Azara Blog: Bob Metcalfe and Hermann Hauser

Blog home page | Blog archive

Google   Bookmark and Share
 

Date published: 2005/02/11

The Cambridge-MIT Institute (CMI) sponsored a joint talk with Bob Metcalfe (in Cambridge, MA) and Hermann Hauser (in Cambridge, UK). Each gave around half an hour talk and then there was half an hour for questions. The subject was how to grow a small business into a large one.

(Metcalfe was one of the main people behind the ethernet standard, and founded 3Com. Hauser was one of the founders of Acorn, which transmogrified into ARM, and is one of the godfathers of the Cambridge scene.)

Metcalfe went first and gave an amusing talk. He said he was in Cambridge, UK in c. 1991-2 and used to wonder with academics here why there were no big companies in Cambridge. "Were the English stupid?" "Did they lack the marketing gene?" "Was there a lack of working capital?" No, none of these. He concluded that the problem was that the best grads wanted to "work for the government". (Well, or more likely financial institutions.) And entrepreneurs had low status.

At the time both Oxford and Cambridge were just then admitting their first MBA students. Many of the business school staff apparently had trouble getting college positions (which Metcalfe said, incorrectly, was a requirement to hold a university position), because they were looked down upon. On the other hand, in Silicon Valley entrepreneurs were treated like "rock stars" and "gods". Metcalfe said he met a few of these "rock stars" and was reasonably unimpressed enough to think "if they can do it I can do it" (it was funny when he said it).

He then gave two fingers (for US readers: the middle finger) to "marxists", who he blamed for the anti-entrepreneur attitude. (Well there have been no marxists to speak of in Europe for over ten years, so they can no longer be said to be the problem. But the chattering classes on the whole are still anti-corporate and anti-business, so there is still a problem.)

Actually he came back later to "marxists" (or "communists"). He said some French MP visited him to find out what the French government could do to try and create a Silicon Valley in France, and he told him "unelect the communists", to which the reply was supposedly "I am a communist". Anyway, Metcalfe said he came to the conclusion that the French government would only be willing to change things that didn't matter. (Lucky for the UK, since many French entrepreneurs have come here.)

Metcalfe said that to start a company that you want to become big you have to start thinking big on purpose and pick the appropriate people, markets, money (from investors), etc. Apparently a lot of techie entrepreneurs (and he included himself at first) do not distinguish between cash and profit. He said "cash is more important than your mother" (although he loved his mother).

There are two points when you need cash, when you are first starting and not yet profitable (cash hard to raise, for obvious reasons) and then when you are in the period of high growth (cash easy to raise).

He said you have to know how much infrastructure to build in advance. You should use a big accounting firm from day one because they will give you cheap rates at the start if they can see you might become big one day. He said you should buy/rent enough space so you can grow into it. (Well, obviously there is a limit to that philosophy.)

He said "Don't hire, recruit. B people hire C people, A people recruit A+ people." You want the best possible person in each and every job at any given moment in time (and who this is changes in time, as the company grows). At 3Com he was the sales boss until they got to sales of $1m per month, then someone else took over until $5m per month, another until $25m per month, and another until $125m per month. (All factors of 5 of course, so no doubt approximate figures.) They needed different people in that role at the different times because the sales channels were different.

Metcalfe said the founders have to decide "do you want it to be big or do you want to run it"? (He listed Ken Olsen as an exception. And no doubt there are a few others. Bill Gates is not far off.)

He said he got his vision while working at Xerox PARC (Palo Alto Research Center). When he left in 1979 to form 3Com he knew that networked computers, mice, etc., were the future. But in 1979 nobody would have said that they wanted to buy ethernet. (This was before PCs even really existed.) So 3Com did not do market research. "We were market research."

Not surprisingly, he said that you have to do everything well to succeed, including sales, something techie founders often neglected. Sales people were "a different form of life, not a lower form of life". During the period of high growth of the company sales people should have the highest compensation in the company. (Yuk.)

The customer was not always right. Don't give the customer what they want, give them what they are going to need. Choose your customers carefully. "Don't choose customers who go bust."

Raise "ambitious money", i.e. raise money from people who have high expectations. This might be like letting foxes into the henhouse, but it was better to have dangerous people on your team rather than on someone else's team.

Apparently in the US there are customers willing to buy from start-ups, which makes it easier (he contrasted with Germany, where he said don't bother with most contracts if you are not Siemens).

In order to foster the right climate for growth of these big companies you need a cluster of research universities (which he said were much more productive than the private labs of monopolies like Microsoft), and you need pre-existing start-ups (e.g. HP existed for 25 years before 3Com got started).

His final words were his strapline for his "secret of everything": "freedom of choice among competing alternatives", or "focaca". Just as well he was a techie and not a marketing person.

Then Hauser took over. His talk was largely a sales pitch. But what the heck. He mentioned the usual Cambridge suspects, ARM and CSR (Cambridge Silicon Radio), but also mentioned a "future ARM", Solexa, who do ultrafast gene sequencing (the claim being that they were aiming to be able to do an entire human genome in less than one week, for less than $10k, so the start of thinking about doing "personalised medicine").

From the government he wanted, or had already gotten,

Of course ARM is the great Cambridge success story so far (but nothing compared to the American giants). Apparently the ARM Risc chip is in 90% of mobile phones, in most PDAs (including all of Windows Pocket PCs), in Nintendo GameBoys, and in iPods. With 1.3 billion units shipped in 2004. That's a lot.

In the discussion period questions were alternated between Cambridge, UK, Cambridge, MA and emails from the webcast audience. Someone asked how do you know what your company should do by itself and what to use external contractors for. Hauser quoted an Austrian general: "a general should only give orders that nobody but a general can give". Metcalfe said that at 3Com they often switched between the two modes (dependent on the specific operation at the specific moment in time).

Someone asked how do you know which customers are the "right" ones. Metcalfe said that with the ethernet, originally they were told by sales people that lots of people with Apple 2 computers wanted it, but he concluded that was not powerful enough a computer, so decided to ignore these customers. The IBM PC was the first computer which was just powerful enough, and those were the customers they aimed for (luckily). Apparently they provided Sun with equipment early on in Sun's life, but then were told by Sun that eventually they would put the functionality on the motherboard themselves, so Metcalfe took that as an indication not to bother upgrading any of the products for Sun.

_________________________________________________________
All material not included from other sources is copyright cambridge2000.com. For further information or questions email: info [at] cambridge2000 [dot] com (replace "[at]" with "@" and "[dot]" with ".").