07 June 2010
Special Report: Start Ups and Spin Offs
Air Semiconductor founders Stephen Graham, left, and David Tester
Being a semiconductor start up involves more than just chip design.
All entrepreneurs start with an idea, but the route forward can be very different. Designing a chip and then getting it designed into a product by large, multinational customers isn't enough – finding the right team at the right time and keeping the investors happy is just as important. And timing and luck play a key role – the dotcom bubble and the recent banking crisis have been influential factors for several young electronics companies.
But does the team or the funding come first? David Tester set up Air Semiconductor in 2006 with the idea of allowing cameras and mobile phones to access location information from the GPS satellite system.
For Tester, it was the funding. He was introduced to Mike Gera of UK venture capitalist Pond Ventures at a SiliconSouthWest networking event in Bath. Gera introduced him to Stephen Graham, who became the VP of sales and marketing and his cofounder. With the first round of funding – $1.75million – raised in just six weeks, the two built a team to develop the chip, but it wasn't easy. "The day I got the signed agreement in my hand – before the money was in the bank – I had a call from an investor asking why I hadn't already recruited a team of 20," recalls Tester. "Building a team with the best people is vital and takes time."
Huw Davies took a different approach at the start. He was returning from running Conexant's video chip business in the US and looking for opportunities. He met engineer Clyde Witchard in 2005, again at a SiliconSouthWest networking event. Witchard had an idea for an audio amplifier that could be 20 times more efficient than today's devices and run a wireless speaker for months off a simple battery.
With a £75,000 grant from the local development agency, matched with their own investment, they started the development. Initially, they used Davies' contacts to meet venture capitalists in the US. "We first went to Silicon Valley to raise funding, but they said 'we like what you are doing but you are 5000 miles away'." said Davies. "So then we went to the UK funds.
"From April 2007 to the A round took five months and we raised $8.5m. We hit every milestone in the first 12 months, built a nine man team and everything was going well. Then the banking crisis hit and our board decided this was a terrible event. There was a recruitment freeze, but the goals for the next 12 months didn't change.
"Then we became that terrible thing – an orphan investment," said Davies. "Our two lead investors left their funds, so we had two new board members, one of whom didn't even have semiconductor expertise."
Audium, as the company became, had been working closely with pc peripheral maker Logitech, as well as some manufacturers in Taiwan. "We had 25,000 pieces on order and a lead reference customer in Logitech.
But being an orphan investment meant there was no appetite for further backing and our investors gave us six weeks notice that they were not going to follow on their investment."
So the company closed in May 2010 and the intellectual property was sold to UK flat panel speaker developer NXT.
"We delivered everything we said we would on the original business plan. Four or five years ago, that would have got us to a B round of funding, but the world has changed," he said.
Similarly Air Semiconductor had its chip designed in with three Japanese camera makers, but after an investment of $9m, including some debt, it couldn't raise the next round of funding and the IP was sold to u-blox in February 2010.
Peter Claydon did get over that huge barrier of the second round of funding with picoChip, which he set up in 2001 with Doug Pulley to develop a programmable platform for wireless infrastructure equipment. Claydon raised $7m from Pond Ventures and was the chief operating officer until last year.
"Like Air, the A round was the easiest," he said. "Like Audium, we went to the US first, but we met Pond over there and came back to the UK. Since then, it has been a process of continual fundraising. It does help to be completely ignorant and believe that you can do it."
Claydon pointed out that the investor is the biggest customer. "We have been through a lot of ceos; it's become a bit of a revolving door. But if you don't spend enough time on your investors, you can easily have a falling out and, once they decide things aren't going well, the only power they have is to change the ceo."
Like Air and Audium, picoChip didn't have a complete team at the start. "The VCs like to help with that, but the more of the team you have at the start the better," said Claydon.
Nick Sturge was part of a complete team made redundant from Inmos in the early 1990s that set up video telephone equipment maker Motion Media. He now runs the SETsquared Innovation Centre in Bristol.
"There are 4.7m private businesses in the UK and 99.3% have less than 50 people, so we have always been a nation of startups," he said.
Motion Media went to the smaller AIM financial markets as an alternative source of funding, but highlighted some of the pitfalls. "We IPO'd in 1996 and raised £1m to productise our technology. Our shares rose from £1.20 to £27, but then the dotcom bubble burst and each day our shares went down," he said.
"It's very, very different starting up as a group to starting on your own. In a group, your low points will be at different times and you can balance them out. But you have 'group think' and you don't deal with the problems that are there," he said. "Everyone advised us not to buy a US company and we didn't listen. They advised us to put a Brit in to run it, and we didn't listen. Eventually, we downsized quite a bit and merged with an Austrian company."
The team and the board was critical, said Davies of Audium. "The big lesson for me was that we didn't have an independent chairman to hold things together. Then it got difficult; in hindsight, that was a mistake."
And there need to be other ways of funding. "You need to identify sources of funding right up front, beyond traditional VCs," said Claydon. "VCs in the UK will not be very fruitful at the moment." Davies agreed: "The more people I speak to, the more I find there are fewer VCs who are supporters of chip companies," he said. "Perhaps the better way is to fund start ups from large corporations and you can leverage their existing infrastructure."
But all four would happily do it all again – and three of them are.