Implementing an enterprise resource planning system sometimes looks like an irresponsible gamble. So many high-profile implementations have gone wildly over budget and schedule, and more than a few — some worth tens or hundreds of millions of dollars — have simply crashed and burned.

That didn’t

stop All Weather Windows, however. Two years ago, All Weather, an Edmonton-based custom door and window manufacturer, damned the torpedoes and went full-speed ahead with a project to install a full Oracle ERP system — despite industry estimates at the time suggesting that fewer than 25 per cent of ERP implementations were a success.

Today, All Weather is starting to see the return on its investment with increased profitability, reduced staffing and improved customer service. “”It’s really working now and starting to pay off,”” says project champion Henry Banman, a co-founder and executive vice-president of the company.

So how did All Weather beat the ERP odds?

It helps to understand where the company came from. Founded in 1978, it grew from one plant in Calgary to seven sites across the country. It now has over 700 employees and brings in just under $100 million in annual revenues. All Weather produces about a million doors and windows each year.

In 1999, when management began investigating ERP, the company was in dire need. It did have an aging custom-built order entry system and other basic business systems, but nothing was integrated. Employees were constantly having to re-key data which created too many opportunities for error. More important, managers couldn’t get information they needed to make good business decisions.

“”A little business that does a million dollars a year, maybe you can control it manually like that,”” says Banman.

“”At some point we realized that we couldn’t grow All Weather into the bigger company we wanted unless we had information that let us know what we were doing and what we’d done.””

The company lacked a detailed, up-to-the-minute picture of when orders were required, what materials were needed and when, and how much pre-built and partially-built stock it had in hand. Sure signs of problems were the burgeoning requirement for inventory space and high production lead times.

The company knew an ERP system in concert with an integrated third-party front-end order entry system could solve its problems. The question was: could All Weather escape the ERP jinx? Was it worth the risk?

ERP analyst Brian Zrimsek, a research director at Stamford CT-based research and consulting firm Gartner Inc., doesn’t believe many companies steer away from ERP because of scary stories about failed implementations.

“”What that reputation has done, though, is to make people much more business case-conscious,”” Zrimsek says. “”And much less aggressive with the scope of their projects.””

There has been a downturn in the ERP industry, he concedes. Total worldwide ERP sales have dropped from a peak of $7.4 billion in 1998 to a projected $4.6 billion in 2003. But that has more to do with market anomalies and product life cycles than fear of failure. And Gartner is predicting steady if gradual growth in sales through 2007.

Companies certainly shouldn’t balk at ERP for fear of failing, Zrimsek insists. For one thing, many ERP projects were only counted failures because they went over budget and schedule. Many still delivered valuable business benefits in the end.

“”Besides,”” Zrimsek says, “”it doesn’t matter if every other ERP project is a failure. It still doesn’t mean yours will be. If you drove the same car as your neighbour and one day your neighbour drives into a tree, it doesn’t mean the next day you’ll run into a tree. Good project management always overcomes fear of failure.””

All Weather spun its wheels on the ERP project for months until it brought in director of IT Gerald McLennan, a veteran of the ERP wars, to make it happen. Six months later, on January 2, 2002, McLennan did just that, with a company-wide big-bang cut-over to a full implementation of Oracle 11.

The system has gradually given the company the kind of visibility of operations that it lacked. It could now schedule more tightly. Production lead times for orders, running at 25 days “”and climbing”” in 2001, are now down to 14 days and falling.

Shorter lead times also allowed the company to reduce inventory of completed orders. Now it has time to take an order out of the schedule if a contractor has delays. Because production managers know more clearly what they’ll be manufacturing on a given day, they don’t need to keep as many raw materials standing by.

All of this has increased productivity, while the company has reduced head count. Staffing in manufacturing has gone from 549 to 417. Revenues and profits, meanwhile, are up 10 per cent this year so far — some of it attributable to the ERP system, Banman says.

So how did they do it? McLennan’s tough leadership was crucial, Banman says. Equally important, McLennan says, was full-bore commitment from upper management.

So passionate was Banman’s commitment that he now says, “”If it wasn’t going to be successful, I probably would have resigned as a partner.”” Company president Gord Wiebe and CEO Harry Buhler were also solidly behind the project.

The absolute necessity of getting buy-in from senior management is one of several key lessons McLennan has learned from bitter experience about how to do ERP. “”It’s critical”” he says, “”that executives say to employees, ‘We’re changing. Get with it. This is a go!'””

The need for committed sponsors shows up on a list of “”Six Critical Success Factors for Implementing ERP II”” in a Tactical Guideline Zrimsek authored for Gartner last year. The other five success factors in Zrimsek’s list: manage expectations, know

your scope, pick the right approach, manage change and avoid customizing the package. McLennan echoes many of them.

Another of his rules of engagement was that the project had to be short in duration — six or seven months. “”You cannot freeze a company for any longer than that,”” McLennan says. “”The business changes too fast.””

Zrimsek thinks six months might be extreme. He warns, “”It is possible to try and do too much too soon, and wreck it completely.”” Clearly that did not happen in the case of All Weather.

Part of the reason is McLennan was adamant about checking “”scope gallop”” — where business users keep asking for new capabilities to be added, either requiring costly custom programming or constant tinkering with package configuration.

“”One thing I’m famous for here is saying, ‘No you can’t have it,'”” he says. Senior management backed him up every time he said it.

ERP is still tough to implement successfully, but it can be done. What’s needed is discipline and fierce determination. For Banman, it’s summed up best in the slogan embroidered on a ball cap somebody brought back from NASA in Houston during the ERP project: Failure is not an option!

“”That,”” says Banman, “”is the kind of attitude it takes.””

Share on LinkedIn Share with Google+