Google and Apple have different innovation signatures | Hacker News Reader

Google and Apple have different innovation signatures

fastcodesign.com

113 pointshunglee231 comments

I'm not sure I buy the argument that Apple's inventors per patent means they're a more open or egalitarian companY. The hotspots seems to suggest that Apple just has a culture of giving management credit on the patents.

Apple has an "invent this for me" culture at the very top. That does not mean those are not real patents, but I doubt if someone travels back in time from 2040 and we posit they are now are at the top of Google today, you can speak three sentences and give the instruction "invent this for me" and have your name at the top of a patent filing that essentially just prettifies those three sentences, turns them into claims, and does the prior art research.

You should be able to, because by 2040 there are lots of things that would be considered genuinely novel advancements to the state of the art today, but which can be described in three sentences.

It's a question about whether top executives can ask someone to invent something for them or not. At Google, no, at Apple, yes.

interesting point, which raises the question of which structure works better at bringing inventions to market. Wasn't it Thomas Edison, credited with tons of patents and produced many actual groundbreaking products, who had an army of inventors working for him? Perhaps if the people with financial and management resources to deploy the organization behind the patents, are the ones who own the patents, this means a better chance of stuff coming to market? Xerox Parc, by contrast, seems to have been quite separate from Xerox itself, and so we have this situation where many of its amazing work was never brought to market properly by Xerox itself?

>> At Google, no, at Apple, yes.

Why ? Is this the cultural, or more related to the type and complexities of those inventions ?

Wonder if part of it has to do with incentive structures for patents? My understanding is that at Google, if you have <= K (2? 3?) inventors on a patent, they each get $X,000, and more than K, $X*2,000 is split N ways. (I have not received any patent bonuses personally, so this is all second- and third-hand.) This means that it pays to keep the number of inventors on a patent small.

If Apple has a different structure, this may account for the differences.

How interesting, that the company gives out bonuses for patents. I would not be surprised if Apple doesn't do this, much like how their retail staff don't get paid in commission.

Apple does give bonuses for patents, but they are of a mostly symbolic size.

I know at least two more companies that give bonuses for patents: IBM and Motorola (before being bought by Google).

My employer pays a decent bonus (> the $4000 cited for Google earlier) and we have a pretty small patent portfolio - I always imagined it was pretty typical for tech companies to but maybe that's not the case?

This is exactly right, and the X was 4 in 2014. Not sure if it's changed since I left.

This entirely ignores the politics that lead to people listed on a patent that contributed nothing. It happens a lot, if not almost always, on patents filed by employees at a large company.

Although having an inventor that you can't attribute a contribution to invalidates the patent.

Doubt this matters in practice when you have the legal and financial resources that Apple does

Sure, but say I come up with a novel idea. I run that idea past another lead as a sanity check, but I do most of the work myself. At one large company we were at the point where if the patent folk came around again it was our plan to milk the process. (Especially after we worked for three years on what was supposed to be a common data format for interchange between manufacturers, OEMs and customers and some halfwit thought it would be a great idea to patent the whole thing at the end. Asshole. We became highly motivated to publish our docs to start the patent clock ticking and eventually got the whole idea canned.)

In a world where the company is going to patent my work no matter what my opinion is,, 'collaborating' with people I didn't really need to collaborate with is a net positive for me.

One, I can let them make or review draft documents so that I'm not the one doing all of the bullshit paperwork. If they participate in the vetting and documentation process (especially the 'arguing with the patent attorneys' over scope and wording), then they legitimately collaborated on the patent. We extract more dollars from the bounty system and we make frivolous patents slightly more expensive.

Then when one of those two people think up an idea on their own, why not put me on it, especially if I offer the same kind of help?

yeah I was thinking the same, and wanted to look at IBM because of that reason to see if the larger nodes matched middle meddling managers that hold the keys to the patent process, but the search turned up only some hundred patent, which is weird because it's touted as the largest patent holder ever... oh well.

https://www.google.com/?tbm=pts#tbm=pts&q=inassignee:%22Inte...

I know there are, I don't now how to get them to show in the render thing the article pointed to

Can we stop attributing "innovation" to the number of patents one company has? We know by now 80% are bullshit, especially if they are software patents.

This seems to say more about the incentives around patents at Apple and Google than it says about innovation or anything else at those companies. In particular, it seems to say that patents are more highly encouraged at Apple and that there's more incentive to put them in your own name at Google.

> "Over the past 10 years Apple has produced 10,975 patents with a team of 5,232 inventors, and Google has produced 12,386 with a team of 8,888," writes Wes Bernegger, data explorer at Periscopic. Those numbers are, frankly, pretty similar in terms of proportion.

Ummm, Apple's 2.0977 patents per inventor and Google's 1.3936 patents per inventor are pretty _dissimilar_, I'd think, unless many companies have yet more or far fewer.

Close enough, no?

Or: if we've got data let's go with the data. If all we've got is opinions let's go with mine.

If you think 1.3 is close enough to 2, I have some investment options for you to consider.

> unless many companies have yet more or far fewer

Does anyone have a source showing the distribution of patents per employee across e.g. the S&P 500, Russell 2000 and NYSE Composite members?

Nice graphs, I wonder how Samsung graph of patents looks like.

  Apple: ISD/1/1/2016->12/31/2016 AND AN/Google - 3389 patents
  Google: ISD/1/1/2016->12/31/2016 AND AN/Apple - 2535 patents
  Samsung: ISD/1/1/2016->12/31/2016 AND AN/Apple - 10163 patents

USPTO search: http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=H...

Better yet, how does the Samsung crystal ball of patents looks like, if I filter the set for washer dryer, for example. Hairy perhaps?

Related question: how many scientific papers does an Apple researcher get to publish, versus a Google one?

Agree. Publishing research is a more accurate way and removes the fact the two companies fundamentally view pattents differently.

Think this article misses a fundamental difference between Google and Apple and that is fundamentally the two view patents differently.

Seems to me, that apple bosses want to put their name on every patent created in their division.

Is this linked communities in R ?

It would be interesting to see Microsoft's pattern compared to the 2

This article makes a deeply flawed assumption by including "outside" patents, that is to say, patents which were acquired (rather than developed/filed/patented by the company's own employees). Notice this tidbit disclaimer buried towards the end of the article:

"In Google's case, we get a clue. One of the company's largest super inventors lives out there, in the periphery, disconnected from other products. That inventor is Kia Silverbrook, who sold the company 269 granted patents on cameras and printers in 2013. Obviously patents that have been recently acquired, rather than developed in house, would lack the interconnections with other employees that centralize the largest bubbles."

This is not a clue. This is a sign that you're biasing your sample. Compare this statement to the premise of the study:

"Over the past 10 years Apple has produced 10,975 patents with a team of 5,232 inventors, and Google has produced 12,386 with a team of 8,888."

Wrong, not all of those people were even employees, see above. And what about the huge patent purchase Google made by buying Motorola and then selling it off (Arris, Lenovo) while keeping the vast majority of the patents. [0] Is it accurate to count Motorola’s patents as Google-developed innovations? Is it accurate to group Motorola inventors along with Google inventors?

"This seems to indicate a top-down, more centrally controlled system in Apple vs. potentially more independence and empowerment in Google."

I disagree, this really doesn’t indicate anything and mis-characterizes the inventors. Google and Apple have purchased a sizable number of patents whose inventors are not employees and not involved in Google's or Apple's R&D. If anything, this may indicate that Google purchased more outside patents than Apple. Some examples of Google's purchases beyond Motorola are IBM [1], Silverbrook [2], IP3 [3]. One of Apple's biggest set of outside patents is Nortel [4].

"Google, on the other hand, has a relatively flat organizational structure of many small teams filled with empowered individuals."

Not disputing this but the more obvious explanation is that Google has been working on a wider variety of products and services than Apple which would mean a greater variety of patentable subject matter by inventors working in different technologies. Don't use patents or that visualization as any indicator of R&D organizational structure.

"But that would take further sleuthing to confirm."

In the patent search community, it's pretty easy and quick to filter a search to only include company-invented patents (it's as quick as filtering down to a specific company, just compare latest assignee to original assignee). So why not do that here? This is such a simple search that I'm surprised that the data provider "Portland-based data visualization studio Periscopic" didn’t team up with any of the dozens of patent analytics providers [5] who probably would have done the work for free for some publicity, not to mention address the glaring flaw of including outside patents (that is, non-employees as inventors).

"Our intention behind PatentsView was to create interfaces that could inspire the public to explore patent data," says Periscopic cofounder Dino Citraro. [...] Indeed, Apple and Google may only be a start. You could probably write a book on the corporate organizational structures revealed in PatentsView."

I hope no one uses this tool to write a book. I think it's a noble goal to open up the data and visualization to the public, but users should be very wary of what they’re really looking at.

[0] https://en.wikipedia.org/wiki/Motorola_Mobility [1] http://www.iam-media.com/blog/detail.aspx?g=54821371-e57d-47... [2] http://web.archive.org/web/*/http://www.seobythesea.com/2013... [3] http://www.prnewswire.com/news-releases/calling-all-patent-o... [4] http://www.reuters.com/article/rpx-rockstar-ip-idUSL1N0U713M... [5] https://www.piug.org/vendors

Reply on news.ycombinator.com