Waterford Institute of Technology (WIT) spin-out Structure101 is focused on improving the efficiency of developers by organising their source files into a cogent ‘software architecture.’
“Software development is hard enough, but when a codebase grows beyond a certain size, developers start drowning in the complexity of an ever-expanding sea of interdependent source files,” explains Structure 101 co-founder Chris Chedgey. “This problem is pervasive and costs the software industry billions every year.
“Typically, a huge sum has been invested in a codebase that works, but is increasingly expensive to maintain and extend. In this case, organising the existing code should involve much less cost and risk than struggling on … or starting over.
“But making this possible has required an entirely new kind of technology – technology which we have invested man-decades developing, and that we have now released as Structure101 Studio.”
Structure101 is targeting what it describes as ‘a modest slice’ of the software development tools market.
IDC estimates there are 11m developers globally, and Gartner put the development tools market at US$9bn.
“We can take some of that,” Chedgey says.
Chedgey has an MSc in computer science from Trinity College Dublin and 30 years of experience in the software industry, notably with large military and aerospace companies in Canada, where he lead the development of the controls software for the International Space Station’s robotics systems.
“I am a strong advocate for the principles of modularity, and have spoken at dozens of conferences and user groups in Europe and the US.”
Co-founder Paul Hickey has a BSc in computer science and mathematics from University College Dublin. He started his first computing job back in 1985 at Generics Software in Dublin (one of the first Irish ‘start-ups’), working with Chedgey on a research project – the application of visualisation and formal methods to software development.
“It seems visualisation was always part of our DNA. Paul has since lived in Sydney and Boston while working for Wormald Advanced Systems Engineering, Lotus Development Corporation and Iona Technologies.
“In 2001, we hooked up again, in pursuit of our passion for reducing the complexity of software development projects through the use of advanced visualisation techniques.”
Structure101 is an ‘architecture development environment’ that guides the software architect in the organisation of the thousands of source files in his or her codebase into loosely-coupled containers.
This is made possible by an interactive, visual model of the system components and interdependencies designed specifically for this purpose and unique to Structure101.
“From the model, the architect creates a series of diagrams which communicate the architecture to the team, so that code changes that violate the rules expressed in the diagrams are flagged immediately.
“Being at all times explicitly mapped to the source files, and continuously verified as the code evolves, the Structure101 architecture is at all times in sync with the code, and the architects in sync with the programmers – a dream come true for software architects.”
Realising the dream for software architects
“Our goal is simple – to make architectural control real and standard practice across the software industry,” says Chedgey.
“A vast amount of source code exists today, close to all of it is a horrible tangled mess. It’s simply not humanly possible to manage the architecture of a codebase without the right tools. The potential value of a technology that can organise all that code into a form that is substantially easier to understand and change is incalculable, but very big.
“Structure101 is well positioned to turn this potential into a reality.”
Structure101’s progress in this regard, Chedgey says, is proven by thousands of customers who use it to control the architecture of billions of lines of code.
The technology has won two Jolt awards, and a JAX award for innovation – highly valued and competitive international endorsements.
“Most of our sales come over the web, but an increasing proportion also comes from technology partners who sell complementary technology with which we integrate Structure101.
“These guys have the global sales force and channel to sell at the enterprise level, which means bigger deal sizes. Last year, we signed up our third partner, and we are aiming to increase revenues from channel partners to about 50pc of total this year.”
Chedgey says the company has stuck to a model of growing organically. “Growing organically without investment is hard work, but things are really looking good right now.
“We launched a major new version of the product last year and increasing revenues allowed us to double our team to eight. This growth is accelerating in 2014.
“Our model of a relatively small, primarily technology-focused company, with a light web-sales model combined with strong partner sales channels is working well for us. It lets us get on with what we’re good at. So I wouldn’t rule out a round of investment at some point, but it will be at a point when it is an easy decision for us, and for the investors.”
Asked if the company has faced many challenges along the way, Chedgey doesn’t hold back. “Yes, many! Structure101 is our second kick at a similar can. It is born from a venture called Headway Software that launched in the dot.com days of 2000 with substantial VC and private investment. This crashed and burned in 2004, but Paul and I still believed in the dream and really wanted to keep it alive.
“So we spun into the Waterford Institute of Technology and took the time to redevelop the technology on the back of what we had learned with Headway. As well as building a better product, we positioned it lower-cost and web-sellable.
“This is what we felt the market was looking for, but also it would allow us to grow the company without building a salesforce, and so without substantial investment. It worked a treat – in 2007 we spun out of the WIT with sales literally doing the hockey-stick thing. Until late 2008 when the global economy imploded and the hockey stick inverted.”
At this point, the Structure101 team decided it was going to be hard to sell the product in those grim times – customers were questioning every expenditure, no matter how small.
“Our product measured the messiness of a codebase, and let you define how it should be, but it did very little to help you get from one to the other. We badly needed a product that actually did something more than just depress our prospects,” Chedgey says.
“We had accumulated a reasonable war chest during the good times, and if they had continued we would have started spending it on more sales and marketing with reasonably predictable impact on revenue growth. But we were pretty sure if we did that in 2009 we would clear the bank account very quickly without the required impact on revenues.
“So we decided instead to put pretty much all our focus and resources into product development – we hired engineers to build the product to fill the gap we knew we needed to fill.
“We very nearly didn’t make it to release of the new product. Enterprise Ireland funding that looked likely, briefly, fell through – they required matching funding from investors (in 2009!) In the end, Paul and I went to the bank for personal loans to keep afloat until the new product shipped. It was tense but it shipped, and revenues picked up on the back of it.
“By the end of 2010 we had the product that could transform pretty much any big ugly codebase – this was true, demonstrable value.
“Then it was time to start promoting it in anger. In the developer world that doesn’t mean handing cash to a marketing company. It means getting on the road and talking at every conference, user group meeting, company lunchtime talk, wherever you can get in front of a few developers who might find what you say interesting enough to tweet, blog, or even buy the product.
“Paul lined them up, and I got on the planes and knocked them down, week after week, for almost two years. And it worked. Now a sizable portion of our site visitors get to us by Googling ‘Structure101’, which means we’re known. And importantly, the authors, thought-leaders, and our customers mention Structure101 regularly, and that means we don’t need to do it all ourselves.
“With a level of recognition and the recent release of Structure101 Studio, we’re hoping for a year that is just as busy, but in an exciting high-growth phase kind of way.”
No distractions, please
Chedgey believes Structure101 is more disconnected from the Irish start-up scene than most start-ups would be.
“Fundraising can easily become a full-time activity which totally distracts you from the business of actually getting started. I take the view that if there isn’t a quick indication of likely success, I am better putting the energy into getting real revenues out of the business one way or another, at least it’s somewhat in my control. Our brief exploration of funding opportunities in 2009 were not promising.
“Our early research and development in the WIT was supported by Enterprise Ireland, and frankly we would have struggled to get out of the blocks without that. But once we actually started as a commercial company, EI support was conditional on matching funds, and getting VC funds for anything was a non-starter when we most needed it in 2009. Although funding for even a couple of engineers could have made a huge difference at times, we felt it was too risky to invest the one scarce asset we did have – our time.
“So I’d say the start-up scene helped us get to the first crucial rung, but no further.”
According to Chedgey, software start-ups all have the same problem. They need to build a product and they need to sell the product.
“The problem is deciding when to start selling. Nothing gets rid of money like trying to sell a product that isn’t ready. We made this mistake once. Obviously big cash up front can kick some projects into orbit. But the odds of this happening are minuscule.
“In most cases, projects would be better to spend more time spending less money, getting the product and sales model right, until it is reasonably clear that any fuel you put into the engine will move it forward without too much resistance.
“You might start slower, but the odds of success are higher, with more control in the hands of the people that know the space and love the business – you!”