So you want to start a software company

Starting a software company isn’t an exact science, but Ron Baecker says he knows how entrepreneurs can get past the experimental stage.

The Bell University Laboratories chair and University of Toronto computer science professor,

who offered his advice to a meeting of Communications and Information Technology Ontario (CITO) this week, says timing is crucial for startups who lack the clout and cash of the Microsofts and Oracles of the world.

“”The idea is to find a niche that can be defined with precision,”” he says. “”That have customers that you can access, that’s growing fast enough and is not already owned by an established vendor.”” He cites capacity planning software for mainframes in the 1970s and tax packages for Canadian accountants in the mid-1980s as good examples.

Many of the best-known incubators, like StartingStartups and NRG Group, have turned away from growing early-stage companies, while the economic downturn may have scared off IT managers from branching off on their own. For those still willing to take the plunge, here are some things to keep in mind:

You are not alone in the sandbox

Baecker says poorly chosen markets have one or more of the following characteristics: they’re too vague (i.e., software for doctors), too large (word processing), premature or past their prime. Some are already dominated by others or a “”category killer”” (Windows) or in the sights of major players.

“”It has to be large enough, or potentially large enough, for growth and profitability, but small enough to be affordable and to discourage competition,”” he says. “”You don’t want it to be so huge that all of a sudden Microsoft says, ‘Gee, we better be in there.'””

Matt Golden agrees. The vice-president of business development for Toronto-based Tira Wireless, a Java 2 Mobile Edition applications publisher, says smaller prey need to look over their shoulder to see if they’re being stalked.

“”Some of the larger players might take a big shift into that space thereafter and either gobble you up or crush you, depending on how they feel that day,”” says Golden, the former director of new ventures at the venture capitalist firm Brightspark Ventures in Toronto.

“”But it worked in the lab!””

A well-defined market alone isn’t a licence to print money, of course. There is still the business of creating and delivering reliable software. Baecker says the inability to deliver software on time and on budget is the death knell for startups, but he understands how things can go wrong. Writing code is a creative process, he says, and anticipating all the problems is impossible. The best way to estimate how long a project will take is to break the project into pieces and estimate the time required for the individual elements and integration.

Reports also play a key role, and filing one a week isn’t obsessive. “”If you have a decent reporting mechanism, it’s not going to take people more than 10 minutes to write a report,”” Baecker says. “”They’ve got to write reports, they’ve got to tell us how they’re doing””

Baecker says a startup must be able to convince potential customers the quantitative and qualitative benefits of the software. A demo version of the product is a good way to show it off, as is finding beta testers. While it may sound difficult for an unknown company to find beta testers, it’s not impossible if you’ve done your market positioning properly.

“”If you have something that people need enough, you can sometimes convince them to try your software, even though you are a new company and they don’t know you,”” Baecker says. “”Particularly in some large organizations, there are individuals within it who are willing to take risks because they think, ‘We’re a big company, we should help stimulate new, small companies.'””

To market, to market

Once the software is working the last step is getting your product in the hands of as many consumers as possible. That doesn’t mean you should focus on the technology itself.

“”What you need to do is boil it down to one single statement. How do you differentiate from other people? How are you better?”” he says. “”Customers don’t buy features. They buy value and intangibles, like reputation service, price, interface, documentation.””

Without hard and fast rules for starting a software company, Golden recommends operating with a Zen-like philosophy: equal parts focus and flexibility.

“”A lot of companies start with an idea and then they stick to it no matter what, even though the landscape around them is changing and clearly it makes sense to head off into another direction, but they’re stuck to that original idea and passion,”” Golden says.

“”Your business plan should be ever-evolving. Those who write a business plan believe it’s the perfect business plan and keep polishing it without changing are pretty much doomed in my mind.””


Share on LinkedIn Share with Google+