Do Iron Mountain And Cirtas Exits Mean The Cloud Is Falling?

Last week, market research firm Gartner let it slip that Iron Mountain was shuttering its cloud storage service, and reliable sources are telling me that cloud storage gateway startup Cirtas has let its marketing and sales staffs go so the company can concentrate on improving the product. Is this a trend? Is the cloud falling? Will we be left choosing all our IT products from the five families ... oops ... remaining suppliers: IBM, HP, Dell, Oracle and Cisco?

Iron Mountain's decision to drop its low value-add, low markup cloud storage offering, while unexpected, made a lot of sense to me. Iron Mountain never seemed to have its heart in it and never made the offering attractive. While Amazon and Nirvanix offered options to store multiple copies in multiple data centers and advertised its prices, Iron Mountain apparently felt its trusted name would be enough and that corporate users would ignore the single data center architecture.  

Cloud storage is a game of scale. Large providers can use lower-cost but failure-prone components, and make up for it in networked fault tolerance, amortizing the development costs over many petabytes to offer attractive pricing. If you can't scale, or you can't sell enough storage to scale, you can't make money.  

My friend Greg Knieriemen was hard on Iron Mountain for leaving the cloud storage market, arguing that any exit was premature and would erode the trust customers must have to use cloud storage. While there's truth to that, I think Iron Mountain made as gracious an exit as anyone could. They stopped sales to new customers on April 1, and will discontinue the service by 2013. Compare that with Mozy's announcement that it was ending unlimited storage and boosting prices effective immediately. A user with a petabyte on Mozy could take a month to frantically copy it somewhere else, while an Iron Mountain user has a year or more to find a new solution.

Iron Mountain's management is also facing attacks from a hedge fund manager who wants the company to give up on the low-margin digital world entirely and reorganize the highly profitable business of managing boxes of paper records and backup tapes by the warehouse-full into a REIT (Real Estate Investment Trust), which would be more favorable tax-wise. Becoming a REIT would also require that Iron Mountain distribute 90 percent of its profits to its shareholders, which would be good for the hedge fund guys looking to generate as much income, and as many fees, as possible. However, this will leave Iron Mountain without enough cash to replace the warehouse business as enterprises shift away from archiving records in boxes to some sort of electronic vaulting. A week after it killed off the cloud storage business, Iron Mountain also sacked the CEO and brought back the last guy, whom we can assume knows more about managing warehouses full of boxes.

Note that Iron Mountain isn't getting out of the IT services business. Connected Backup, LiveVault, Mimosa and the value-add--and therefore more profitable--archiving solutions aren't being killed off, just the commodity cloud storage.

In a more common occurrence, cloud storage gateway startup Cirtas let a substantial portion of its staff go and suspended its sales and marketing efforts. Apparently, while Cirtas' Bluejet iSCSI-to-cloud gateways could manage a good demo, they weren't really ready for prime time. Management believes the market will be more receptive to a revamped product and wants to preserve what's left of the $22 million they raised earlier this year so they can revise the product and try again.

Iron Mountain is out of the basic cloud for business; Cirtas left because the product just needed work. That still leaves Amazon, Nirvanix and a raft of ISPs and hosting companies offering cloud storage, and Nasuni, TwinStrata and StorSimple with gateways. Looks like the cloud isn't falling, but these exits reinforce my opinion that IT folks need to plan their cloud adventures carefully. After all, sometimes clouds bring rain.