in 2008. 2.3bn of that on NSN (ie non-handset)

    I would very wildly guesstimate that Apple may have spent about USD200-300m R&D on the iPhone (but probably more), cumulative over a couple of years, starting from scratch. Remember, though, they only have 1 model on sale at any time. HTC has 6-7. Nokia has maybe 50. Apple presumably will have a bunch of models on sale in 2009, and that will expand the budget.

    A decent browsing experience has not really been a major focus at Nokia. Squeezing cost out of the low end, putting GPS in the handsets, catching up on thin, MP3 and mega-camera trends have been bigger priorities. Probably now, a decent browsing experience and touch is something like a Manhattan project for them.

    Basically, Ramster, your points are good. But the essential truth remains, which is scale matters. Had the USSR found out that only half of its nuclear arsenal actually worked, it would still have been able to blow us all to smithereens with a small fraction of their resources.

    Similarly, when Nokia diverts 10% of the R&D of each of its existing projects into the “get Apple” project, (say EUR350m to add to the EUR1.2bn they spend on “multimedia computers” (old acctng format) they will outspend
    Apple’s entire R&D budget of USD1bn by a factor of well over 2x. Got that? Say iPhone is 50% of Apple’s R&D budget, Nokia could spend 4x as much to whack Apple on the head in iPhone as Apple can spend on iPhone.

  4. Scrolling.

    Why is the iPod still so dominant? It has a patented scrolling system which no one else has managed to approach in ease and intuitiveness.

    And now, astonishingly, with its finger drag and flick stystem, it’s come up with something even better. Which, again, it’s patented.

    So never mind R&D, Apple has IP which is much more valuable.

  5. Felix’s comment is germane though I disagree with its IP emphasis. What’s so special about the iPod, and other Apple device UIs ? At the risk of lapsing into fanboy-dom (did I mention that I find Steve Jobs loathsome?), Apple gets it and executes beautifully on delivering a great user experience. And contra Felix, their success doesn’t depend on IP. Patent protection is helpful but even if it didn’t exist, their competition would have a tough time replicating the user experience that Apple regularly delivers. Why? It’s in their DNA and that’s a tough thing to replicate, no matter how much $$$ you throw at the problem.

    This gets to my main issue with baruch’s argument: the implied determinism of the efficacy of R&D spending. Spending twice as much (or even 10 times as much) on R&D doesn’t necessarily yield correspondingly better outcomes. Consider the iPod example. How much R&D did Sony spend trying to match Apple? And look at their results (SonicStage or ATRAC anyone?). They suck and Apple killed them in the music player market. The dubious marginal value of additional R&D spending is particularly true in the software space. And the software side will have a huge impact on the market success of smartphones. Again, the Apple advantage here doesn’t have to move mountains. It just has to justify a modest price premium (10 or 20%) that will directly boost EBIT compared to the competition.

    How’s this for a counter-example to your scale argument: GMs clearly a helluva lot bigger than Honda and has much higher R&D spending. Who’s shares would you rather own? (I realize that this is unfair and that it’s a slander on Nokia, a fantastic company, to compare them to a POS like GM).

  6. Felix, check out the HTC Diamond phone for touch, scrolling and drag and Apple OS ripoff in handset UI. I played with one last week and it was a bit like my ipod (which i don’t use anymore). IP, schmi schpee.

    Particle, no it isn’t.

    Rhamster, you argument is a fine one, clearly some R&D dollars are more efficient than others. But it’s sort of built into my argument –you are also right that Nokia is a fantastic company, where GM was a near total monopoly in a captive domestic market. Also the oil shock may have helped get Honda (Toyota is an equal or a better case) get a foothold, which is another nail in the coffin your example deserves to lie in.

    As for your Sony example, iPods are a completely different case, a new category which Apple eneterd at the same time as others and reinvented. Mobile phones and PDA-smartphones have been around for a long time.

    I can almost guarantee you Apple will trade at a price premium anyway vs a similarly specced Nokia.

    Your best point, the one I have the hardest time defeating, is about the SW. There I have to grant you an advantage, and there Nokia and HTC and Samsung and Sony Ericsson and LG and the rest will have to work really hard, and 1 or other of them won’t make it. That’s what’s given Apple its “in” in handsets, and that’s why they have a chance now.

    My only issue is that too too many people think it is a done deal already, and we are only at the start of the battle.

  7. Baruch, doesn’t your “virtuous circle” argument also apply to the desktop OS market? In which case, why hasn’t Apple been crushed by Microsoft yet?

    You’re betting on Nokia, Sony, & Samsung, all hardware companies, to defeat Apple in the software arena?

  8. Dr No: The virtuous cycle argument is an interesting one, however I’ve only ever seen it applied in circumstances where volume begets volume, and choice is effectively eliminated. Think google search – once people began using it, it pays to pander to google search parameters, which leads to more people using google search, which leads to more pandering – there is no benefit in using anything other than google search except in niche areas. Another example are online classifieds – once a site attains a critical mass, then it will attract more visitors, which in turn attracts more viewers.

    The difference in the desktop market is that buying an iMac isn’t necessarily disadvantaging yourself – not only is leopard software exceptional, but with the addition of bootcamp you can run all the windows versions of the software if you wish anyway. I would agree with you that 5-10 years ago it was a virtuous cycle, hence the dominance of the windows OS.

    Virtuous cycles don’t allow choice for the rational customer; they necessitate.

  9. Doctor, actually right now I would be getting long AAPL, or at least thinking about it. Note though that Nokia does actually have quite decent SW expertise; it’s bought enough companies in the space. HTC is pretty good too.

    trick is doing HW and SW together with scale. Some of these guys have HW edges. Some have scale. I said and still think that AAPL has a SW edge.

    Some say MSFT actually SAVED Apple’s arse in DT OS in the mid 1990s. That’s really why AAPL didn’t get crushed by MSFT OS — the antitrust case.

    Oz, I have seen HW scale come through in a virtuous cycle in loads of industries. PNDs, Dongles, iPods, phones, NBs, NW equipment (CSCO, Alcatel, Ericsson) — wherever you have a large dominant player or duopoly in an tech HW industry, that’s pretty much how they got there.

  10. As a professional mobile software developer I think a lot of people miss the point with the iPhone.
    It is not really a phone. It is a decent mobile computer that happens to do phone stuff (which is why Safari was a relatively easy port and why Nokia still hasn’t achieved it properly and is having to license and extend its own version of webkit).

    Apple has provided something that is missing from the mobile space – which is a decent OS platform/SDK combination for third-party developers to use. A similar state change that Microsoft initially provided on the PC.

    JavaME on the handset is poor and so fragmented that it is a nightmare to develop anything useful to run on more than one phone model. Symbian is similarly a poor development platform. Current APIs also suffer badly from not only vendor lockdown but also further carrier control of the firmware versions available so that the same models on different carriers are restricted to different firmware versions.

    This is why there are so few decent mobile apps. The manufacturers (Nokia included) still think of things in terms of phone hardware, as do the carriers. This lack of experience in the wider PC software ecosystem has severely limited both their ambition and innovation, which both Apple and Google (with Android) have identified.

    Apple has seen that a multi-platform strategy that targets desktop,mobile and the consumer space simultaneously is the way forward as data and SW become seamlessly portable beween environments in a way that isn’t easy with the PC/Phone separation we currently have.

    Uptake of the next set of mobile devices will be driven by software not yet more powerful cameras, and in this Nokia and the others have a lot to learn from Apple and Google, as their own offerings to third party developers are just not strong enough to provide a reasonable user experience.

    This is something that RIM has capitalised heavily on – but I think a number of competing Exchange integration applications on the iPhone and Android will remove the main driver for the Blackberry’s success, as has happened previously when generalised software capable devices become prevalent in a previously limited function device market

  11. You see, Bento, we get a better class of commentor on UB. Even I, genius that I am, am learning something from this exchange.

    Steve, what about Linux? I hear that Nokia is dumping Symbian, or at least thinking about it, and getting all Linux-y as a way of getting some more of the software edge that everyone says is the next frontier in mobile phones.

    But Steve, let’s face it, iPhone IS a phone. It’s sold by phone companies with voice packages, subsidised like a phone, and they even call it a “phone”. Most people are going to spend most of the time they spend on it using it as a phone, jabbering away in trains saying loudly “I’m on a train!” It has some exciting bits that mean it is better than other phones at non-core tasks like web browsing. Phones have been about HW scale for ages. That may be about to change, maybe we are in a transition phase.

    But if it does become all about the SW, won’t the HW eventually become commoditised? And then, won’t the winner in a commoditised HW space be 1) the player with the cost advantage, and 2) the one with the most open-source or freely available SW stuff, and the most apps?

    Apple may not do well in that environment. It didn’t in PCs. Would a Nokia-Linux combo be the best horse to bet on?

  12. Part of the problem with this analysis is that we’re in uncharted waters and it’s just not clear what it takes to be successful in this space. Smartphones will become the norm over the next 5 years in the developed world and 10 years everywhere else. The truth is, everyone wants a great mobile browsing experiencing, as long as it doesn’t cost them much. And if they can get a media player with tons of storage, a camera and GPS in a nice small, yet usable form factor, why not! (as long as the marginal cost of all those things is reasonably close to zero). That’s the looming future mass market of smartphones.

    On the hardware side, we just have to wait for Moore’s law to kick in, in both handsets and even the wireless infrastructure (these days, 2/3 the cost of a wireless base-station is signal processing and other computing hardware). So we’re going to have a world of cheap hardware and huge wireless network capacity (from HSPA, 1xEVDO, LTE and mobile WiMAX). This is the landscape on which all these players are fighting. Given this, it’s hard to see how the SW won’t be the deciding factor.

  13. I think that the reason Apple did not progress in the PC market was because they hung onto their antiquated OS kernel (pre OSX) for far too long which limited 3rd party application development and integration into the business environment (where most people got their initial PC exposure).

    20 years later we are seeing the backwash as people’s preferences at home are pushing their way into the business arena (hence the rise of Linux and OSX) so integration is becoming the key (and MS is losing ground rapidly as Vista is a car crash).

    Most developers I know now use OSX or Linux machines, and OSX is gaining share because you can do pretty much everything Linux can but you are not forced to constantly deal with the ugly bits of the OS (You can also run XP in a VM if you need to).

    Funnily I think that Nokia and RIM are both in a similar space now in the mobile device market to the place Apple were in the early PC market (see my first paragraph). And if they do not realise this now they will in a few years when they follow Moto down the pan.

    As I said before Apple’s trick has been to take a stable and robust BSD foundation OS and apply a unified GUI that is pitched at the consumer and can ported to mobile/PC and desktop environments for a unified experience.

    This is something that pretty much all Linux ditros are still struggling with (who wants to mess about with the insides of the OS when you don’t need to).

    Additionally a reasonably unified SDK on the variants of the OS is a great cost saving as it makes it easy to produce mobile versions of desktop apps, without having to learn a whole new set of APIs or hire new dev teams.
    The fragmented nature of Linux is going to make this a lot harder as even if Nokia adopts it for phones and achieves a decent UI layer they have no presence on the desktop and I can’t see how they can then leverage the device APIs in a meaningful way. This means 3rd party companies are going to have to have very different skills to write the same desktop and mobile SW. Google is also looking at this with Android and their Google apps projects.

    I think the hardware is pretty much already at the commoditised stage (most phone parts are already from a variety of suppliers) it is how you put them together and the SW running on it that counts (i.e the iPod and the iPhone).

    The carrier subsidizing the phones and bundling with voice packages is a reality at present but once the voice becomes a secondary function of the device and wireless is used for a lot urban voice traffic instead of the GSM network I think that the carriers will lose this control as AOL and Compuserve did with their initial control of the ISP market.

    Like I said it is limiting to look at just the mobile phone dimension and project forwards just on that basis. It may be valid in the short term but as ramster said it is uncharted waters and as much as I think that Steve Jobs is an egomaniac he does seem to get it and you can see this in Apple’s overall strategy.

  14. Steve, that is one of the more informative comments we have had here. Thanks. So your argument boils down to OSX is easier to write SW on than Linux, therefore everyone will write programmes on it and the SW differential will make more people buy iPhones, creating a virtuous circle?

    If so, while I don’t know my GUI from my BSD SDK (I am honestly impressed you do, though) , I would note that Esperanto is supposedly a much easier language to master than English, yet has never quite caught on.

    If Nokia, Samsung, and everyone else took up Linux, so your addressable market is so much bigger, might you not bite the bullet and write your version of Donkey Kong in Linux, even though it is clunkier and will take longer and be more annoying?

    I can see why eventually just looking at the phone dimension might miss the point, but I have been looking at this space for a while now, and have become very bored waiting for our mobile data computing future to be born. At the very least, I can foresee a lag from now until then.

    You see, your argument viewed properly eventually comes down to the same as mine. You need a chicken to lay the egg. Apple needs to get its hardware scale as a phone down first before its SW advantages can truly come through.

  15. Hi Baruch,
    I take your point about the chicken and egg thing. But I think Steve Jobs is ahead of the curve on this and is not waiting for the scale to arrive before placing all the pieces of his strategy in motion (I am sure he lives in a volcano somewhere and has a white cat :-).

    It is not that it easier to write on the SDK (to be honest Objective-C is not that high up the popular languages list at present). It is that Apple and other developers can leverage this on all its offering simultaneously, which will also helps drive its PC uptake when the usefulness of this becomes apparent.

    The struggle the phone manufacturers have is that they have no foothold in the PC and online markets at all and so are very limited in trying to compete in one dimension, while Apple can use its presence in PC/retail and online data environments to get a compelling integration offering in place while everyone else is trying to squeeze yet more megapixels out of their devices.

    Granted not everyone wants this and not everyone can see the strength of it. I am probably wildly wrong here but it just looks so obvious to me I am surprised the telco analysts just don’t see it.
    I think that the business world tends to see things in light of existing models and misses the sea change until it has happened all around them.

    It sort of reminds me of the story about Gordon Moore at Intel who was so concerned about Motorola’s 68000 CPU features and the race to commoditisation that, when presented with the 8080-based destop PC concept by Ted Hoff and Stan Mazor, Moore said it was not a ‘serious’ computer and he could not see the point of it and pointed out that Intel would never be able to spend the time writing all the apps for it let alone provide the low cost screens and printers.

    At the same time Jobs and Wozniak brought out the Apple II and Visicalc arrived and changed the game. Admittedly MS won the game by doing the IBM deal but I don’t think that is likely to repeat.

    Moore was so focussed on the HW he could not see that for the user at some point HW features are not much of a difference it is the SW that counts and the same software on all your devices is a big pull through.

    Linux just can’t provide this as I don’t think its penetration on the desktop will get much higher until some company manages to get the UI right. If only MS would adopt Linux then Apple would have some competition.

  16. Oooh, right I think I get it — so the developer scale exists in the iPhone SW already because whatever they write for iPhone will work on their Mac. They are trying to get the halo effect again. That’s also why they have this MobileMe thing going on.

    My problem with that: say you wrote a version of Donkey Kong for the iPhone. Why would I want to run that on my Mac. It would be like taking a ferrari to the shops. I can buy Halo4 instead. Basically my question is: what are the apps I would want to run on both machines? I can’t think of one I really want. In my own luddite mind I separate both computing worlds. The internet, now THAT’s what I want on my phone. I want to buy the iPhone for browsing, SMS and texting. But that’s about it. Am I a narrow-minded clod?

    What I think Apple should do is get out of HW altogether. Nokia is desperate to get some decent SW and a nice OS. License the OS to phone makers! LG and Samsung can make the HW for cheap, and have no pretensions to owning the value chain. Make the Apple OS as ubiquitous on phone-like devices as MSFT is on PCs. Charge $50 for it, the HW commoditises — 100m smartphones a year in a couple years time, that’s a potential $5bn revenue, but at a super-duper margin, without any of that nasty hardware R&D.

    I think I mentioned that in the other comment chain that it was a good idea. The more I think about it the more I think they should do it. What do you think, Steve?

  17. Hi Baruch,
    You seem to have a bit of a Donkey Kong fixation here!
    I think that the integration will spawn a large number apps that were not possible previously – we just don’t know what they are – but a lot of development at the moment is around this area so it should be pretty interesting over the next couple of years.

    I am not sure about the licensing thing – Apple did try in the 90s on the desktop – but it fell flat and Jobs killed it when he rejoined. Maybe the lack of a decent competitor in the mobile world is a good reason to do it but I think it still assumes the balance of power is going to stay as it is.

    The drawback I could see is that the more interested parties have a say in something the more fragmented and bloated it tends to become – this is essentially what happened to J2ME.

    So Apple holding onto the HW and SW allows them to keep a more focussed approach and SJ can run his evil plan without having to negotiate with any other parties. Maybe he thinks in the long run they will own both areas as Nokia et al gets eaten by their focus on the hardware.

    Who knows – if I did I could make a killing :-)

  18. The browser is the only significant app for the mobile phone.
    I don’t believe that the pc experience with all of its peripherals
    and connectivity issues is relevant here. There is minimal synergy
    between the two platforms – otherwise MSFT will have won by default.
    The quality of the browser will be important, the SDK not so much.
    BTW is anyone here aware of the open source browser project?
    Ironically osb is a cooperative effort between AAPL and NOK.

  19. I stand corrected by non other than nokia itself. Yes the SDK
    is vital, still I maintain that on average, the browser
    and only the browser will be the vital app. And as to the future
    of the apple SDK, how will it compete with now free source Symbian?

Comments are closed.