Custom software: Not so scary

Written by Jennifer Rivkin

When Themis Papadopoulos launched Triglobal Inc. in 1997, he bought the best financial software on the market. Yet he was soon frustrated. The CEO of the Westmount, Que.-based financial-products brokerage found that he was spending more time on administrivia than engaging with clients.

So the former programmer began writing software to ease his paperwork burden. After several years of tinkering, he had a tool that, in just minutes, could spit out reports on his clients’ investments that might have taken days or weeks with packaged software. The result? He could keep more on top of those investments, freeing up time to connect personally with clients. When he saw the advantage this gave him over other brokers in wooing and retaining clients, Papadopoulos decided to turn his patchwork program into a firm-wide initiative called The Harbour: “Software would be the tool [we’d] use to differentiate ourselves from our competitors.”

That tool cost more than $3 million to develop. But it paid off big. Since rolling out The Harbour four years ago, Triglobal has added 650 new advisors to its original base of 50, and assets under administration have soared from $200 million to $1.5 billion. Papadopoulos says brokers flocked to Triglobal largely because The Harbour automates many of their daily tasks, and he credits it with boosting profits by 10% to 15%.

While off-the-shelf programs satisfy most business needs, custom applications such as The Harbour can give you a big leg up on the competition. They can spotlight what’s distinctive about your business, allow you to offer products your rivals can’t match and help you cut costs and raise productivity. And they don’t have to break the bank: some cost as little as $10,000, with only a few reaching the expense Triglobal incurred.

Fine, but what about the risks? A mismanaged custom-software project could devour time and money without yielding any of these benefits. Yet, by planning ahead and following a few simple steps, you can avoid becoming one of the horror stories.

You’ll also need to be clear on the program’s purpose. It might be to allow you to analyze your own data in a way that a standard packaged program can’t, for instance, by customizing an accounting program so you can answer questions about pricing, sales performance or time management. It might be to make life easier for your clients. Bob Fabian, past president of Canadian Information Processing Society Ontario, points to Oakville, Ont.-based Procor LP, which manages Canada’s largest private rail-car rental fleet. Link-Net, a proprietary fleet-management application, gives Procor clients online access to a maintenance report on each tank car showing when it will need servicing. Or your purpose might be far more ambitious. Custom software might become the product or service that differentiates your firm, as The Harbour does by allowing Triglobal to promise brokers fewer hours eaten up by administration.

Some custom-software projects have another mission: to save you a bundle. Jim Koutougos, CEO of Discovery Computers (Canada) Inc., a Richmond, B.C.-based computer retail franchisor, was able to reduce his payroll by developing a custom order-fulfillment system to handle much of the work previously done manually. The tool, not available off the shelf, merges data on suppliers, pricing and product availability into a single, easily searchable source.

Koutougos says it was a necessity for managing fast growth: “You can spend $50,000 or $100,000 on a piece of software, which you’re going to be able to keep and use for [years], or two or three times that on adding staff and not even get the results you want.” The $50,000 project has saved Discovery $15,000 a month in operating costs since its completion earlier this year. “More importantly,” says Koutougos, “it has made running our business easier, and you can hardly put a dollar figure on that.”

But there was a time when he wasn’t singing the praises of the custom route. Koutougos was earlier forced to scrap a custom project he calls “an absolute disaster.” He had made the mistake of trying to create a fully customized application instead of getting the most he could from packaged software and building only the features he needed. The resulting program was too complicated, buggy and way over budget. Koutougos wrote off $90,000 when he abandoned it after 12 months.

Although no figures are available on the flop rate for custom projects, Forrester Research of Cambridge, Mass. estimates that more than half of all software projects fail to meet their business objectives. The failures are seldom technical. “The big issue is getting the right business model on which to base the new system,” says Fabian. “The major off-the-shelf software systems come with a pre-defined business model. With custom software, all too often the users don’t really understand the business model they need, [and] the developers take a best guess.” A happy outcome is far likelier if you define your requirements up front and hire a developer who understands your business goals. And be ruthless in analyzing the payback: if there’s no return on investment within three years, forget it.

Above all, don’t customize more than you have to. If you adapt pre-written software, only use the tools provided in the package. Completely overhauling a program could cause major headaches in a future upgrade and make it virtually impossible to get support from your software provider. Instead, advises Fabian, keep your generic package intact, then develop custom modules that work with it. You’ll find it fairly easy to add these features later to upgraded versions of your packaged software.

In the end, using standard software with customized add-ons beats trying to contort an off-the-shelf package to do everything you need. Following this logic, there’s only one answer to the “build or buy?” question: do both.
Papadopoulos continues to discover more uses for his mined data and has vowed not to stop building until he has automated every process in the firm. “You can’t say, ‘We’ll stop where we are’,” he says. “This is something that will change constantly as the size of the firm changes [and] as the needs of the market changes.” Is that a problem? “I would only consider it a drawback if it doesn’t increase profitability.” And The Harbour certainly can’t be faulted on that score.

© 2005 Jennifer Rivkin

Originally appeared on