If you’d asked me which company spends most on research and development worldwide, I’d have said IBM (famed for filing lots of patents), or possibly Microsoft (it employs lots of researchers, even if it doesn’t always show), or perhaps Honda (humanoid robots and jet aircraft don’t come cheap, after all, especially when your main business is building cars, motorbikes, marine engines and lawnmowers).

Of course I’d have been wrong, as the good people at IEEE Spectrum have shown. They’ve done the leg work (or mouse work) necessary to comb through all the annual reports and work out who spends what, compiling the R&D global top 100. Their numbers are for 2006 fiscal years, as it’s tough to get good data for the year we’re still in.

The surprising top 10, in pure pile-of-dollars terms, is as follows:

1.    Toyota (Japan) $7.49bn
2.    Pfizer (US) $7.42bn
3.    Ford (US) $7.2bn
4.    Johnson & Johnson (US) $7.13bn
5.    Microsoft (US) $7.12bn
6.    DaimlerChrysler (Germany) $7bn
7.    GlaxoSmithKline (UK) $6.61bn
8.    Siemens (Germany) $6.6bn
9.    General Motors (US) $6.6bn
10.    Volkswagen (Germany) $6.03bn

As you can see, Microsoft and Siemens are the only tech companies in the top 10, with five carmakers (none of them Honda) and three healthcare companies completing the roster. The US and Germany dominate, with the UK and Japan making up the numbers.

IBM comes in at number 14 with $5.7bn spent in 2006, and Asimo fans will find Honda at 19 with $4.6bn.
Other notable entries include Intel at 12th place and $5.9bn; Finland’s Nokia at 17 with $5.1bn; and even acquisitive Oracle at the number 48 spot with $2.2bn spent in the labs.

Frankly I'm surprised that the tech sector is so poorly represented.

Of course Toyota happens to be the world’s largest automaker, and so presumably can afford to splash a bit of cash in the labs. Interestingly, IEEE Spectrum has also provided data for what it calls R&D intensity, or R&D spending as a proportion of turnover.

I put the numbers into a spreadsheet and reordered by intensity, and of course the top 10 looks rather different. These are the firms spending the most per dollar of revenue, and therefore the most committed to making business progress through research:

1.    Qinetiq (UK) 44.9% - defence
2.    Electronic Arts (US) 33.7% - tech
3.    Broadcom (US) 30.5% - tech
4.    Amgen (US) 23.6% - healthcare
5.    AMD (US) 21.3% - tech
6.    Schering-Plough (US) 20.6% - healthcare
7.    Qualcomm (US) 20.1% - tech
8.    Eli Lilly (US) 19.9% - healthcare
9.    NEC Electronics (Japan) 19% - tech
10.    Daiichi Sankyo (Japan) 18.4% - healthcare

This list (and the rest of the top 100 by intensity) seems to better match my own impressions and prejudices about tech sector spending.

Attempts to relate R&D spending to business growth are, of course, far from straightforward. As IEEE Spectrum observes:

“Apple, the one company perhaps most closely associated with innovation, doesn't even show up on the R&D leaderboard this year [and] hasn't appeared there [in the last five years]. Its absence can't be attributed to size, because Apple's sales of  $19.3bn surpassed those of 30 of the list's 100 firms. Comparing those sales to the relatively meager $712m Apple spent on R&D in 2006 yields an R&D Intensity of just 3.7 percent.”

So while it may feel like a forward-looking supplier ought to be diverting a healthy proportion of profit into creating tomorrow’s products - it ain’t necessarily so.

I’ve been using the beta version of the Bloglines RSS reader, since I learned about the pending upgrade through Drew’s PR blog a few weeks ago.

I like the changes a lot and find the beta a big improvement over the current mainstream Bloglines, which hasn’t really changed in a long while.

The Bloglines beta developers have clearly taken a long hard look at Google Reader before designing their upgrade, as the two function in a very similar manner. I prefer bits of both. Google’s ability to easily share content with others through a web page or RSS feed, with a simple tick, is very useful. If you’re a blogger you can add the resulting feed to a box in a sidebar, say. In an intranet context it’s very easy to provide recommended reading lists to colleagues, who can slurp up your recommendations through their own choice of content aggregator.

However, I prefer the way the Bloglines beta keeps track of what I have and haven’t read. Bloglines won’t fetch items I’ve previously read unless I tell it to. Reader, by contrast, automatically appends them to the bottom of the list of unread items if you scroll down, with a subtle colour change the only indication that you’ve seen the item before. I find this tends to waste my time.

Neither app seems inclined to let me personalise how I view new articles. I’d like to read in reverse order - ie, starting with the oldest item I’ve not yet read and working forward in time, whereas the assumption seems to be that I will always want to read the newest item first.

This is not necessarily the case, particularly in the Google aggregator. If you read and tick to share items as you go along, your shared feed ends up reading backwards - the newest items from your reading list each day become the oldest items in your shared list, and vice versa.

I’m sure both Bloglines and Google Reader will become more flexible in the future. Neither yet does what I’d really like them to do, though. What I want is an aggregator that genuinely aggregates - that can take multiple feeds and group stories together by content. For example, if six of my feeds all include a new article covering the same topic, I’d like to see them grouped together in the style of Google News.