Apple Is Getting More Bang for Its R&D Buck | Hacker News Reader

Apple Is Getting More Bang for Its R&D Buck

bloomberg.com

42 pointsbko60 comments

I wish Apple would spend some more R&D money fixing up their Mac application-level software, which in my opinion has been on a gradual decline since about 2005.

More seriously though, Apple’s revenue is much higher than these companies, and their R&D budget is still in the same order of magnitude (larger than Qualcomm or Facebook, about the same size as Google). So the answer here seems to be that they have a more efficient business model or participate in larger markets than Facebook or Qualcomm or Google.

>So the answer here seems to be that they have a more efficient business model than Facebook or Qualcomm or Google.

Or perhaps the sorts of issues they're working on are simply lower-hanging fruit? (no pun intended)

Would it be rational to assume that even Apple has a hard time coming by quality people and all their A players are already busy laying the ground work for whatever product category they are about to enter next?

For example, if that might be a car they will need much better software and that from day one, otherwise people will die. The lack of polish on OS X might be an indication that they are working on something big. Not necessarily big for us, but a huge effort for them.

I assume that’s a big part of it. It’s disappointing though, because there are huge possibilities for user interface improvement in systems like file managers, “office” software, web browsers, image/video/audio/&c. tools, developer tools, and so on, for someone willing to invest serious time and money in getting the core abstractions right instead of chasing shiny new small features.

It’s depressing that for example the file manager from BeOS is still better than anything we have in current machines, and that the OS X file manager isn’t fundamentally anything but a mashup of the ones from NeXT and from original Mac, with slightly different chrome.

What's stopping a third party from making the software-with-better-core-abstractions you're talking about, and selling it—on the Mac App Store even?

In most cases, nothing in particular (the file manager is a bit different, as it needs work at the operating system level). But software projects competing in established markets which require many years of development to catch up to “industry standard” tools are hard to get off the ground, because new untested ideas are risky and there are few short-term payoffs. I would love to see new products competing in all these areas.

The one I’m most familiar with, Photoshop-style image editors, is based on core abstractions dating from research done on 1970s graphics workstations with far-too-literal adaptations of darkroom photography practices with 19th century ideas about how color works embedded deeply, and there is huge room for improvement if someone is willing to reimplement everything from scratch on better foundations. It’s a several-year multiple-developer kind of effort though.

I’m sorry about Apple in particular because it seemed like their Mac application software teams were on a roll from ~1997–2005, but have sort of stalled since then. (Some of what they’ve done since has been awesome, but overall in my opinion there’s been substantially less great stuff in the past 10 years than in the 8 years before that, and many apps feel severely neglected.)

I am thankful that Apple (with help from Qualcomm, Samsung, Intel, etc.) keeps pushing the hardware side though. The new stylus, “retina” displays, better battery life in thinner computers and tablets, etc. are wonderful.

Careful about ditching too many paradigms at once; you'll end up with something like Kai's power tools, which everyone agreed were revolutionary but always remained a curiousity. This industry does not like having to retrain.

I created a file organization app Blik (http://www.burntcaramel.com/blik/) sold through the Mac App Store.

It’s purposefully very simple, as I hope to not expand too quickly and make something new that is also complicated. But I actively develop it and have some interesting features in the works, such as something similar to Smart Folders. Also, I love users’ feedback!

This already happens - e.g. xtrafinder.

But... it's still odd that Apple cares so little about OS X that there's been almost no effort put into the UI of Finder since forever. Even the most vocal Apple fans don't think much of it. (Nor did Steve Jobs, by all accounts.)

In 10.6 there was a rewrite from Carbon to Cocoa, but not much changed on the outside. So it's been reliably disappointing for nearly 15 years.

Considering Apple's exploding cash mountain and the relatively tiny resources needed for an update (at the absolute max, a month or three for design, and a handful of developers for a year for implementation) it's hard to understand why there's been no movement.

It seems to me that you're ignoring some of what they're doing. For example, the latest version of Final Cut Pro is a big step away from the tradition of track-based video editing. As for developer tools, they've invested a lot in clang and even created a new language (Swift). Even Safari was aimed at cutting out the cruft from the browser (originally) to make it more streamlined and standards-compliant.

Why spend resources focusing on the Desktop, when Mobile dwarfs it in size by orders of magnitude? Mac had around 7 billion in revenue last quarter report (http://images.apple.com/pr/pdf/q4fy15datasum.pdf) but Mobile had 36 billon. 5x more.

Would it matter that much if you spend a bunch of resources to increase that 7 billion to 7.5 or 8? Are those resources better spent making the next 30 billion, or even just further servicing that 30 billion? No matter how much money you have, resources are always finite and must be spent accordingly. Even if its just mental resources.

May as well apply the same logic to Nvidia & AMD on Discrete GPU chipsets. Are the 30 million desktop discrete GPUs worth chasing compared to the 300 million mobile ones? Desktops seem to further increase their reliance on onboard GPUs as they service the majority of desktop needs well enough. Seems like everyone is chasing Mobile.

I don't see Apple as any different viewed in this light.

Desktop is still the tool to create Mobile. Both software and hardware. When I look at it this way it makes sense to me that Apple might want to invest in making it better.

To follow-up : maybe they are still more laser-focused than Google, Facebook or Microsoft are at the moment ?

I think the car business is not where Apple should be looking. After all, once the driverless cars become a reality, people will not need to own their car. Transportation will become a service, rather than an "owned" product. Apple would have to totally reinvent itself to become active in such a market.

The issue is really incentives. If you have access to the capital of Apple, Google, etc. then it's simply giving a green light to spending more than might make sense, and doing so in a way that makes sense.

I wonder how much Apple's draconian employee secrecy policies hurt their recruitment. Developers that like to blog and contribute to open source are probably wary of joining Apple.

Instead of a more efficient business model, I'd offer that it's actually the difference between being a hardware and software company.

Being a year or two late in software can often mean you product never has a chance to disrupt the first to market (assuming roughly equal quality product). Contrarily, Apple has shown time and again they can be several generations behind in specific technology (until they can "get it right") and still maintain demand for their products.

Comparing Apple to Qualcomm, Facebook, and Google is a mistake.

Apple is a consumer electronics company. Typically consumer electronics companies don't R&D their components. iPhones use Qualcomm baseband chips. Apple does design some of their components, but that's actually atypical for a consumer electronics company. Even then, Apple needs someone like TSMC to fabricate them.

So a lot of Apple's revenue goes to buying components of companies that are spending a lot on R&D. Qualcomm is making the components. They are doing the R&D because they have to. TSMC is R&Ding building 14 nm chips. Hynix is creating bigger NAND flash drives.

Google and Facebook are different because they sitting on monopolies that print money. They also do much more in house because it's easier to.

Apple has some ability to use their size and existing products to capitalize on new technologies. But it's probably better to just return their profits to their stockholders and let the market fund new R&D.

Then again a lot of cool tech came from monopolies playing around excess profits (Bell Labs, Xerox, etc.).

Then again a lot of cool tech came from monopolies playing around excess profits (Bell Labs, Xerox, etc.).

But notoriously, a lot of cool tech that was developed at Bell Labs, Xerox, and Kodak ended up being so severely misunderstood by upper management there that outside companies were the ones to develop and sell products based on the new technology, while efforts to do so in house were practically or actually discouraged. These include such game changers as Object Oriented Programming, WYSIWYG word processing, Graphical User Interfaces, and Digital Cameras.

However it happens, Apple is good at taking R&D, no matter where it originally happened, and creates game changing products out of it. One could argue that this is baked into the very DNA of the company. (Two of the oldest examples: Microprocessors ==> Personal Computing, and Graphical User Interfaces.) If you like, think of Apple as the anti-Xerox. It's not the company that spends tons of cash on internal R&D, misunderstands it, then winds up being unable to capitalize on it. It's the company that can take anyone's R&D or idea, understand the essentials of it, then executes great products.

Why be the company that invents something, when you can instead be the company that profits from it?

Apple designs its own microprocessors, numerous sophisticated connectors, extremely small speakers, clicking hardware, touch sensors, metal plating processes, etc.

While they partner with folks like Samsung, I don't think its fair to say that they only design "some" components. They are industry leading in many areas, including batteries.

While Apple does do a lot more engineering than the average phone maker, it still couldn't even make the A9 processor themselves.

What part of the A9 does Apple not make?

It's an ARM core manufactured by TSMC or Samsung, but neither of those caveats remove "maker" from Apple's role.

Note that Apple is part-owner of ARM Holdings, and that no one except IBM, Intel, and the government have their own fabs these days. Samsung has their own fab, but Samsung also makes tractors -- they're in hugely disparate businesses.

All of it. Semiconductor processing is a supremely complicated endeavor.

Of course it is. That's kind of the point.

Just like everyone else, Apple does not fab their own chips. But unlike most others, they do design their own chips.

If Apple saw a competitive advantage in running their own fab, they would do so. I'm not sure what definition of "make" you're using here, but if it's just the production part you're talking about, then ok.. Correct but not strategic, in the scheme of things.

What about their margins ? It seems commonly accepted that Apple has much better benefits for every piece of hardware they sell than any other players.

Here's an article by Horace on Aymco in 2012 on this same subject [1]. Back then Apple was spending just north of $600 million a quarter on R&D (in 2011, which is when he had figures for). Since then to reach $8 billion a year, Apple would have had to increase spending by over %30 per year. That's a very impressive ramp up in spending, especially when sustained over a substantial period of time.

The thing is their revenue growth has been so spectacular, it's swamped their otherwise dramatic increase in investment in R&D on a proportional basis.

[1] http://www.asymco.com/2012/01/30/you-cannot-buy-innovation/

R&D accounting for tech companies should somehow factor in M&A spend as this is becoming the de facto way to "generate" research and IP for large tech companies.

> Apple’s relative thriftiness extends to its vaunted advertising and marketing operations. The company spent $3.5 billion on advertising and marketing over the past four quarters, while Google spent about $8.8 billion in the past three. R&D spending isn’t worthwhile if it never leaves the lab, says Tim Swift, a business professor at St. Joseph’s University in Philadelphia. “Apple’s products are going through some of the most effective and sophisticated marketing we’ve ever seen,” he says. “That’s the other half of what makes Apple the most productive R&D spender.”

Surprising to see that Apple also spends less than 50% of what Google does in direct advertising. Amazing what kinds of costs you can cut when you have such strong marketing already built into the name, look, and design of your product.

Google's sales and marketing expense category consists of the following items (from their 10K) - some (perhaps most) of it is actually stock comp for sales & marketing employees.

• Labor and facilities-related costs for our personnel engaged in sales and marketing, sales support, and certain customer service functions;

• Advertising and promotional expenditures related to our products and services; and

• Stock-based compensation expense for our employees engaged in sales, sales support, and certain customer service functions.

I read this as "how Apple relies on the R&D of other companies" For example, the quote about how IBM spent 100x as much on R&D as Apple.

IBM has produced 6 Nobel prize winners that lead to revolutionary advancements in the industry which Apple "innovation" relies on. There would be no iPhone without the investments in basic research and process that has done been done at IBM, Bell Labs, Intel, et al. When I worked at IBM, I couldn't wait to read the next IBM Systems Journal, it was the only piece of "corporate" internal communications in any company I've ever worked for that I looked forward to. I worked at IBM T.J. Watson in the 90s and one of the things told to us during orientation is that our budget, and our research was separate and independent from the commercial divisions of IBM and we were free to work on stuff that wasn't necessarily commercially viable.

Something rubs me the wrong way about this article and Apple's R&D spending. R&D spending is wasteful, it is a gamble, but every once in a while, it produces something incredible that leads to a huge net positive externality for the world. Some of the greatest inventions were done on research no one ever thought would have any commercial usage. These days politicians decry studies on bear DNA, or Shrimps on Treadmills, but who's to say these won't be vitally important things 20 years from now?

Part of relying on the research advancements of others, standing on the shoulders of giants, IMHO, is to take your success and your profits, and invest them, continuing a kind of scientific "social contract" for those that come after you.

Hoarding profits and bragging about how you can get away with making lots of money without having to hire many researchers is IMHO, a kind of parasitical attitude.

On top of that, is the tendency of people to credit Apple for all advancements, even those that were tweaks on hard work done by others. Do I credit Apple for their graphics performance, or Microsoft, AMD, PowerVR, NVidia, SGI, and everyone else who over the past decade have invested massively into consumer 3D acceleration? Do I credit Apple for amazing battery improvements, or Intel or Samsung investments into NAND? Is the A9X more important than TSMC/Intel/Samsung fab innovations? Or IBM's introduction of STM, Magnetic Storage, Eximer Lasers?

Science and Innovation are social endeavors relying on a community of shared progress. The secretive way Apple does R&D, their lack of publishing, and the way they promote this 'lone genius/inventor' view of the company rubs me the wrong way. I'm not saying Apple hasn't innovated, especially in mobile software, but won't should brag about not doing basic research, and banking a hundred billion in offshore cash to be used for what, stock buybacks?

Just what is wrong with say, spending some of that money to fund research into say, quantum computing or medicine? Does every damn thing have to be applicable to the next iPhone?

>Shrimps on Treadmills

Serious question, meant with no sarcasm at all: Can you give an example of something as ridiculous as the shrimp on treadmills study that cost a fortune and ended up being incredibly important in a way that wasn't immediately obvious at the time the study was conducted?

The shrimps on treadmills study didn't cost a fortune, that was ridiculous and dishonest conservative demogogueing over it with fictitious accounting.

It's like when a politician will say "This government project spent $1 billion to build a highway, and only produced 1000 jobs, at a cost of $1 million per job!" without looking at the other thing the project produced: a new highway!

The shrimp treadmill cost only $47 to make.(http://chronicle.com/blogs/conversation/2014/11/13/how-a-47-...) and was trying to figure out a way to determine food safety, you know, like whether the shrimp we're eating are diseased.

I'm pretty sure if you total up the money spent on number theorists over the past several decades, it all looked ridiculous numerical masturbation, up until the point where public key cryptography was invented.

By contrast, military R&D is HIDEOUSLY expensive by several orders of magnitude. The people whining about a $50 shrimp treadmill need to get their fucking head in order and look at the cost of the F-35.

Thanks for answering. Now it makes a lot more sense. I wasn't aware of the follow-up articles, only the original politically-fused exaggerations.

There was a time when X-rays were considered useless or nuclear physics. In addition, if you took a closer look at a shrimp on treadmills study and its rationale, you would often see the use pretty quickly.

The British government gave Charles Babbage the 1830s equivalent of 2.4 million USD to come up with a more efficient way of making printed logarithmic tables, which Babbage never managed to accomplish.

> Just what is wrong with say, spending some of that money to fund research into say, quantum computing or medicine? Does every damn thing have to be applicable to the next iPhone?

Things probably need a business plan backing them, which involves making money. Usually for social advancement we lobby the government to fund said research for our betterment. The private industry tends to fund research for its own betterment.

Also phrased another way, I imagine AAPL shareholders would be upset if Apple spent those profits on research into medicine, unless Apple was trying to build a product to sell said Medicine.

I think Society in general can benefit but it can be harder to see and the motives may often appear "after the fact", they may not have been an actual goal. Are people healthier due to health fitness trackers? Will doctors have better care due to the deluge of data? Can researchers actually get better data from cloud sourcing informatics, allowing their published work to happen faster or iterate on ideas quicker?

I wouldn't put a business plan on those goals, but I'll take them as a side effect for a gadget that everyone buys for whatever reason.

These seem like different horizons on investments. Apple seems to be dedicating a lot of resources to projects with 5 year horizons, while Google seems to be dedicating a lot of resources to projects with 20 year horizons. Neither strategy is wrong, but of course you would expect better ROI in a short period than money spent on fundamental research at a (relatively) new company.

Technology goes really fast. 20 years is insane, I'm not sure I'd bet a technology business on where we'd be in 20 years. 5 seems risky but worth thinking through. Probably not worth planning for except in the most abstract of ways.

I agree - but I'd describe things like self-driving cars or curing cancer as fundamental research rather than technology.

TL;DR: 3.5% of $233bln is more than 21% of $12.5bln. Is this just an article about basic math, or am I missing something?

Your missing something.

Reading beyond the infographic, the article is really about how Apple is able to offset the need for R&D dollars (which as expanded from $4b in 2013 to $8.5b in 2015) by being an overwhelming large consumer of components ($29b committed next year).

This causes the suppliers to invest heavily in R&D to be able to "wow" Apple and lock up big contracts with their innovations, instead of Apple needing to do it directly.

The most interesting part to me is if this speaks to a counterpoint of the slow direction Apple appears to be taking moving more and more in house. This appears to indicate there is significant value by keeping key partners separate and in the free market, even with the lessened control of being a 3rd party.

Seen that way, Samsung Electronics is basically an Apple subsidiary kept at arm's length to keep the market forces going.

Apple is a marketing arm of Samsung.

I see what you did there. "... marketing ARM..." Clever. ;-)

No, you're missing something.

Tl;dr: Apple only spends 3.5% of its revenue on R&D versus companies like Google, Qualcomm, and Facebook who spend 15%, 22%, and 21% respectively. (2015 stats)

Well, their absolute R&D spend seems to be working very well for them!

That said, I wish Apple would invest more in their web services. I find iCloud, Siri, etc. to be not as good as services from Google and Microsoft.

Because of (at least what I percieve is) Apple's fairly strong stand on the basic human right of privacy, I would probably award them all of by business is their web services were much better.

As it is, in addition to Apple gear I also use a Chromebook, Android phone, and even a Windows 10 laptop (I was curious to kick the tires of Windows 10).

I'm pretty sure I have read about some pretty strong stands for privacy from both other companies you mentioned too.

I am a huge fan of Google's services but they do make money from our personal information.

I am hopeful that Microsoft will evolve into a strong supporter of privacy, but with some (hopefully temporary) issues with Windows 10, I will wait and see what happens.

They make money from personal info but not necessarily in a way that would automatically be considered a violation of privacy.

The best way I can think of it is to compare to cable TV. Insert some other service if you've never had or wanted cable TV. Right now I pay something like $40/mo for cable TV and still have to watch something like 15-18 minutes of advertising per hour of programming.

The "Google version" of cable TV would either be free or cost a nominal $5/mo subscription and their algorithm would automatically use my viewing history to determine which ads might be applicable to me. As a result, I'd only have to watch a single 5 minute ad segment in the middle of every hour-long program. The better odds of a given ad being applicable would make the ad time more valuable so they wouldn't need to run as many to make the same amount of profit. Likewise, people and companies that want to reach potential customers waste less of their advertising and publicity budget on viewers that would never buy their products.

So, gross oversimplification aside, there's "making money from personal info" and there's "using data to make the same money from less advertising than traditional methods". I'm sort of on the fence myself and somewhat playing devil's advocate here but it's something that's often lost in discussions.

My main concerns are that government, law enforcement, or criminals would gain access to data that I only agreed to share with one company's algorithms. My main point is that "making money from personal data" can sound a lot like "selling my personal info to god knows who" which I've never known to be the case regarding Google or any of the info that's collected by Microsoft in newer versions of Windows.

My main conflict is between two things: a reflexive distrust of the motives of business and an acknowledgement that some really cool things (and plenty of mundane but still good things) only work when you have the data to power them. It's definitely a complex issue and not something I'll solve in an overly-long comment while I waste the last 10 minutes before it's time to go home.

Thanks for your comment. It is a complex issue for me also. I get a lot of value from Google Now and even FB has value because it lets me know what family and friends are doing.

Is whatever was spent on something like buying FaceShift counted as R&D spending?

That's a polite way to say that apple is spending less on R&D.

They're spending less per dollar of revenue, but the amount they're spending in absolute terms is totally comparable to other big tech companies.

apple is not a standalone research company, it is market driven, with its purchasing power for components and tooling budget is making its vendors innovate and that shows up

Are acquisitions considered money spent on R&D?

Thanks for the autoplaying video ad!

I read this as: how successful you are has nothing to do with how innovative you are.

The Apple formula: keep innovations at a minimum, but market all your products as the coolest thing that anyone has ever seen anywhere. With enough of a marketing budget, everyone will believe you.

From the article:

Yeah, no. If all they had was marketing, and the products were not actually good, then no one would buy version 2.

The only thing I can assume is that people like you haven't ever really used Apples full line of products and seen how seamless the experience is in comparison to anything else on the market. Or used one phone for 2-3 years and not felt completely crippled in that time. I have rarely used an Android device that didn't feel sluggish.

Reply on news.ycombinator.com