March 13, 2005

What's Behind the Retek Battle?

By Evan Schuman, eWEEK

As Oracle and SAP engage in their half-billion battle for who will takeover retail software vendor Retek, the most obvious reasons for the fight are not the true ones.

The biggest news this month in retail has been the battle between $9 billion SAP and $3 billion Oracle to take over retail software vendor Retek, which had barely $174 million in revenue last year and less than $7.7 million of profit. But that's pretty good compared with the prior year, when it lost a little more than $20 million.

So how much are SAP and Oracle offering?

SAP started the bidding at $496 million and Oracle tapped it slightly with a $525 million offer. SAP has yet to announce if it will counter.

What do SAP and Oracle see in Retek? Quite a bit. It goes beyond that retail is a huge space with a very aggressive growth curve projected for the next decade.

Yes, it's true that Retek has an impressive list of marquee clients, including Abercrombie & Fitch Co., Nordstrom Inc., Hallmark, A&P, Best Buy, The Kroger Co., RadioShack Corp., Eckerd Corp., Tesco, Zale Corp., Sears Canada Inc., Gap Inc. and Kohl's Corp. But that's not all that there is to Retek's allure.

Any look at why Retek is attractive needs to start with why Retek is seen as likely to avoid the almost clichˇ-d downfall of most mergers, such as HP-Compaq, AT&T-NCR, AOL-Time-Warner and Netscape-AOL. has been reporting recently about some of the integration problems of the AT&T Wireless-Cingular merger as well as Sears-Kmart.

Mergers of so-called equals have a huge hurdle in front of them because someone needs to be seen throughout the company as the acquirer. When multi-billion players like Oracle and SAP are involved in a much smaller player such as Retek, that issue shouldn't be a problem.

Another key historic technology merger problem is the loss of momentum during a transition that can last a year or more, which is a golden window of opportunity for rivals to steal nervous customers. IBM's legendary FUD campaigns (fear, uncertainty and doubt) were tailor-made for that "right-after-acquisition" phase when corporate IT executives get the most nervous.

That loss of momentum happens when there is a big question about product direction and/or strategy and—as is the case today with AT&T Wireless and Cingular—difficulty integrating the technology to form the cohesive whole that was sold to shareholders.

That's the heart of the argument for the analysts who believe that an Oracle takeover of Retek would be far less disruptive than would an SAP takeover.

That addresses why a Retek acquisition might not be bad. But why do SAP and Oracle think it would be so good? Much of that answer comes from looking at Retek's revenue stream.

For a company that has held a strong favorable reputation within U.S. retail for years and has amassed such an impressive link of major retail clients, why is their revenue so relatively small?

From one perspective, their plight is not that different from the hypothetical car company that makes the ideal automobile, which lasts for 50 years and never needs maintenance nor fuel. The company makes a killing its first two years, but then revenues quickly plummet. After it has sold to most of the prospects who could afford such a vehicle, it's business model provides no recurring revenue from its customers and it goes bankrupt.

This is not to suggest that Retek's products approach perfection. Indeed, Retek received some unwelcome publicity just last month when Retek systems sold to Coles Myer and Britain's Sainsbury chain developed problems. Sainsbury even claimed that it needed to take a $700 million writedown from IT last October and placed some of the blame on Retek.

But the very nature of the large merchandise planning and merchandise operations management systems—coupled with the investment-averse nature of retailers in general—means that Retek's cashflow much more closely resembles that fictitious 50-year car than the once-or-twice-a-year upgrade licensing fee structure that Oracle, SAP and the likes of Microsoft have gotten used to.

Many retailers can go 10 or more years before replacing a Retek system, in the same way that the Point-of-Sale (POS) systems they are replacing today are often 10-15 years old.

The reason why they last so long is another cashflow yin-yang Retek fact: highly-customized. Retek's team crafts packages precisely to that specific retailer's situation. On the plus side, it's a hypnotizingly-persuasive lure when making that initial sales pitch and clients cough up impressively big bucks in the beginning.

"Retek brings to the table a lot of marketing pizzaz," said Greg Buzek, president of the IHL Consulting Group. "It demos very, very well."

But like a high-quality custom-made suit, it lasts a very long time and the alterations are often free. On a great suit, however, the hope is that the customer will buy many more like it. With retailers, it's not necessary. The best that Retek can hope for is word-of-mouth referrals, which it's been getting. But ultimately, there are only so many Tier One retailers. Retek's only hope is to find some way of getting those retailers to pay for additional services—or to be bought out and to let the new owners worry about it.

"We're not talking about the XP model, where every two years you get a new model. This stuff doesn't turn over," said Jeff Roster, a principal analyst for retail at research firm Gartner. "If you want to be doing a lot of updates, that may not fly (in retail). This has to last ten years because there is so much customization that goes into it."

And when it comes to milking customers for more money, few can hold a candle to Oracle's Larry Ellison. (Ok, Bill Gates can, but Microsoft has a very different view for retail: capture the low-end. But that's another column.)

With this level of customization, retailers don't replace it until they absolutely have to. What they will do is tweak the programming every so often to adjust for business changes, but that's handled using in-house coders.

Retail is different from many other verticals because of a notorious cheapness, but it's a stinginess out of necessity. IHL's Buzek estimates that the typical healthcare company runs at about an 8 percent profit margin and a traditional insurance company tries to bring in six percent margin. The average retailer can barely hope to bring in 1.5 percent, Buzek said.

Despite the industry's efforts to paint this Retek battle as an ego war, it's much more than that. Although I'll grant you that it's hard to envision an Oracle strategy that isn't somehow pegged to id-protection. In the new Webster's Dictionary, under "ego," it offers Ellison as a synonym.

What is really going on is the technology push for the next generation of retailers. With ERP, business intelligence and CRM initiatives being seriously debated in every major retail boardroom today, it's clear that a lot of high-end software dollars will be up for grabs very soon.

Oracle and SAP are hoping that if they have tight integration with merchandising systems from Retek, it will help deliver large amount of purchases: their kind of purchases, with lots of upgrades and new per-seat licenses and technical support and more upgrades and cross-platform support and more upgrades. And did I mention even more upgrades? Oracle is hoping to start naming upgrades Oracle Business Intelligence 10g.5PM and 10g.6PM.

One of the key reasons retail is so attractive to Oracle and SAP is the vast amount of data being collected and how little of that data retailers tend to truly understand.

Aberdeen Group retail analyst Paula Rosenblum phrases it well when she says that many large retailers "are almost savant-like. They know every detail of how they put their business together. But, in fact, some of these details matter and others don't and they can't tell the difference. Ask 99.99 percent of retailers what their core competency is and they can't tell you."

SAP appears to be bracing for a loss with Retek, with their CEO late last week pulling a page from Aesop's The Fox and the Grapes and saying that it's not that critical if they lose Retek.

"Retek was, is, a firm that meets our requirements, and we made an attractive offer. It's not the only possibility," SAP CEO Henning Kagermann told Reuters. "It's not that we have a hole in our company that we necessarily must fill."

Uh-huh. No, they just wanted to drop a half-billion dollars on a lark. SAP has been desperately trying to get into retail for years, with little success. If they officially lose Retek, look for them to quickly grab a Retek rival, very possibly Finland's Aldata Group. The U.S. stockmarket has been boosting the stocks of JDA Software, Catuity and Retailix on speculation that they could be a backup plan for SAP.

Let's not forget that one of the most real reasons so many are looking to grab Retek is that Retek has made itself such an enthusiastically willing acquiree. This may be one of those cases where the Retek board concluded that it would be best to letter larger and greedier heads prevail.