Over the past few decades, governments have decreased their investments in original research, with corporations taking on a greater role. There are plenty of problems with scientific research funded by private, shareholder-driven companies. Perhaps the worst, however, is that corporate R&D, at least in the technology market, …
I was under the pression that Microsoft bought the Kinect technology from an Israeli firm.
And I think the reason Apple is so good at R&D is that they have focus. They attempt to innovate (or borrow/buy and improve -- see Fingerworks) within a relatively small area. MS, Google and Nokia indulge in the 'throw shit at a wall' method.
> Microsoft bought the Kinect technology from an Israeli firm
and Rare in the UK wrote the software.
> Microsoft bought the Kinect technology from an Israeli firm
More licensed really - but MSR developed a very different and much more efficient approach to body tracking than that used by NITE/PrimeSense. Mostly in Cambridge, UK as it happens.
>and Rare in the UK wrote the software.
As much as I hate MS...
the conclusions are 99% junk. Here is why:
- significant variables unaccounted for are:
1. Scope of business; it is easy to squeeze every last cent out of your customers in a particular niche (Apple)
2. Disruptive technological change; your curve is much steeper if you start form nothing and in a totally new area (Google)
3. General health of the economy; people start looking for alternatives when they are short on cash (open-source
4. Corporate idiots; "Our customers do not need tabbed browsing", " We are not interested in Internet", "Touchscreen has no future", etc., etc. How is this R&D's fault, if management take the wrong path?
-So the conclusion is R&D does not deliver, but the companies are cherry-picked. Look at Redhat, some Biotech's, oh, yes, throw in Google too. The conclusion is going to be the opposite.
Is there waste? Absolutely! But these conclusions are so far fetched and radical, I think they are mostly junk.
In a successful company you have all levels working together for the same goal. This conclusion I will accept.
Actually RIM's PlayBook is pretty impressive...
...and I'm usually pretty quick to diss Blackberry as a pretty junk OS. I know they bought QNX which was always kinda amazing - I remember running the entire OS from a floppy in the 90s including full GUI, networking, browser etc - but they had to to a LOT of R&D work there to make it happen so at least RIM can show *something*, unlike MS or, hah, Nokia...
One can argue MS sits on the top and blows mountains of R&D money on stupid things, only to find 1 out of 20 failed stupid projects that helps to maintain the leading position.
Also Apple has nothing to show: their modus operandi is to take someone's idea or even existing feature or technology and make it better, smoother, integrate it into the OS. Apple never invents anything - they never did and I can hardly imagine them ever doing it. But of course, they love to patent everything and quickly sue others...
"Actually RIM's PlayBook"
RIM's what? Where can I buy one?
Apple is a marketing and sales company more than anything.
Apple is a marketing company. The acquire technology, package it to look pretty, send the designs out to be manufactured by someone else, and do marketing and sales.
Apple is a marketing and sales company more than anything.
What product category have they ever produced that someone else didn't produce before them? Nothing. They just did a better job of marketing.
"Innovation" isn't a concept limited solely to technology.
It's perfectly possible to innovate in the fields of design and integration. This is what Apple do: integration.
My only quibble with the original article is that it seems to be making a generalisation: that all corporate-funded R&D is "bad". I disagree with this. It's also unfair to criticise Microsoft's fumbles without mentioning their successes: they've had a huge success with their XBox consoles, which came out of nowhere. The XBox360 is Microsoft's only current example of what they can do with a walled-garden ecosystem approach. If Microsoft were to build their own PCs and laptops, I think it's fair to say they'd do a pretty good job. But they can't: their corporate structure won't let it happen.
Similarly, Nokia are still selling well in the low end mobile phone markets around the world.
That Nokia and Microsoft are going through turmoil and transitional pains is to be *expected*: this is exactly what you would expect a company to go through when its obsolete business models are being systematically wiped out and replaced.
A coherent vision and focused management is what some of these companies lack. All the R&D in the world is no use at all if you're researching and developing the wrong stuff.
infrastructure projects != mobile phones
I guess Nokia has also another "proble" they are active in infrastructure development which is costly and not as rewarding as sexy Jobs' projects.
"corporate R&D, at least in the technology market, doesn't appear to work."
You should tell that to 3M. All of their products came out of R&D. Of course in their case R&D isn't limited to one department but to the whole company. Anyone in the company can have an idea and 3M will help them develop it into a product.
That is the exact problem here.
Software, IT, etc is innovative and inventive as a whole. We are not yet at the point where it will become a utility with 3-4% year on year growth and sub-15% utility margins. As a result it is essential for a software or IT company to be innovative as a whole in order to be able to deliver a long term perspective.
3M is a very good example, quite a few others.
The problem of Nokia and Microsoft is not that they use 14% on R&D, it is that they use _ONLY_ 14% on R&D instead of having the entire company streamlined around innovation and delivering innovation into the market.
In a company like 3M (or Apple for that matter) the actual R&D number becomes purely arbitrary because R&D transitions into concept to market, product and delivery without clear Microsoft/Nokia style borders between them.
Like it not the military (funding) has been the most productive R&D ever. When it comes to protecting the nations security or ways of (not) killing people the unlimited funding power of the military (government) with its black holes and secret research created the internet. Wireless, bluetooth all military projects with beginnings back in the 1960s. So in general I would agree the problem with greedy shareholders is they expect everything to make money rather then funding crazy scientists to think up/invent crazy shit in labs free of controls. Unleash the scientists, set aside an amount for 'The crazy ideas factory' and you never know what might pop out the end.
This is how Bell Labs worked. A huge amount of their research was military work, basic scientific research, and on weird shit. Some of it was telephone related; such as the transistor, com satellite, and cell phones, and some of it was WAY out there, such as radio telescopes (which started out using telephone microwave repeater gear)
Why the USA is a world leader in government subsidies
Because of its huge military.
RIM does M&A too
That's where the Webkit browser and QNX came from, right?
Title .. Schmitel
wrong, WebKit started as fork of KHTML, which is part of KDE
WebKit is Apple's fork of KHTML.
I don't want to agree with this.
Unfortunately, i think I have to. Corporate R&D doesn't work, with one exception: Skunkworks.
You see, corporate R&D doesn't work because corporate R&D is still bogged down by the extant corporate culture. They are mired in the thought processes of the existing company; "outside the box" is rarely rewarded. Startups don't have this issue; they are founded on a new idea and because they are starting from scratch, they are still just creating the corporate culture that will one day slow them down.
Skunkworks are a different breed. Arms length, yet with all the corporate power and funding that a large company can bring to bear. By being arm's length from the main corporate body, they can (and do) have a completely different corporate culture than the behemoth that requires their existence.
The barrier to R&D's effectiveness is quite simply that of corporate inertia. Past a certain point, everyone stops being in love with the technology and becomes far more infatuated with covering their asses. It is this risk aversion that is the death knell of any R&D.
I think you're exactly right. Its classically viewed as a mistake to start investing lots of resources outside of your core area, but this is exactly what startups do because they don't yet have a core area. I often wonder how effective a big company could be if it stopped being so damn integrated and started to embrace a "lots of startups" culture (with central support). It would be a really interesting model of which I'm not aware of any examples.
I am aware of an example
I recall something about them swimming in filthy lucre because of it.
Growth isnt everything, especially with antitrust people waiting to pounce
I guess if a companies only goal was growth or market share this would be valid. But in the case of Microsoft, growth like Apple has would be beaten down by antitrust lawyers in the US and in the EU, and massive growth of market share is not something they really need. Its more important to perfect the product and give "ease of use" and NUI to the users, than grow the company.
Heck, they might even try to split the company in pieces, god forbid!
Being #3 in smartphones, and #2 in tablets is not such a bad thing, it makes MS designers work harder, and competition is healthy!
RE: "#3 in smartphones, and #2 in tablet"
Just who the hell are you talking about...? And since when we count anything in tablets when there is no tablet market whatsoever so far...?*
*: no, having an all-conquering Apple model and nobody else even close to it is not a market yet - you need real competition to call it a market. Right now it's Apple only, even if it's largely useless due to its inherent iOS shortcomings.
Apple doesn't have to worry about anti-trust
Apple doesn't have to worry about anti-trust because it always has competitors -- the companies who were making competing products before the Apple marketing machine came along.
What utter utter twaddle!
…and the evidence for the inverse-correlation between R&D spend and corporate success… Apple Computer. Maybe I’m just cynical, but I have a hunch that low-cost Asian labour contributes just a little more to Apple’s profitability that Mattie gives them credit for.
Maybe Powerpoint really is what makes MS Office so successful, and not the staples of Word & Excel.. but I think most people would agree that it is more likely that Mattie is peddling twaddle.
Kinect was also an acquisition (of talent not a company). it was not internally developed initially, but was a base idea already developed by others that MS acquired and finalized in house through added development. Yes, Kinect was developed primarily internally (contributed to most heavily by Geiss), but the initial momentum was through bringing in people who had an existing idea developed at another firm. I do not know if they presented it to MS, or is MS caught wind and made an offer that could not be refused. What I do know is that the brainchild behind most of the root code, including the skelatal tracking system, took his pay and left as soon as it was productized, and as of October 2010 is no longer with the company, so good lick seeing any significant enhancements to Natal down the line....
It may seem like a small thing to a non-programmer, but for programmers, the addition of generics was a significant feature and it came from MS Research.
That said, I recall reading that it was one of the first things that came out of research unchanged.
In Canada, the government gives tax breaks for R&D activities, so there may be a lot of R&D that isn't research, just development and it doesn't really lose the amount of money that it seems to because of the tax break.
Generics were not invented by MS funded research
Microsoft's funding of research led to the invention of generics in the same sense that it lead to the invention of object orientated programming. I.e. it didn't. This really is just another case of MS buying in pre-existing expertise on a pre-existing idea that had pre-existing implementations. And unfortunately it is not clear that the claims for type-safety in the .NET implementation are all that they appear (it is not easy to retrofit genuinely type-safe polymorphic methods to a run-time environment that was not designed to support the appropriate type-constraints), not that such technical niceties seem to bother those who write the marketing blurb.
Even so, putting aside concern about dubious claims to have "invented" things, and to own "intellectual property", I think it is good for companies to have research groups, even they don't do the hard conceptual ground-work themselves (that is what the universities should be doing, if only the government would let them).
Invention and Innovation are separate activities with different risks and costs.
Traditionally, invention happens with individuals or very small groups - primarily because it is a creative process and it is people who have ideas. Innovation is the process of taking an invention to market, and it requires vast resources and many different activities - hence it is essentially the province of large corporations because only they have the resources to do it.
The cost of R&D - essentially the cost of innovation, is well recorded in the accounts of large corporations, and the full costs are exposed. This is not true of invention. Most inventions come to nothing, but the wasted efforts and expenses are not recorded, except perhaps in bankruptcy proceedings!
The economics of the two processes then are essentially different. Inventors are like writers (authors) whereby a very few strike it rich, but most fail to make any meaningful return on their work. Innovators are like film or theatre production companies, they take the works of a very few authors, and then spend huge sums of money trying to create a blockbuster.
Now look at the return on investment. The Oscar winning author makes a huge return on investment and the film company makes a big profit, but as return on their enormous investment, it is far more modest.
Of course, if we factor in the total costs of all the failed authors, I would suspect the the film companies actually turn in a better ROI, and this is why they purchase the creative output of independent authors.
I suspect it is the same with technology megacorps. Don't try and invent everything 'in house' because it is much cheaper to cherry-pick from the creativity of others, and by doing so you avoid the costs of failure.
Using the term 'Skunkworks' without mentioning Lockheed, and in lower case, was a poor decision.
Re: several points
Didn't you know that "skunkworks" is the new "agile"?! Everybody wants the coolness of the original Skunkworks to rub off on their tired bureaucracy, the latter being almost entirely the reason why various big companies aren't delivering on their "R&D spend".
Not that Mr Asay gets as far as investigating all this before veering off course as usual.
They're not the only ones....
Some years ago I was an accidental Intel employee. This was during the time that Intel was developing their low power chipsets and an important part of that development was the wireless interface. They threw billions at this, both in direct development costs and technology aquisitions, sums that were beyond the comprehension of someone reared in shoestring R&D outfits like myself. They were eventually successful but the amount they spent was over an order of magnitude larger than a startup like Atheros spent getting their own (vastly superior, IMHO) product up from scratch.
Large companies have huge cashflow and lots of financial people pulling the strings so they tend to have a weird attitude to R&D -- they'll inventory the paperclips while they're hosing money at what looks like the "cheap, safe" route to their desired goals. You can't change this behavior so you just need to figure out how to get on the other end of the fiscal firehose (after all, that's what Bill Gates did with IBM -- he had what they needed and they weren't too careful about how much it cost or under what conditions it was purchased).
Don't make me laugh! They only developed ONE original thing. Bob! And we all know how successful THAT was.
Everything else has been somebody else's idea, expanded upon.
Remember: as long as there is a C: drive, it will still be a derivative of CP/M.
It's not the spend, it's the freedom
If a VP tells the R&D people "Develop this", they will develop it. Even if it's a godawful idea that can't possibly ever sell.
If the R&D people are told "Make something cool and surprising" and more or less left to it, they're far more likely to come up with something cool and surprising that sells like hot cakes.
There is also another aspect to R&D which isn't mentioned here at all - product maintenance.
Once you've got software product in the field, you have to continually improve it, or at least bugfix it. Otherwise nobody wants to buy your 'latest and greatest', because they'll think "The last thing they shipped was a piece of crap because of bugs x/y/z and they never cared enough to fix them."
Of course, the more products that you're supporting, the greater your maintenance spend...
So one really effective way of radically reducing R&D spend is not to support products for more than a couple of years. Guess which of the example companies do that?
R&D isn't white coats
What is corporate R&D depends on the company accountants, what Wall St wants to hear and the local tax laws.
So company A counts almost nothing as R&D, everything is product development, to Wall St analysts it looks very lean and productive.
Company B counts everything that hasn't actually shipped yet as R&D, it writes off all it's development costs as R&D tax credits and tells Wall St it is a major innovator.
Look at the positives
It is easy to slag Apple off for just taking other peoples technology and using it. Such is the case with any commodity/manufacturing company, however, it misses the point.
Apple are Designers par excellence. This is not just slapping a few components in a pretty box with a bow on the front, anyone can do that and many have tried. What design, at its best, represents is careful and insightful thought into how to connect the bits together and importantly how the user may use the item in the best way not only in terms of raw functionality but to make the use of the item intuitive.
It is this insight that pushes Apple out ahead of the pack and in itself this is inventive - it certainly provides a very tangible benefit to the users. In many ways Apple provide the usability benchmark (which they keep moving) and as a result everybody has to improve. This detailed inventive, creative thought process (bordering on obsession) of course does not stop there - the internal and external form and function must also be right.
Design at this level is analogous to the architect. You have to see how the building relates to its environment, make sure it looks good, functions well mechanically, and ensure that the building works well for the users of it. You are concerned with the whole design from the widest view to the function of the bolts on the toilet door.
Clearly no one is perfect and I would guess that, to people who get irritated by an inelegantly routed wire, the recent woopsie with the iphone antenna was the cause of much angst and soul searching - these people will care!
To be a large successful company and still be innovative and also have a productive R&D arm is a tall order and requires different cultures in different parts of the organisation. If the prevailing culture is that of the dead hand of the bean counters, or the superficiality of the marketers you will struggle. Some companies manage it to a greater or lesser degree. Companies with a strong vision, often young and owned by the founder can have a purity of purpose. This is helped if they are privately owned as decisions are make quickly and you do not have to worry about the short term city gamblers reactions.
Apple is an established company that is, unusually, led by its founder - a man with a very strong vision and an aesthete to boot. Probably uniquely as a technology company they are design led.
For all it's faults Apple has a role to play challenging the more stodgy companies to raise their game. The world could do with a few more fruit in other areas.
In the UK, anything that is R&D is a tax write off
Leading to the policy at $JOB that everything 'new' we write is R&D. Leading to all the devs refusing to sign off work as R&D. Beancounters now assign whether it is R&D, based upon what the project is - hey, at least my signature isn't on the thing.
When Bureaucracy meets R&D
Spending up - useful results down
Bozo the Clown
This goes back to Xerox Parc and WIMP interfaces. It's NOT a problem of R&D; it's a problem of incredibly crap management. Scientists and Engineers can create entire new businesses IF the management recognises & rewards what it gets. However the management tend to be short term fixated and imagination (and intelligence) lite. 99% of work is screwed up by them and thus never sees the market.
Startups are capable of keeping these bozos out, taking the time and focus to develop the novel solution AND CAN BENEFIT FROM THEIR work, if they sell it off to these same bozo managers for £xxm later on.
To fix it, and get return from your 12% R&D investment, sack the bozos.
Having pure salesmen as tech managers and directors
Having pure salesmen as tech managers and directors, and they can't see the possible uses of abstract concepts.
Unless they see a completing product, they can't see a market.
It is what differentiates a high tech company or high tech division from a low tech company or low tech division -- is it run by salespeople or lawyers, or by engineers and scientists?
Most R&D spend is driven by fear in the marketing department, which is not a basis for rational or careful decision making.
Extremely unfair assesment!
The whole mobile phone scene would not exist without some very, very expensive R&D by Nokia, Ericsson and traditional large telecom firms. Not just the phones, but the whole GSM and 3G network infrastructure. Without that groundwork, Android and iPhone would just be MP3 players.
You miss his point
His point is that it is more profitable to be a marketing company than a high tech company.
R&D leads to patents
There is another reason why these companies spend money on R&D without this leading to actual products: patents! They do some basic research, register some vague patent and then sue the ass off everyone who's trying to be successful.
For a real glimpse of how things have changed read the 1987 NYT article cited regarding the Microsoft Office acquisitions.
"Some industry officials think such ''desktop presentations'' have the potential to be as big a market as ''desktop publishing,''
"Microsoft is already the leading software supplier for the Macintosh."
Wow. I think I heard this plot before
This sounds a lot like a work of fiction by Cory Doctorow. Big company (I think it was a hypothetical merger of Kodak and Duracell in the post-film era) decides to switch to primarily an angel investor scheme for all the innovative little companies, including a lot of their own engineers.
Lots of little groups of people making cool products as quickly as they can imagine them.
I liked the idea when I read it before and I still like it now.
http://craphound.com/makers/ ( and if you want there's a download now link at the top)
Size isn't what matters
Some R&D programs do generate successful products. And intuitively, it seems like some problems are best addressed in a formal research environment. But other companies do poorly, despite huge spend. The key variable seems to be management and focus.
IBM has it - a brief look through their research websites makes it clear that for each problem they've got a very good idea of what they are trying to accomplish and often what they'll do with the knowledge gained.
The impression one gets from the Microsoft Research site, though, is of a softer focus, without a clear link from the research to product, nor even a clear description of the final goals of the research.
Bayer, Intel, Texas Instruments, DEC, Mercedes, Bell, Boeing, GM, Toyota,
Lots and lots of successful products indeed.
While no company is successful forever, they've all had really good runs.
The answer is just so obvious I can't see how these companies don't see the solution staring them in the face. Fire everyone in R&D, hire some corporate spies, wait for their competitors to waste 6-14% of their revenue on tons of products until they finally come up with something everyone is willing to spend a lot of money on.. then copy it, changing the colour and moving the components around a bit.
Tax definition of R&D
Remeber the drugs companies example:
Marketing Sales teams bribing vendors(Doctors) all counts as R&D. So paying vendors to add time limited versions of Office to every Widnows computer is R&D.
You should not be surprised...
...that corporations cannot do R&D well. That is because large corporations don't do anything well. They are often choked by large bureaucracies, restrictive policies and unimaginative management.
When you have management who are always afraid of doing anything wrong, you typically never do anything at all.
- Comment Renewable energy 'simply WON'T WORK': Top Google engineers
- Useless 'computer engineer' Barbie FIRED in three-way fsck row
- Game Theory Dragon Age Inquisition: Our chief weapons are...
- 'How a censorious and moralistic blogger ruined my evening'
- Amazon warming up 'cheapo web video' cannon to SINK Netflix