Upon first hearing of a new, companywide computer system known as enterprise resource planning (ERP) going in at a promising-looking company, the portfolio managers at Disciplined Growth Investors of Minneapolis may decide to sit tight before buying stock. With an ERP project underway, they reason, the stock soon could be a lot cheaper.

“Of the 13 or 14 of these that we’ve experienced over the last couple of decades, I can think of maybe one that has gone as planned,” said Scott Link, lead portfolio manager. “I’m not exaggerating.”

Maybe an optimist would see an opportunity to buy shares last week after Patterson Cos. announced that an ERP problem — among other factors — led to a disappointing quarter. Its stock ended the day about 24 percent cheaper.

That an ERP implementation was underway at Mendota Heights-based Patterson was no secret. The dental and veterinary products company announced it several years ago. Its release talked about productivity gains and enabling the delivery of best-in-class customer service.

Last week in its quarterly conference call for investors, executives talked instead about “challenges,” “disruptions” and “internal headwinds.”

ERPs were created for manufacturers to use one big computer system to efficiently bring together all the parts and materials needed to build products. These systems became more sophisticated and took over more and more work across a company, linking into accounting and other departments to take care of things such as billing customers.

Buying an ERP system is by itself hardly news, as all companies use some technology system to run their business. But the very thing that gives ERP systems their power to enhance efficiency — tying together the whole business — creates a very big headache when something goes wrong.

Of all the technology used companywide, “ERP is the one that will really damage a company’s financials,” said Allen Debes, chief operating officer with Aeritae Consulting of St. Paul and a veteran of big-company technology transitions.

ERP is the system that automates the heart of a business, he said, containing information on everything a company builds or sends through the warehouse. And even with so much at stake, a new system can’t easily be tried out, like in a pilot program, or rolled out one piece at a time.

As for what can go wrong, an unwillingness to change some processes to better match the new software system is often one, Debes said.

That leads to another, related problem of having to knit the new system together with parts of older systems a company is unwilling to throw overboard. Not only is this costly and can take longer than anybody planned, that linkup is one place where errors can get introduced into the new system’s store of information.

There are so many other ways for things to go wrong, too.

Pacific Gas and Electric Co. made news when an outside technology analyst easily found a database with the details on more than 47,000 PG&E computers, servers and other machines — all suddenly available to anybody with an internet connection.

Target Corp. shows up in discussions of ERP disasters, too. Its short-lived expansion into Canada had so many wounds it’s difficult to say for sure what was fatal, but ERP might have been it.

Target didn’t have time to adapt its proven system for Canada, so when it expanded there five years ago it took one off the shelf from the big ERP vendor SAP. One result, according to a postmortem in the publication Canadian Business, was opening up more than 100 stores with inaccurate information in the system for most of its 75,000 products.

The Sleep Number Corp. didn’t have a great experience with SAP either, pulling the plug on a project in 2008 and writing off nearly $28 million. Then several years later a new ERP project didn’t seem to go much better. Executives on the company’s fourth-quarter 2015 investor conference call talked mostly about what they were doing trying to get their ERP system to work and catch up on deliveries.

This second, “successful” ERP implementation still exasperates Minneapolis investor Adam Wright, a onetime dissident Sleep Number shareholder. This project at Minneapolis-based Sleep Number Corp. might have cost more than $200 million including lost profits, Wright estimated this week, and he wishes he could easily identify the return on investment.

Patterson did not return calls to discuss its ERP problem, so it’s not easy to tell how much pain it may be in. Sales declined in the most recent quarter, and so did its operating profit margin.

The company told investors last week that its challenges are mostly inside of its four walls, not out in the market. It’s important to note, though, that Patterson has had to contend with heightened competition from competitors selling online in its traditional dental supplies business. Patterson had a good case for making a commitment to technology that could reduce cost as well as make it easier for customers to buy.

Patterson CEO Mark Walchirk offered investors last week the assurances one would expect, saying that “once we really stabilize the situation” the new system will allow Patterson to better manage its mix of products, make sales reps more effective and so on.

Wright made the same point about ERP, that companies can have a project that creates chaos and still quickly recover by working hard on cleaning up the messes. Wright said he still owns shares in Sleep Number, and not long ago the stock closed at its highest price in more than 10 years.

“It can be a blip,” Wright said. “It doesn’t need to kill you.”


lee.schafer@startribune.com 612-673-4302