This article can also be found in the Premium Editorial Download "Storage magazine: What are the real benefits of data storage management software?."
Download it now to read this article plus other related content.
The company's Patrol Storage Manager (PSM) had reaped praise for its innovative application-centric approach. Even after a tough negotiation where the price dropped by nearly a third, BMC walked away with a $2 million deal for PSM and related products. But in February 2003, BMC pulled the plug on PSM, leaving PSM customers with no roadmap for distributed storage management.
|When the partner stops dancing|
A storage manager at an unnamed, large manufacturing company was stunned to learn when browsing technology business headlines on the Web that BMC had suddenly killed its Patrol Storage Manager product. The manufacturing company had only purchased the product some months before and had been diligently trying to implement it. In fact, the effort required considerable development work on their part.
"We had been investing a lot of time and effort into this," he recalls. Their efforts also put them in pretty frequent contact with BMC technical people. No one had said a word about this--not even a hint that BMC might kill the product. Even their account manager didn't know what was going on.
Implementation efforts immediately stopped. After spending months to select the product and negotiate the deal and then spending months more trying to implement the product, the manager, his staff and the executives above him were furious. "We are all too busy to waste our time spinning our wheels for nothing."
Although there won't be any negative career ramifications at the manufacturing company as a result of what is now seen as a fiasco, everyone agrees that they were used, abused and deceived by BMC.
"We had been giving them a lot of valuable testing and feedback," the storage manager says.
Now they have to restart the tool selection process from what amounts to square one. And more importantly, critical management capabilities they needed yesterday won't be available for months--at the earliest.
From a financial standpoint, the situation is even murkier. The diversified manufacturer does a lot of business with BMC. In the end, corporate purchasing folded the deal for Patrol Storage Manager into a much larger package that provided a huge aggregate discount while blurring the price distinctions among various products. Whatever they paid, the storage manager assumes, will be applied to some other products in the overall mix.
The end of this story is not yet in sight. "Now we're wondering about all our other BMC products," he says.
Gary Pilafas, senior storage and systems architect at UAL Loyalty Services, a subsidiary of UAL Corp., recently needed to automate the management of 21TB of storage. He shopped aggressively and bypassed most of the household names in management software before settling on CreekPath, the Longmont, CO-based startup, for a sale that cost approximately $300,000.
Unfortunately, many storage managers are facing a similar dilemma: either buy a large, expensive product from an established player or buy a more focused, less expensive product from a startup. In either case, you better hope the product works, the company stays in the market and that it lives up to its value proposition.
It's not that vendors are getting fat off of you--most complain they're not making any money in storage management, sales take too long and development costs are too high. So what gives with storage management software? Is it too expensive? Is it too hard to buy? Or is it too complicated?
BMC's problems seem to have been a mix of their own private woes and ones afflicting a number of other storage software vendors. BMC acquired Mountain View, CA-based Remedy, a help desk software company, for more than $350 million last September. A bad economy flattened growth and reduced revenues in areas the company had recently entered and marked for investment, such as distributed storage management. But some of the reasons behind BMC's lack of sales were common afflictions among software vendors:
- PSM was a complex product that had a lengthy evaluation process.
- High sticker prices led to prolonged negotiations and dramatically lowered selling prices, but at the end of the day, it was still a lot of cash.
- PSM had a lot going for it, but it took a long time to deploy--particularly with recession-restrained staffs--and still, it was only a partial solution that didn't always work.
To make the price of storage management software more palatable, vendors are throwing out their list prices and adopting a variety of pricing models (see "Storage management automation pricing models"), which means that prices are all over the map. If your vendor rep says "that depends" when you ask for a price, they're probably telling the truth.
For example, when BMC was still in the open systems storage management business, it listed the price of managing a terabyte of storage at $40,000, but even when that price was discounted, it didn't reflect what the actual cost would be. First, you needed BMC Patrol as the system management foundation. Then you needed specific management functions--such as storage resource management (SRM)--which were bundled together at an additional price. And then you'd have to buy various knowledge management modules to manage the equipment of different vendors. So, you could drive down the initial list price only to see the price bounce way up when you added in everything you actually needed to do the job.
Similarly, Computer Associates (CA) recently introduced BrightStor Portal, which provides a single point of management control for the enterprise storage environment. The cost of the portal is $35,000 with a 50-seat license, but that doesn't begin to include the cost of the functional modules that enable you to actually do anything.
Nigel Turner, senior vice president at CA's BrightStor storage solutions group asks: "What's the value in ensuring your storage environment is always available and under control?" Well, unlike the stuff you buy with MasterCard, it's not priceless. "A large enterprise wanting to deploy the whole breadth of functionality is looking at a cost of millions of dollars," Turner says.
Like most software, storage management automation prices include high margins--70% or higher--according to a former vice president of marketing at a storage management service provider who wanted to remain nameless. Storage management vendors, therefore, are more than willing to cut prices, he explains, although they'll try to make up for it through the sale of accompanying professional services.
Young companies will cut prices even more, he adds, because they don't have a huge legacy code base to maintain and are desperate to build a client roster. The typical deal price for large enterprises--when he left the storage management service provider in mid-2002--ran about $500,000, but a few deals ranged as high as $2 million. Again, the extra money buys the management of more vendor platforms and more devices--not more or different management functionality.
Beyond a willingness to deal, vendors also have been experimenting with various software licensing schemes. Traditionally, management software is licensed based on the size of the server on which it's installed. Some storage management software vendors now offer pricing based on the number of managed switch ports or the managed storage capacity.
InterSAN, which focuses on automating storage area network (SAN) storage provisioning, prices its licenses based on capacity. As the number of terabytes of managed storage increases, so does the cost. "In the past, companies typically spent 10% to 14% of the hardware cost on management. We try to keep in line with that number," says Karen Dutch, InterSAN's vice president of marketing. The payback comes from the automation of the provisioning task. Using InterSAN's automation, it takes less than an hour to provision one volume to an application running on one server, she claims. InterSAN sells its products directly and through Hitachi Data Systems and its integrators and VARs.
This was first published in April 2003