Do or die for local developers?

Government identified the software development sector as one of 10 potential high-growth sectors under an older DTI scheme that aimed to identify and assist potential high-growth industries. Government also has a stated policy of using open source software wherever possible, something that met with fierce resistance from closed source vendors, and a policy that many in the industry, seeing their lunch walking out the door, continue to attempt to undermine.

“Government,” says IndigoCube MD Ziaan Hattingh, “is one of the biggest spenders in IT and a large portion of that spend is on software development. If you were going to grow or develop or strengthen a local industry, it would make sense to use that spend to develop the industry. Sita, for example, is spending a lot on its own software development.”

Which begs the questions why it's not spending that with the local companies government aims to develop?

“The software development industry in South Africa has grown immensely over the years and is characterised as one of the top 30 software development outsourcing destinations, having established itself as a quality and low-cost industry leader, according to international research group Gartner,” says Dariel Solutions MD Malcolm Rabson. “However, with software development becoming increasingly more inexpensive and outsourcing more convenient from countries like India and China, the local software industry faces the threat of 'extinction' in just a few years to come. It has become crucial for the South African government sector to intervene in the local software industry and set up strategic initiatives to promote skills development and provide financial support and aid in ensuring that this industry continues to impact the economy in a positive manner and remains functional, thereby not following the same path as our local hardware sector.”

What hardware sector, you might ask. To which one would respond: precisely.

“We need a holistic approach,” says Hattingh. “SA needs to specialise on one aspect, such as Java-based open source technologies. We need a common business forum for all software and service companies. We need to increase the number and quality of software graduates by school initiatives to improve the uptake of maths and science.”

The skills issue he raises is a big one, although just how big is disputed. Pretty much everyone agrees we don't have enough skills, but what's often up for dispute is the quality of skills. Many commentators bemoan the poor quality of maths and science skills coming out of our school system, others the lack of standardisation across tertiary institutions. As Hattingh notes, you have no means of judging the quality of a candidate's degree because it really depends on the university that awarded it.

Prof Barry Dwolatzky, director and CEO of the Johannesburg Centre for Software Engineering at Wits University, says the quality of skills is good, we just don't have enough of them.

“SA isn't a newcomer in the software world. We've been around since the start of computers. Through the 50s, 60s and 70s, we were breaking new ground. Sanctions were fantastic for us in that sense because we got to do what software people love best - reinventing the wheel.

IDC is no different from the other vultures that give you money and take your company away.

Lee Child, founder and MD, Relate Technologies

“Our experience is fantastic, our skills are fantastic, we see South African IT professionals and software developers all over the world, and new ones are snapped up internationally. Our formal skills legacy is the positive side. On the negative side, over the last 15 to 20 years, the barriers protecting the industry have gone. We're re-engaging the world and that re-engagement has thrown up the problems in our world. We're not good at building big systems predictably and repeatedly. A lot of companies sensitive to the risk associated with big projects, like the banks, look to India if they want it done repeatedly and predictably, on time and within budget. If the local industry is going to compete, it has to be world-class.”

Hattingh echoes this: “We're operating in an unprofessional fashion; we haven't professionalised the industry or made it desirable to be in, so we see the industry being mainly a body-shopping industry. Many companies say they are developers but few do project work; they're shop bodies, and when they do project work, it tends to be small. Companies aren't comfortable outsourcing big projects. We don't see local companies with Capability Maturity Model Integration (CMMI) level 5 and highly certified staff. Banks, telcos and government all look at the industry when they've got a big project and go out to get project managers from Joe's Software Development and developers from Sandy's Software Development and testers from India and put them in a room and, boom, they run the project. We lack credibility in our own country. It's a certification issue. We need to put our money where our mouths are and be serious about certification.”

The other common criticism is that for smaller development houses trying to get established, or break into the big time, support is scarce and expensive.

Lee Child, founder and MD of Relate Technologies, thinks a fund that supports innovation and the creation of local IP that is uncomplicated to access and focused on kick-starting creative concepts and technology-driven ideas are what the software industry needs.

Says Child: “We're an SME ICT company with great ideas and products that we want to take international, which have been already well-received there. We now need to beef it up, extend our resourcing, and bring our plans to fruition more quickly. Typically as a start-up, you fund yourself, squirreling profits away so you can fund development. The minute you get some success and need to scale up and get cash, you, from my point of view, start talking to people. You start finding out who the investors are, and how to go about this. Government has plans, it has the Industrial Development Corporation (IDC) and networks in place that enable small companies to get cash.

“The first thing you hit - we're about technology, and have no experience networking around understanding who investors are - are consultants who come in to help you address these challenges. But they want to make money from a slice of the money you raise. So if you need R2 million, they say no, no, you need R15 million to get investors. What about IDC, you ask, but they want a lot - the same as any other investor. They want you in brackets where they can give you a lump of money and if you're looking for an amount that's below their range, it becomes a difficult debate. IDC is no different to the other vultures that give you money and take your company away.”

Child and his partner have spent the past year trying to get funding to grow their business, with increasing amounts of frustration. Ultimately, he says, they've decided to fund it themselves and remain frustrated at having to pass up opportunities because they can't get their hands on the lump of cash they need to increase their developer pool.

Viadata MD Craig Byren has also hit a ceiling, but of a different variety.

We need a holistic approach.

Ziaan Hattingh, MD, IndigoCube

“Businesses use installed client base and availability of resources as a major measure of 'maturity' of product and 'sustainability' of vendor. South African vendors struggle to build up significant numbers in the local market and therefore always perform poorly in RFP evaluation on these criteria. Businesses perceive this to be a major risk factor, even though many local developers with a very small installed client base have been in business for over 15 years and have a small yet highly satisfied client base,” he says.

“To overcome this prejudice may require some form of official accreditation, government or industry backing. In other words, rather than actually providing incentives to businesses, local industry needs to help business mitigate the perceived risk in opting for a local solution, either with some form of guarantee or with some form of local accreditation,” he says, echoing Hattingh's call for the industry to professionalise.

“Organisations like CITi (in Cape Town) are making a strong effort to publicise the successes of the local development industry, but this is not reaching the broader business community throughout South Africa,” Byren adds.

Child says initiatives like Silicon Cape haven't been helpful to his company either. “If you take a cynical view of the layer of consulting guys making money helping you get money, then Silicon Cape is a great networking opportunity for them. It's full of young hopefuls trying to get out there. There isn't a really easily accessible or well-defined angel investor network. You fall in the consulting space of Want Hasso Platner or Here Be Dragons and need coaching and mentoring and help putting a plan together, which the consultants do. So Silicon Cape, as much as we're excited about it and buy in, we were in the mix at the time we were most aggressively looking for funding and it didn't address our concerns and issues. It didn't put us in contact with great people - it put us in contact with the usual crowd.”

To overcome this prejudice may require some form of official accreditation.

Craig Byren, MD, Viadata

Dwolatzky's JCSE has several initiatives aimed at increasing the skills pie (see www.jcse.org.za) and Accsys' CEO Teryl Schroen says there is help available for start-ups and software developers, if you can access it. “Maybe one of the challenges is the fact that there is assistance but it's under-marketed,” she muses.

Either way, industry probably needs to take its own initiative, form a representative body, and work together to get things moving. Waiting for someone else, or something else, to do it is never going to lead anywhere.