Wednesday, September 15, 2010

HP R&D and Palmissano

The dirty little secret is starting to get wider currency. Sam Palmissano, in today's Wall Street Journal, p B1, is headlined at "IBM's Chief Thumps HP". Spencer Ante's story uses the wrong numbers (the same numbers that journalists have used for these past two months -- I have been flabbergasted by how everyone uses everyone else's words/numbers as though they themselves had done the work!), but makes the point that HP has been lowering its R and D investment steadily for a long time now.

Palmissano said "HP used to be a very inventive company.... (but re) buying 3PAR, HP had no choice. Hurd cut out all the research and development..." Damning, but of course, just words from a competitor.

Ante then repeats the mantra that HP under Hurd cut R and D from $3.5B (4% of revenue) to $2.8B (2.5%) in 2009 while IBM has maintained their R and D percentage at 6% ($5.8B). True numbers, but he (like others) uses a foreshortened lens to look at the facts.

I first reported this systematic reduction in R and D investment in an May 2009 interview with Ashley Vance of the New York Times, who featured a barbed response from R and D VP Shane Robison, related in my May 26, 2009 blog entry "Whither HP Now". Our book, THE HP PHENOMENON, documents this in overview form on pp. 503-4, submitting that the Valley 'buzz' is that HP has become "innovation hostile".

It is perhaps noteworthy that HP tried to stop publication of the book, citing "numerous factual errors , inappropriate use of confidential materials, and unwarranted conclusions", none of proved to be supportable when we engaged on specifics. The errors we have found in HP website materials, pointed out carefully, have been ignored (such as how many oscillators they sold to Disney and the price point, etc., not that it matters except you'd think they'd care enough to have their own history correct).

Since Ante and others are insistent to show that Hurd screwed it up by going from 4.0% R and D to 2.5% R and D investment, I would at least like them to use the right numbers. For example, it is wrong to use the 2005 fiscal numbers as Hurd's baseline, since Carly was gone for eleven months of calendar 2005 and nine months of fiscal 2005. And Hurd was there for nine and seven months respectively that year. And we have three quarters of 2010 already reported, measurably lower than 2009 for R and D percentage.

First of all, the six decade continuous investment at 9.0% plus or minus 1% was ended by Lew Platt, who took it from 9.0% to 6.0% in seven years, 1992-1999 (5.8% 4th Q 1999). Carly continued the trend, at a somewhat slower decline, taking it from 5.8% to 4.9% (1999-2003). Under pressure from her Board in 2004, she held R and D dollars constant, and the percentage shrunk to 4.4%. Counting the first six months of fiscal 2005, she averaged 4.34% R and D investment for her last eighteen months (counting the 2nd quarter of 2005, ere Hurd was fully on-board).

Hurd quickly started reducing, cutting $46M out of the 2nd half R and D actual dollars, taking the percentage to 3.9% from 4.3% in six months. The carnage continued through his last days, the average for three fiscal quarters of 2010 is 2.3%. General selling and administrative costs are now 437% of R and D costs, vs. 308% when he arrived.

So, for the last three CEO's, here is the track record:
Platt reduced R and D percentage by 34% in seven years
Fiorina reduced it by 27% in five and a half years
Hurd reduced it by 47% in five and a half years

The claim at HP continues to be of the form that "it is hard to spend more money on R and D and expect better results; we're spending plenty...". But Apple seems to find new exciting products to invent, with higher investment rates and more risk-taking. Iterative R and D is really just D; memristors are indeed R, but somewhere inbetween is what used to be the sweet spot. Even when they out-invent the competition (great examples include 3D Printers and DNA Printers, plus HP Halo), they today have a tendency to license or outsource or forget to market their advantage. What gives?

"What gives" is clearly a focus on today, not tomorrow; pandering to quarterly results throughout the company which emasculates not just R and D, but the possible uptake or technology transfer from R and D to successful marketing launch. Palmissano got it right. He closed his statements by praising Ellison at Oracle, "Oracle invests". That might become the contest between Hurd and Ellison, which has not yet been picked up in the trade press as the possible hot button.


4putt said...


Correct me if my recollection is not accurate. HP has removed the last four CEO's. John Young was terminated by Dave Packard and Lew, Carly and Mark were removed by the Board.

If true, HP has not had a CEO make it to retirement since Packard.

Larry Potter
A HP retiree (Agilent technically)

chuck said...

4putt, you are correct that the last four CEO's have been replaced by the Board (Packard got Board approval re Young, but yes, it was Dave who drove that). pp 358-59, 374, 379-81, 384 describe Packard's (and Hewlett) angst re Young.

The other CEO to retire, of course, was Hewlett. Page 201 hints at the idea (without proof) that Packard persuaded Bill to step down also.

I think in fact that there has never been an HP CEO since Dave who retired solely by their own decision. To add to the irony, I think it is also true (again, no proof) that Packard wanted to return and run HP after his defense department days, but Hewlett stood in his way. Pages 190-197 hint at this; there was of course much more 'between the lines'.

Thus, if true, even Packard was 'not returned to power' when he sought it, through actions of 'the Board'. Six for six?

Algosome said...

Comparing percentages of R&D spend is really an exercise in accounting trickery, relating to the mix of hardware and software development in the company. Even discounting how today's accounting rules allow much of the time spent coding software to be counted as R&D, when the same step in the hardware product cycle would be counted as manufacturing, software costs nearly nothing to manufacture, so the fraction of expenses in bringing the product to market for everything else is proportionately higher. Consider how Microsoft's R&D expenses are 14% of revenue vs only 6% for IBM! Why? Because although IBM is still the largest software company, it has a substantial fraction of its income from other things like hardware and services while MS is pure software.

chuck said...

your comments are true enough, but actually give some poignancy to the box in which HP now finds itself.

HP does, if you dig a bit, break out R*D by sector, which doesn't allow finding the 'manufacturing R&D' hidden in the PC side, nor does it allow seeing the code massage being buried in the software number.

But all things considered, sector for sector, I think it is patently clear that HP has, under three successive CEO's, steadily reduced R*D, to the point that it today is not much of a weapon, but just a holding action.

I'd love to believe otherwise