May 20, 2005

Wawa CIO: Upgrade Fear Dictated Multimillion-Dollar SAP Purchase

By Evan Schuman, Ziff Davis Internet

The CIO of the $3 billion Wawa convenience food chain was frustrated that various business units saw business data in different formats—and, consequently, were basing decisions on different numbers.

So, he sought bids to consolidate all retail software systems and present a more unified view for the 13,000-employee chain with stores in Pennsylvania, New Jersey, Delaware, Maryland and Virginia.

CIO Neil McCarthy eventually awarded the project to SAP, for somewhere between $5 million and $9 million. But along the more-than-year-long bidding path, he found himself caught in the middle of the firefight that Oracle and SAP launched during their struggle for control of Retek.

And, he said, he was ultimately quite happy that he ended up working with the loser of that bidding war.

When the bidding process began about 18 months ago, McCarthy's goal was to deliver consistent data across the company. It's not that any particular data snapshots were necessarily inaccurate, but because they were looking at different aspects of the business and using different criteria, the numbers different managers studied didn't always match up.

"We needed one version of the truth. Today, we have a bunch of different data repositories, seven or eight different price books and decision support systems. We want one common repository," he said. "Today the store managers see different [data] than marketing and finance. [With the new software,] everybody will be looking at the same data."

McCarthy's goals include more sophisticated and rapid-response pricing to secure universal access. "We want to be able to quickly answer what [products] we should be carrying and what we should be discontinuing and should we raise these prices and lower these other prices" he said.

"You'll be able to get to this information anyplace you can get to a Web browser," he said, adding that he wanted to avoid a VPN approach so that employees could truly access the data from any Web connection.

Instead, he would like to have beyond-password security such as a one-time-password keyfob, where the password would be displayed to the authenticated user and then quickly change. The security premise is that it would make a stolen password useless to an intruder as it would be voided in a few seconds.

McCarthy dubbed such an approach more convenient to employees than a VPN, which would permit access only from a machine that had the chain's VPN software installed.

As long a bidding process as this was, McCarthy estimated that it could take "four or five years" before installations and integrations were complete throughout all of the 550-store chain's operations, with financial units the first to be deployed and human resources, warehouse and facilities among the later ones.

With that long a rollout, the CIO was concerned about upgrade disruptions, especially given Wawa's history of needing major upgrades every 18 months to two years. "I've had 18 years of managing upgrades and best of breed solutions," McCarthy said. "It is a distraction."

They quickly focused on SAP and Retek as options, but SAP became the favored choice for, among other reasons, the belief that its upgrades would be less painful than Retek.

"As long as we stayed within the guidelines, the migrations would be made much more easily," he said, referring to SAP guidelines that dictated development and standards limitations—such as how information is passed along from system to system—that would allow less disruptive upgrades.

"Some of the companies we looked at, when you do an upgrade, it's like starting from scratch. It's like a custom installation with every upgrade," he said, referring specifically to Retek.

In many respects, Wawa's decision reflected a typical upgrade strategy. The more customized and fine-tuned to a user's site and business needs an application is, the more difficult it will be when it comes time to upgrade to a new version or migrate to a different package or platform.

McCarthy agreed with that IT balancing act, but said that his situation allowed for an easy decision. SAP's package out-of-the-box, he said, pretty much delivered all of Wawa's initial needs, so a lot of customization was not needed.

"Our needs are not as complicated as some other retailers'," he said. "We don't have 200,000 SKUs. We have 5,000. What [SAP] has is enough for us right now. We felt that we would be able to grow with them."

Had Wawa opted for Retek's more elaborate and customizable packages, he said, "it would have taken us years to really leverage that."

When it came time to lock down the contracts and finalize the deal, McCarthy said he retained a retail industry consulting firm—The Lakewest Group—to finalize figures and to prepare an ROI (return on investment) argument for presentations to other senior corporate executives. "They delivered a really good business case" for the SAP deal.

That's when Wawa found itself in the middle of the tug-of-war between SAP and Oracle for control of Retek. "The day we sat down for the lockdown in the negotiations" with SAP was the day before SAP announced its agreement to buy Retek.

The merger—had it ultimately happened—would have combined the two top contenders for the bid, in effect giving SAP a lock on the bid and theoretically weakening Wawa's negotiating position. "It was kind of odd because, suddenly, they were holding all the cards."

After some discussion, though, Wawa's IT team concluded that the merger didn't really matter and that it might even be favorable. "We thought it could have been a good thing because SAP was going to get some talent from Retek."

When things later flipped and Oracle ended up winning the bidding war and gaining control of Retek, McCarthy said he was relieved to have made the deal with SAP. "Thank God we didn't go with Retek," he said, because of all of the changes and uncertainty that will surround Retek "for the next two years as Oracle tries to integrate it all in."

Wawa is also an Oracle user, and McCarthy was asked whether that would have made a Retek-Oracle combination more easily integrated. He said it wouldn't have any impact.

"We're using Microsoft [operating system]. Does that mean that we should use SQL Server, too?"

Over at Retek, a senior executive said his company was sad to have lost the Wawa bid, but questioned whether Wawa dodged the merger bullet or is walking into an upcoming merger train.

The executive—Joe Polonski, Retek's vice president for enterprise strategy—was referring to widespread industry speculation that now that SAP lost its bid to purchase Retek, it is going to acquire a different retail company. "That news will come one day," Polonski said. "SAP didn't get what they wanted, so they are still shopping for second-best."

In November 2004, SAP bid between $6.75 and $7.25 per share for control of Retek and—when bidding against Oracle—ultimately offered $11 a share, only to be beaten by Oracle's $11.25-a-share winning bid.

Polonski made the argument that it's no sin to be rejected because your software is too fine-tuned to a customer's business objectives and that its feature set is more rich and robust. Asked if it was necessarily fact that Retek's package was stronger than SAP's, he said, "Obviously SAP thought Retek was more functionally rich."

Polonski said that during the bid for Wawa, the company saw data that Wawa's "user community had really latched onto Retek" and gave Retek a higher score.

Asked if he thought Wawa's decision made sense, given the concerns about business disruptions during upgrades, Polonski said it depends on the business' goals.

"SAP would be more comfortable if you don't change their software. You should change your businesses process to map to the software and all is good," he said, adding a parting shot that Wawa's decision made sense to him if the CIO "is willing to compromise on his business requirements."