Deduplicating Customer Choice

On May 20th, NetApp announced a friendly agreement to acquire Data Domain.  NetApp would gain a very strong product line with incremental revenue and growth.  In addition, NetApp, with its established distribution presence in EMEA, APAC and many top enterprise accounts, could open new doors around the world for Data Domain, accelerating its growth.  Put simply, this deal is a win-win for all parties involved.

 

On June 3, EMC weighed in, panzer style, (loved the YouTube video) with a hostile tender offer for Data Domain at $30/share, upping our offer of $25 a share in cash and stock.  Clearly we touched a nerve because, in about a week’s time, Data Domain became something so precious to EMC that they were willing to spend $1.8B on it.  NetApp revised our offer to match the EMC price and the Data Domain board accepted our revised offer and maintained our agreement.  

 

So what is this all about?  Why does NetApp want Data Domain?  Why does EMC so badly want us not to have it?

 

First, a little background.  Data Domain builds backup storage appliances that provide a disc-based alternative to tape for backup.  They have pioneered the use of data deduplication for backup data, allowing them to store 20x the amount of data as non-deduplicated storage.  This makes their appliance more cost effective than tape for a wide range of backup applications.  Mostly, Data Domain competes for the budget that would have been spent on tape infrastructure, but they also compete with Virtual Tape Libraries or a limited set of other disk-based backup appliances.

 

So why is this interesting to NetApp and our current and future customers?  We have built data protection into the very soul of Data ONTAP, so customers using NetApp for primary storage can easily add a second controller set to keep a secondary copy of the data either through mirroring or vaulting of snapshot copies.  This is a great solution for data protection, but only for NetApp primary storage.  In cases where NetApp is not the primary storage, we offer a strong VTL product that does an excellent job of augmenting tape-based backup solutions – but we do not have a product that truly replaces or minimizes tape when backing up data from EMC, HP or other storage systems.  Data Domain has a great solution for this specific problem – and we liked the idea of being able to offer that solution to a wider set of non-NetApp customers.

 

Clearly, EMC did not like this idea at all.  This is understandable.  EMC and NetApp are fierce competitors and it is natural for EMC to take action to try to slow us down.  But this move did, frankly, seem to make less sense.  After all, EMC already has so many other competing products in the backup appliance market.  They have an Open System Virtual Tape Library product called EDL based on technology from Falconstor.  They have another EDL product based on deduplication technology from Quantum (in which they have recently invested $100m).  And they purchased Avamar, which builds a deduplicating backup product that is now both standalone and integrated with Legato.  This is “Big Love” for backup products with deduplication. 

 

When you own or distribute so many of the products in the market, you tend to acquire high market share.  According to IDC, the worldwide market for open systems VTL was $630.6M in 2008.  EMC has about 42% of this market, based on NetApp share calculations and in the deduplicating backup appliance (non-VTL) market, their Avamar solutions has about 19% share.  By themselves, these share numbers are not a great concern.  But Data Domain has 17% share in the deduplicating VTL market and about 52% share in the deduplicating backup appliance market (based on NetApp share estimates and Taneja Group market size of $270M).¹ So the combination of EMC and Data Domain would have 59% share in deduplicating VTLs and 71% share in deduplicating backup appliances.  

 

Beyond the overwhelming market share that EMC would gain with Data Domain, EMC also would gain control over more than half of the patents that are vital to these fast growing markets.

 

Deduplicating data is good for the customer.  Deduplicating customer's choice of vendors is not.

 

Joe Tucci wrote a very sincere and heartfelt letter to the employees of Data Domain pleading with them to believe they have a future with EMC.  I wonder what letter he wrote to the employees of Avamar.  Or those who work at Quantum and Falconstor?  EMC already has made financial commitments to three other deduplication-enabled backup systems – but if EMC plans to keep Data Domain employees and products, what does that say about EMC’s existing commitments?  My guess is that the letters to the employees of the existing businesses would start with “Dear John.”

 

I would love to hear from customers who purchased Data Domain, EMC Disk Library or EMC Avamar solutions.  What products did you consider?  Who was in the final bake-off?  Did having two alternatives give you some pricing leverage?  How would you feel about EMC owning nearly all of the products and technology in this emerging market? 

 

I bet the government would like to hear from you as well.

 

 

¹ "Next Generation Data Protection Market Forecast 2007 - 2011", Sept 2008

Comments

[Bob Sinclair]

We bought Data Domain and it's a brilliant product that has saved us 44% on storage already, saving us around $430k per year - we'll be sticking with them and look forward to seeing what the new companies come up with next!

[Mark Kulacz]

I think the EMC acquisition of DataDomain was brilliant. Its easy to pass it off as a move done to attack NetApp, but once you really dive into the products the DataDomain solution is nice middle ground between EDL and Avamar, with many of the benefits of the Avamar dedup solution yet with the minimal disruption required to install an EDL.