Tue, 14 Oct 2014 - 21:00
Viewed

2014 JJC Bradfield Lecture by Atlassian co-CEO Scott Farquhar, "A Start-Up Nation: Capitalising on the Software Revolution in Australia"

Introduction

Welcome everyone, and thanks to Paul Fletcher for inviting me to speak today. I wanted to share that opening video because it highlights how software is changing 4 different industries – medicine, education, automotive and emergency services.

Today I will make the case that software will profoundly affect not just these 4 industries, but every area of our lives. I’ll argue that this presents both a threat to our current standard of living, but more importantly an opportunity for Australia. Finally, I’d like to make some suggestions for how the government can best prepare for this change.

Software isn’t just about software anymore. It’s about everything we do.

Electricity

First, let me take you back to a different time, when a different technology transformed the world. Just over a hundred years ago, it was the adoption of electricity that touched every part of our lives.

Let me take you back to 1908. At Ford’s Piquette plant, the first Model T would roll out. They were in high demand - with a top speed of 70km/h and ability to run on ethanol, kerosene or petrol. But using only existing production tools and techniques, they could only manufacture 11 of them in the first month of production.

Factories up until then were powered largely by steam. One steam engine would power overhead belts and pulleys that would power each individual machine. It was dusty, dirty and dangerous work, and the end product moved around the factory like snakes and ladders - as the machines were laid out according to their power needs, not in any productive order.

Later, factories replaced steam engines with electric engines and saw some mild gains. But new companies like Ford embraced the new technology, and saw that if they gave each worker their own electric tools, then they could rearrange the factory to be more streamlined, so that work flowed seamlessly and in productive order.

It was the electrification of factories that allowed the modern assembly line. This new way of working improved production time of the Model T by a factor of 8 – from 12 hours to 93 minutes, and cut the cost in half.

So, whilst you may not think about it in this way - it was electricity and the electrification of factories that led to the modern assembly line, without which there wouldn’t have been the popularisation of the Model T, and the entire automobile industry.

Software

Like electricity, software provides an order of magnitude more productivity, paving the way for cost reductions and new products, and transforming how we work as a result. Standing at the start of the 20th century it would have been impossible to overstate the impact electricity would have on our lives. Similarly, as we stand here at the start of the 21st, it is impossible to overstate the impact software will have.

Tonight’s lecture honours Dr JJC Bradfield – whose work as a civil engineer transformed Sydney in the nineteen twenties and thirties.

Today, the work of software engineers is not just transforming Sydney, but the entire planet.

With my cofounder Mike Cannon-Brookes I am lucky enough to lead a company, Atlassian, which is riding the wave of this software revolution.

I met Mike when we both studied the same scholarship program at UNSW sixteen years ago.  

We got the idea to establish Atlassian because we wanted to create the kind of company we wanted to work at.

We were very much going against the accepted wisdom. People told us that IT was going to be “outsourced to India” – that it was low value work we were doing.

We began operations – not in a garage but in our bedrooms – in 2002.

It was a pretty classic start-up experience. We had no capital to start with; we ‘bootstrapped’ ourselves with a $10,000 credit card loan. Our first employee, Owen, was a British backpacker who literally walked in off the street. He slept on my couch for a month.

But from our suburban Sydney bedrooms, we were global from day one. Because that’s the way the world works now.

Our first two sales were to the UK. Our next three were to Sweden the USA and Germany. We sold to in 13 other countries including Puerto Rico and Slovenia before we sold to anyone in Australia.

It is the ability for us to sell globally, and to sell online that has allowed us to grow without outside capital from $18,000 in sales in our first year to over $200 million last year. Only 6% of our revenue comes from Australia today.

When I started my business journey, it wasn’t clear to me how businesses have a positive impact. I thought that it was government and non-profits that did all the good.

I’ve since learned that business can be a powerful force to improve society, and I’m proud to share with you some of the ways in which we’ve helped this great country:

•        We sell to over 40,000 companies in over 130 countries. We estimate that over 10 million people use our products, and we’ve brought in over $800 million of export revenue into Australia.

•        We employ over 1,000 staff in high paying, stable, rewarding jobs. Over half our staff are located in Sydney, and we were recently named the best place to work in all of Australia.

•        All our staff members, every single one, are shareholders - each able to share in the profitability of Atlassian - the profit they helped create

•        We have been able to create a foundation where we donate our products for free, give our employees 5 days a year to volunteer to a charity of their choice, and donate 1% of our profit and 1% of the equity in the company.

•        In addition to our international achievements of building schools and libraries for over 200,000 children in the developing world, we have partnered with Social Ventures Australia and the Australian Indigenous Mentoring Experience to help Indigenous students finish their higher education studies and continue on to university.

As a country, we brag about our sporting achievements, but play down our business ones. I’m super-proud that we help not just our employees and customers, but the broader Australian community.

Just imagine the benefits if we had a hundred more successful software companies like Atlassian! Just think how would that benefit Australia!

Who is Atlassian?

What it is that we do - what product do we sell that has allowed us to have such success?

Atlassian started by selling software to software teams. If there’s a global war in software, we make the ammunition. Because software teams are fast moving, often distributed and highly collaborative, they require tools to help them keep track of their priorities, to brainstorm ideas and manage the development process. Millions of people use Atlassian’s products such as JIRA or Confluence to make their teams more collaborative.

We have some incredible customers. We’ve been able to help NASA build and test the Mars Rover. We help Cochlear to make implants. There are at least two customers mapping the human genome, storing their data using JIRA. Pixar uses Confluence to storyboard the scenes for all of their movies.

With such a diverse range of customers I’m lucky enough to have a bird’s eye view of the way that software is transforming the entire economy.

I can say without a doubt that we’re seeing the same kind of change which the industrial revolution brought in the eighteenth century and electricity brought in the nineteenth century.

With such a big change, there is huge opportunity and huge risk.

Every company is becoming a software company (or being disrupted by one)

The first thing everyone needs to know, and I can’t possible overstate this, is that every company is becoming, or already is a software company. Even if they don’t know it yet.

This is the most important part of my presentation, so I’ll state it again for emphasis. Every company is becoming a software company.

It’s software that powers computers, powers robots. Software is the ultimate lever for human performance. As such, like electricity, it can augment us, enable us, even replace us in many areas. Archimedes said, “give me a lever long enough… and I’ll move the world”. Software gives humanity that lever.

Let me give some examples. I’ve already mentioned some of our technical customers - Cochlear, NASA, Pixar. We also sell to banks like CBA, NAB and telcos such as Telstra and Vodafone. It’s probably obvious how software is used in these companies.

But there are other companies we sell to who don’t just use software - they disrupt the status quo using software. We buy TVs online with Kogan. We bet online with BetFair. We buy cars, property and insurance online with carsales.com.au, real-estate.com.au and iselect. We can get a degree online now with OpenUniversities, and then go on to find a job with Seek. Each of these Australian companies is disrupting billion dollar industries with software.

To illustrate how widespread software is, I went through Atlassian’s Australian customer list.

Perhaps it’s no surprise that Jetstar and Qantas are on there and have huge software teams, but you may be surprised that law firms such as Clayton Utz and Mallesons are customers, and are starting to see huge efficiencies through building better software to support their partners.

To go through the list – in Mining: Rio Tinto, Woodside, BHP; Media: News Corp, Fairfax, Fox Sports; Utilities: AGL, Sydney Water, Snowy Hydro; the list of industries and companies being transformed by software is endless.

Doing research for this speech, I was most surprised by the 257 government departments (we have that many?) who have software teams using our products - from the RBA, to the RMS and the RTA. From the Administrative Appeals Tribunal to the Australian Antarctic Division

Take another industry – logistics. Toll Group’s competitive advantage may have once been the number of trucks on the road, but these days their core advantage is software. They use software to optimise their despatch process, use GPS to provide live, up to the minute updates on trailers in transit. They access real-time footage from 300 traffic cameras to route drivers around accidents or traffic. Their customers can access real-time information about their packages using their website.

The improvements Toll creates over the coming years won’t be by having more trucks, or faster trucks. All their improvements will be from software.

My favourite story is a personal one. I’m close friends with my personal trainer, who runs a small PT business in the Sydney CBD. He believes nutrition is the key to achieving your fitness goals. Over the last few months, he has achieved more success with his clients than he has over the past 10 years.

The secret? Not a new diet, but a free iPhone application that makes it trivial for his clients to track their eating. By allowing people to track their macro nutrients, track their exercise, his clients are losing more weight and achieving their goals like never before.

After 10 years, the biggest advance in personal training is an iPhone app.

Let’s project into the future a little. 3D printers can now print metal. Though costly and expensive now, it won’t be too long in the future where my apartment building will have a 3D printer in the basement. If a tap breaks, I’ll just print another one. If I need a baking tray to cook some muffins - I’ll print one. Or a spare part for my car.

Companies now fit into only two buckets: either becoming a software company, or being disrupted by one.

Competition is global and fierce

My second observation is that software makes the world flat. This is both an opportunity and a threat for Australia. Geoffrey Blainey famously wrote that Australia’s history was shaped by ‘the tyranny of distance’.  For generations of Australian exporters, that tyranny was very real – but software liberates us.

This disadvantage started when the first fleet landed. For many years, there wasn’t anything worthy of exporting. Lumber – too heavy. Fresh fruit would spoil. We didn’t have a spice trade. It wasn’t until we farmed sheep that we found an export light and valuable enough to be worth transporting.

Software is weightless. There are no shipping costs. Software liberates us from the “tyranny of distance”.

We “made it” from suburban Sydney. That’s the good news. The bad news is that any other “Mike and Scott” from any suburb in Seattle or Shanghai, Lucknow or London or Hanoi or Helsinki can make it too.

Today, Myer and David Jones don’t just compete against each other, they compete against Amazon. Free to air TV isn’t just competing with Foxtel, it’s competing with Netflix and YouTube.

Even software start-ups are at risk. Australian developed taxi sharing application GoCatch isn’t just disrupting Taxis Combined, it’s also having to compete with much larger funded international competitors such as San Francisco based taxi apps Uber and Lyft.

But the reverse is also true. For the first time in 226 years, Australian exporters have a level playing field! Software allows Australian software companies to compete on the world stage!

And we have. I know of at least 10 Australian software companies worth more than $100 million who sell have the bulk of their sales overseas. 3P Learning who run global education website Mathletics. Ozforex, BigCommerce, Aconex, 99Designs, Campaign Monitor, Freelancer, Invoice2go, Envato. Companies you may not have heard of, but who compete on a global scale.

But this is a drop in the ocean compared to what we should be doing. Someday soon, robots will be doing surgery. Cars will drive themselves. As a society and as a nation, we have a choice whether we want to be a software producer for the world, or a software consumer -missing this whole revolution and left wondering how we are going to pay for it all.

Automation of jobs

The last observation I’d like to share is the changing nature of jobs. This is natural. We no longer employ 90% of the population as farmers, which is a great thing – being a farmer is hard!  We no longer employ lamp lighters to light the streets each night.

The changing nature of jobs is a natural course and by-product of technology. But the difference with software is its ability to replace not just one industry, but almost any repetitive human task, so the scale of the job disruption will be larger than any that has come before.

Previous transitions have affected mainly low-paid, manufacturing jobs. But software automates the knowledge worker - the doctor, the engineer. The types of high paying jobs that Australia needs to maintain its high wages and standard of living.

I mentioned the Model T before - let’s use another automotive example - the Tesla electric car.

Tesla is a car that was designed using software. It is built entirely by robots in an automated, software enabled factory. They have no show-rooms - you arrange a test drive on the internet. You can review it, configure it, and order it online.

The Tesla sends back diagnostic information automatically, so the manufacturer can diagnose failures, before they happen. The dashboard has one large software powered touch screen, replacing a myriad of buttons and knobs.

When Tesla’s San Franciscan customers complained that when they took their foot off the accelerator, it rolled backwards on hills (like a manual car), they shipped an update “over the air” which fixed the problem.  No recall. No mechanic. A software update.

It is so reliable, they offer an 8 year, unlimited kilometre warranty. It’s been rated “the very best car ever driven” and the “safest car ever tested”.

So whilst Australia is closing down its automotive manufacturing plants, Tesla designs and builds their cars in California - one of the most expensive labour markets in the world, because of the efficiency of software.

Like the Model T, this car is only possible by transforming the manufacturing process. Instead of electricity, it is software driving this revolution. Except this time there aren’t thousands of workers building cars - there are people writing software, and there are robots building cars.

It’s obvious what this means for auto workers - we’re already seen that decline in Australia. But what does it mean for mechanics when cars diagnose themselves? For the entire petrol industry when electric cars can be filled up for $10?

Let’s take it even further. In what was once the realm of science fiction, Google has tested a self-driving car in California that has logged over 1.1 million accident free kilometres.

I have two young children. I worry about their safety as they grow up and learn to drive. I worry about drink driving, about speeding, about dangerous drivers. There are 1200 deaths each year on our roads. In a driverless world, I won’t have to worry if my kids will be coming home.

But what will that mean for the large numbers of people currently working as bus drivers, truck drivers, couriers, taxi drivers, hire car drivers, delivery drivers and the like? The latest Census data reveals that 266,000 Australians are employed in these occupations.

According to a recent study by Oxford University, nearly half of all jobs are at risk of being lost to computers and robots. Gartner predicts one in three. Either way we need to prepare our children, our economy, for the world of software and robots. It’s not the distant future, it’s now.

Australia has a choice. We can lead the world, or be led by it

I think it is clear that the stakes are high – for any country that wants to build or maintain its prosperity as software transforms the world. We need to do something now!

In fact, (and I wish I didn’t have to say this) I think if we were getting started with Atlassian today Mike and I would not have stayed in Australia: we would have gone straight to the US. In other words, I think this problem is getting worse not better.

Some countries, clearly, are doing very well. We need only look at the number of tech sector giants based in the US or China.

It's probably obvious to us why these two are the tech sector giants (size does matter), but did you know the country with the second largest number of companies on the NASDAQ has a population of less than 8 million? Any guesses what country that is? Israel.

We don’t let size stop us when it comes to global athletic competitions. We should take a page from Israel's book and lead in business too.

The well-known book about the Israeli tech miracle, Start-Up Nation, discusses the ‘classic cluster’ of conditions for an innovative economy:

“It consists of the tight proximity of great universities, large companies, start-ups, and the ecosystem that connects them – including everything from suppliers, an engineering talent pool, and venture capital.”

I am very confident that we have these necessary elements to build such an ecosystem in Australia. We already have high quality universities. We already have a talented labour force. Once we get up to speed, it will be self-sustaining.

Talent attracts talent. Cities become centres of gravity. People can live anywhere, but want to live where there is a critical mass of jobs, which requires a critical mass of people. Los Angeles is the epicentre for movies. New York and London for finance. San Francisco and Tel Aviv for technology.

We can also build a tech centre in Australia - and I’d argue Sydney. As we get more examples of globally successful software businesses based here, and they became better known, that will build confidence. Young Australians will think – correctly – that they too can build a global business from Australia, and we will attract talented people from around the world to work here.

As a nation, we have the opportunity to be a tech giant in a world increasingly dominated by tech, and we have the risk of missing the boat entirely in a universe where that very same technology is reinventing traditional industries and with it the jobs in those traditional industries.

I now want to share my recommendations for the government.

ESOPs

I am going to be brave by talking about tax. And I say brave because whenever you talk about tax, and in particular option schemes and the like, everything you say can be seen as self-serving.

But the simple fact is that tax matters in attracting people, tax matters in attracting capital and in a global market for both you simply can’t afford to be too far off the pace.

Credit where credit is due – Australia has some amazing tax incentives in the form of The R&D Tax Incentive Program, and the Export Market Development Grants scheme (EMDG). R&D in particular is a godsend, and one of the only reasons we have a burgeoning development centre in Australia.

But there are some brain-dead things we do that send young companies overseas before they have begun. There are two very simple legislative fixes that should be done immediately.

First we need to fix the tax treatment of employee share ownership plans (ESOPs). These are a vital tool for attracting and retaining talent in the software industry. If Australian companies cannot meet the world standard for rewarding our people, how can we compete on the world stage?

Let me explain what ESOPs are, and why the current system is broken. ESOPs are most commonly stock options offered to employees, the rights to company shares, which accrue in value as a business grows in wealth.

For an employee who is considering whether to join a startup, the option to share in future riches is needed to lure staff from the big banks and large corporate firms, or from start-ups overseas.

Now, in most countries, options are taxed in the year in which the option is exercised. However, following changes under the previous Government, we now have a bizarre situation. Options are taxed in the hands of employees at the time of issue, rather than at the time they received the proceeds.

And to be very clear about this, what can happen to an individual is that they pay tax, in cash, when they are issued options, and then the options may prove to be worthless. Yes. You heard me correctly.

At our company, we decided it was vital to offer options - so vital that we paid the tax on behalf of our employees. So far it has cost $5.4 million dollars in tax to do so. If Atlassian were located in any other country, those millions would have reinvested to grow faster. We made it a priority and were lucky that we could afford it, but many other companies are not in the same fortunate situation we were in. It is a huge disincentive to stay in Australia if you have a tech start-up.

The current proposals to introduce new tax brackets and limit the definition of companies eligible for tax breaks on employee share schemes are over-complicated. In my view we should follow the US model: tax on the options is deferred until the options are exercised and the employee actually makes money.

Venture Capital

Securing top talent is only one part of the equation of course. Access to capital is another critical part. Although the activity level of angel investors and VCs has ramped up – there is still a significant gap in the market in Australia. Particularly once you move beyond an early stage, it is too difficult for a company to raise finance locally.

To my mind, one obvious question to ask is, how can we encourage Australian superannuation funds to invest more into local venture capital. After all, the superannuation savings pool is around $1.8 trillion. Yet only 0.0006% of our superannuation funds are invested in local venture capital. Contrast that with the US, which commits around 2% of pension funds to venture capital as an asset class – representing a 350 times multiple relative to Australia.[1]

So we have one of the best savings pools in the world, and we should make it easier to invest it in areas that will have huge productivity gains for the nation. Instead the government went the other way - they made it even harder for superannuation to invest in Venture Capital.

The new Superannuation disclosure laws require every asset to be reported on a “look through basis”. What this means is that if a $1 billion superannuation fund invests $10 million in a VC, who invests $100,000 in a start-up, the superannuation fund needs to disclose the exact funding details for something that is 1/10,000th of the fund!

I’m hugely in favour of better disclosure! But this is madness. It would turn every VC backed startup into a quasi-public company!

VC funding details are very sensitive for companies. There are many downsides to this disclosure, including the simple one - you don’t want your competitors knowing the details of your last funding round!

There is a very simple solution – have a materiality clause. Any investment made by a VC is likely to be so immaterial to an overall superannuation fund, as to be a non-issue.

Without this, we will further erode the capital base, see the funding move offshore, and we will see more of the promising start-ups of tomorrow follow.

Importing Talent

Beyond the simple and obvious changes around ESOPs and Venture Funding, my second recommendation is to make it easier and cheaper to bring amazing talent to Australia.

Look – I can’t wait for the day when we have enough trained Aussies to meet the employment needs. But today, frankly, there aren’t enough trained people to go around.

Take Atlassian for example. Our Sydney-based head of engineering is from Ottawa. Our Sydney-based head of infrastructure is from Boston. Our head of products was from San Francisco, and recently returned there. If we could find these people in Australia, we would have.

Instead we've had to become quite creative in trying to meet our talent needs. This year we staged a road show in Europe – we called it “steal your geeks” and drove a bus around Europe - to hire 15 developers and relocate them to Sydney, all expenses paid. Every other Australian software company is the same - there aren’t enough people here to meet the demand.

We have two choices – employ staff overseas, and potentially move the company there, or bring those talented people here to Australia where they can train our staff, and contribute to the local economy.

To bring people to Australia, we take advantage of a great program called the 457 visa, which includes safeguards to protect Australian jobs. But there are two problems that need to be fixed.

The first is that whilst larger companies like Atlassian can work through it, the process is so byzantine that almost everyone needs to hire employment lawyers to handle it. Requirements to “train” Australian staff seem obscure to start-ups who are heads down trying to make payroll.

It is so bureaucratic that is often easier to move the entire company to the USA than to bring qualified staff to Australia.

There should be a special class of 457 visas for software start-ups that are fast-tracked, and have a simple online form to apply. We have to make it easier to bring people to Australia than it is to move overseas.

The second problem is that of expense. Some key terms of the 457 visa mean really add to the cost of living of people who come to Australia. Even though they pay the same taxes, they get no access to Medicare so they need very comprehensive private health insurance, and if they have school-age children they are required to pay fees for those children to attend a state school.

These costs are in addition to the high cost of housing, and the need to often maintain housing and pay taxes in the country they moved from.

We have brought many people over under the 457 program, but have also lost quite a few who have returned to their home country for family or because the cost of living in Sydney is too expensive.

Many of these costs used to be offset by the living away from home allowance (LAFHA), which offset some of these costs from the local tax they needed to pay. Unfortunately this benefit was recently removed, with the result that many of our overseas staff had to move back home.

We need to reinstate LAFHA, so we can attract people to live in this country, and grow our talent pool. It’s not just about attracting staff, it’s also about not losing the future start-ups of tomorrow.

Education

If importing talent helps us build a base short-term, the most important priority long-term (and my final recommendation) is our education system. We must give our kids the opportunity to learn programming concepts as early as possible as during their primary and secondary studies.

Some may say that not everyone will be a programmer. And that’s true. Not everyone who learns English goes on to be a poet. But we teach English because communication is a fundamental skill!

Similarly, computing is a fundamental skill for surviving in a software world.

I was lucky enough to go to James Ruse Agricultural High, one of the top selective state schools in the country. The historical rationale for selective agricultural high schools is obvious: agriculture has long been central to our prosperity.

But today, an even stronger case can be made for an intense educational focus on developing computational and software skills. More and more nations are making this a priority.

In the United Kingdom, starting this year, public school students will be required to learn coding at age 5 and programming languages at age 11. Similar programs exist in France and Finland. In Estonia, 100 per cent of publicly educated students will learn how to code starting at age 7 or 8 in first grade, and continue all the way their final year of school.

Estonia. Estonia is kicking our ass in software education for their children!

In Vietnam, computer science begins in year four and continues as a mandatory subject through to final year of high school. It’s because this investment in education bodes well for the long-term health of the country, that Atlassian has opened an office there, and in a little over a year we’re hired 85 staff. 85 jobs that could have been in Australia.

At the tertiary level, the trends locally are going in the wrong direction: enrolments in computer science are actually falling. Despite demand for ICT workers in doubling over the last 15 years, applications for tertiary ICT courses has more than halved over the same period.

This lack of university interest probably starts at the school level, where there is no uniform national requirement to teach computer and software skills. Fortunately, Digital Technologies is currently being considered as part of the national curriculum review – this would be the nation’s first coordinated effort to teach computer science in every primary and high school.

Naturally, teachers would be worried about having the skills necessary to teach programming. This is a reasonably worry.

But let’s not teach our kids software using 19th century teaching methods. Online teaching has to be part of the answer. Look at the success of code.org - in its first year, 33 million students tried “Hour of Code”, a one-hour introduction to computer programming. Code.org have built 20 hours of tutorials, in 30 languages. In the US, code.org has partnered with 30 districts to add computer science to the full curriculum.

We need to make software and programming a mandatory part of the national curriculum as early as possible in our children’s education.

Conclusion

In closing, every company is becoming a software company - whether they know it or not, whether they like it or not. It will transform everything we do in a way not seen since the industrial revolution.

In this change, as a country we have two choices. We can ignore it, and watch it erode our global competitiveness, or embrace it, and, like on the sporting field, punch well above our weight on a global basis.

Now is the time to move. There are simple changes the government can make today with ESOPs and Venture Capital. We need to create critical mass quickly, but we’re competing with other global cities to do it. Biggest thing is to create a talent base, by allowing and supporting skilled migration. We HAVE to invest in education. Or in 10 years you won’t have Atlassian or any other company based here.

It may seem big, and it may seem like we’re late to the party. But I’m reminded of a Japanese proverb:

“The best time to plant a Bonsai plant was 20 years ago. The second best time is today”

With your help we can build a world class software industry in Australia today.

Thank you.

Paul Fletcher MP

Parliamentary Secretary to the Minister for Communications

Commentary on JJC Bradfield Lecture by Scott Farquhar

Sydney 13/10/2014

I am very pleased to have this opportunity to comment on the JJC Bradfield Lecture by Scott Farquhar.

At the outset, I want to congratulate Scott for an insightful and exceptionally well informed lecture. Scott is in an extraordinary position to speak about the software revolution and Australia’s place in it, and it really showed.

In my commentary, I want to highlight four points which struck me from what Scott has put before us this evening.

Australian Heroes

First, something which Scott could not say: in my view Scott Farquhar and his co-founder Mike Cannon-Brookes are Australian heroes.

Scott gave us the facts: they started the business in 2001 straight out of university, and today it employs over 1000 people around the world, over half of them in Sydney. They have brought in over $800 million in export revenue since the company got started.

Let me give you the colour. This is an extraordinary achievement. They have built a scale, global business in a ruthlessly competitive sector. As Australians we should celebrate what they have done.

As Scott rightly observes, as a country we brag about our sporting achievements but play down our business ones.

This story needs to be widely told and widely known: going well beyond the business pages. Many Australians know how Mark Zuckerberg started Facebook in his dorm room at Harvard and Google founders Larry Page and Sergey Brin met doing computer science at Stanford. The story of Scott Farquhar and Mike Cannon-Brookes should be just as well known.

One reason it needs to be well known is to inspire and encourage other Australians who might be considering starting a tech sector business. The more who take that plunge, the better our prospects of building the self-sustaining flywheel as he puts it.

Importance of Individuals

My second observation is that Scott has given us a really sharp reminder of the importance of individuals, of entrepreneurs, of risk-takers, of the private sector, as the mechanism by which prosperity is generated out of technological innovation.

He told us about their first employee who slept on his couch for a month. In an early conversation with Scott about his ideas for this lecture, he told me about going to their first trade show in San Francisco in 2003: three of them shared a hotel room at $80 a night.

This is a really critical point about success in the tech sector: individual entrepreneurial effort is critical and it is not something government can replicate. If you will forgive me for being political for a second, the previous government seemed to wish it were different.

Its default approach was that if a sector was vital to the economy, then we needed to pump billions of taxpayers’dollars into establishing a state owned enterprise to dominate the sector: hence the establishment of the $43 billion NBN and the $10 billion Clean Energy Finance Corporation.

I have not checked the records, but I rather doubt there are any recorded instances of three NBN Co executives bunking down together in an $80 a night hotel room.

The story of the tech sector globally is that start-up businesses which succeed can create a lot of wealth and a lot of jobs in quite a short time. As Scott reminds us, Atlassian as well as some other Australian companies have demonstrated this principle.

He observed that at least ten Australian tech companies worth more than $100 m have the bulk of their sales overseas.

A paper issued by Google earlier this year noted that nine of the top ten people on the 2013 BRW Young Rich List made their fortune in the tech sector.[2]

This is great news for the founders of startups who hit the big time. But it is also important for the broader economy, because of the outsize role of startups in job creation –and in turn the disproportionate role of the high-tech sector in generating startup companies.

A recent report from America’s Kauffman Foundation found that new business formation was 23 per cent more likely in the high-tech sector of the US economy than in the private sector as a whole. Moreover, in information and communications technology (a sub-set of the broader high-tech sector) it was 48 per cent more likely.[3]

The importance of startups was also noted by the OECD in its most recent Science, Technology and Innovation Report: one third of job creation in the business sector comes from young firms with fewer than 50 employees –even though these make up only 11 per cent of total employment.[4]

Serving Businesses

The third thing I found noteworthy in the lecture was the insights into the Atlassian strategy of serving businesses rather than consumers.

Scott pointed out that Australians can compete on the world stage in software: but you need to be good because every other country can do the same.

One key issue we face in Australia is a small domestic market. This is a significant challenge when it comes to building market scale in consumer applications. It is surely no coincidence that the largest consumer facing software and software-based businesses, such as Google, Facebook and Apple, come from the US, a country with one of the largest consumer marketplaces in the world.

But the Atlassian story suggests that in business to business software, this is not such an issue; just as the experience of German-based SAP over the past thirty years also suggests.

People with Skills

The final point which stood out for me in what Scott had to say was the importance of public policy settings in a number of areas to underpin our prospects of being a nation with a strong software industry.

He had some very important points about employee share ownership schemes and venture capital, but the point I want to focus on is the need for the right skills in the workforce.

Clearly when it comes to having enough skilled people to support a software industry, two critical policy levers sit squarely with government: education and immigration.

As the Chief Scientist Ian Chubb noted in a report issued last month:

The global economy is changing. New technologies and smart companies lead…Nations at all levels of development are now focusing on the skills required for building new jobs and creating wealth….At the core of almost every agenda is a focus on STEM: science, technology, engineering and maths.[5]

This is important both at university level, where the Abbott Government has a major focus on freeing up universities to be more flexible and responsive to market demand, but also at the school level, as is demonstrated by some of the examples which Scott gave about what other countries are doing.

Clearly this is a timely point to be raising, when the Review of the Australian Curriculum was released yesterday and Education Minister Christopher Pyne has stated that there will now be discussions with states and territories and other key stakeholders about how to strengthen and refine the curriculum.

Immigration policy settings also emerged strongly in the lecture as a limiting factor on getting the right skills. Certainly the Abbott government is setting immigration policy with an eye towards meeting the skills needs of the economy. In this year’s immigration programme, nearly 68per cent of Australia's migration places are allocated to skilled migration, and there is a reprioritisation towards employer-sponsored visas.

Conclusion

To conclude, tonight Scott Farquhar has given a detailed and well researched lecture explaining the software revolution and what it means for Australia; and setting out some things that need to happen if Australia is to best capture the opportunities it presents.

Scott has put a lot of time and effort into this lecture, no doubt because he sees it as part of his role as a leader within this sector - and he and Mike have consistently made time to give leadership since the early days of their company.

He has given his point of view - not cleared by me, I hasten to add - but that is exactly what I wanted him to do. In my view Scott has had some serious and important things to say, and I am very grateful that such a busy tech sector business leader has taken the time to give his frank and unvarnished views. It is critical to sensible policy development in this area that government is in no doubt about what the sector is saying. Scott - thanks for leaving us in no doubt!

[1] StartupAus Crossroads report, p49]

[2] Start with Code: Australia’s Innovation Generation, 2014.

[3] Tech Starts: High-Technology Business Formation and Job Creation in the United States, Kauffman Foundation, August 2013, http://www.kauffman.org/what-we-do/research/firm-formation-and-growth-series/tech-starts-hightechnology-business-formation-and-job-creation-in-the-united-states

[4] OECD Science, Technology and Industry Scoreboard 2013: Innovation for Growth. Executive Summary, p 13.  http://dx.doi.org/10.1787/sti_scoreboard-2013-en, downloaded 20/7/14.

[5] Australian Government, Chief Scientist, “Science, Technology, Engineering and Maths: Australia’s Future,” September 2014, p 5.