Research Publication Date: 25 August 2009 ID Number: G00170233 EMC Acquires Data Domain, Becomes Deduplication Leader, Signals Deduplication as a "Must Have" Capability Dave Russell EMC's acquisition of Data Domain provides EMC with expanded revenue opportunities, including industry-leading data deduplication technology and a likely alternative and/or replacement for EMC's current deduplication and virtual tape library (VTL) solutions. The purchase makes EMC the dominant deduplication vendor in a move that sends a message to the storage industry that deduplication is now a critical capability. Key Findings EMC significantly bolsters its deduplication capabilities with Data Domain's mature technology, and becomes the market share leader in deduplication solutions by owning the market's top-selling, client-based and target-side deduplication products. Leveraging its large global sales force and partner ecosystem, EMC can quickly sell the Data Domain portfolio, because no new development or integration is required. Expect the current Quantum-based Disk Library (DL) 1500 and DL 3000 deduplication appliances to be supported by EMC, but phased out in favor of Data Domain. Within a few years, expect the FalconStor-based DL 4000 VTL solution to be replaced with Data Domain technology as well. With acquisitions of Avamar for client-side deduplication and Data Domain for targetside deduplication, EMC has made this technology a must-have function, as the efficiencies and cost containment potential for this feature will be too compelling to ignore. Recommendations Data Domain customers should: Expect added security and investment protection for current deployments. Be proactive in managing possible unwanted changes in local support staff or procedures. Be quick to escalate support issues to reduce the time to reach product specialists. EMC customers should: Reproduction and distribution of this publication in any form without prior written permission is forbidden. The information contained herein has been obtained from sources believed to be reliable. Gartner disclaims all warranties as to the accuracy, completeness or adequacy of such information. Although Gartner's research may discuss legal issues related to the information technology business, Gartner does not provide legal advice or services and its research should not be construed or used as such. Gartner shall have no liability for errors, omissions or inadequacies in the information contained herein or for interpretations thereof. The opinions expressed herein are subject to change without notice.
Current DL 1500 and 3000 customers should regard these Quantum-based solutions as essentially stabilized offerings that Gartner believes will be phased out in the near future. Current DL 4000 customers of the FalconStor-based VTL should hold off on the current Quantum-based deduplication option for the DL 4000 series, as Gartner believes that EMC will replace the current deduplication option with a Data Domainbased option. Current DL users should review the planned service lives of their disk-based backup solutions in expectation of EMC functionally stabilizing these products. DL for Mainframe (DLm) customers should weigh future mainframe VTL investments against the cost and capabilities of a Data Domain-based solution combined with a gateway device, such as Data Domain's current partner, Luminex. EMC DL and Data Domain prospects should: Base purchasing decisions only on currently shipping products and capabilities, not on future possibilities. Regarding deduplication in general: In May 2007, Gartner said that deduplication is a transformational technology with the potential for significant cost savings and expanded quality of service capabilities (see "Data Deduplication Is Poised to Transform Backup and Recovery"). We reiterate this assessment, and advise clients to investigate deduplication technologies for use in addressing current and anticipated storage challenges. Although deduplication currently is used largely for backup activity through a file or VTL interface, this technology can be used for archive and primary data and should increasingly be considered for such applications. Publication Date: 25 August 2009/ID Number: G00170233 Page 2 of 10
TABLE OF CONTENTS Strategic Planning Assumption(s)... 4 Analysis... 4 1.0 What You Need to Know... 4 2.0 Event Facts... 4 3.0 Data Domain... 4 4.0 EMC's Previous Deduplication Technology... 5 5.0 EMC's Acquisition and Partnership History... 6 6.0 EMC and Data Domain... 7 7.0 Data Domain Implications From the EMC Acquisition... 7 8.0 Future EMC Implications... 8 Recommended Reading... 9 Publication Date: 25 August 2009/ID Number: G00170233 Page 3 of 10
STRATEGIC PLANNING ASSUMPTION(S) By 2014, deduplication will be a standard feature in all backup products and applied to 75% of backups. By 2014, some form of primary data reduction, such as compression and/or deduplication, will be used for at least 20% of all enterprise primary storage workloads, up from low, single digits in 2009. In 2014, deduplication will be widely available for blocks and files, and pervasive in application software, middleware, operating systems, appliances and storage arrays. ANALYSIS 1.0 What You Need to Know EMC's July 2009 acquisition of Data Domain, combined with the November 2006 purchase of Avamar, now makes EMC the market leader in backup deduplication technologies. Gartner reiterates the assessment that deduplication is a transformational technology, and we advise clients to investigate deduplication for use in addressing current and anticipated storage challenges ranging from cost containment to improved quality of service. 2.0 Event Facts On 23 July 2009, EMC announced that it had completed the acquisition of deduplication vendor Data Domain. In its second-quarter 10Q filed on August 6, EMC indicated that it acquired all the capital stock of Data Domain for approximately $1.9 billion in cash, which was net of Data Domain's cash and investment balances of $240.5 million. The completion of the acquisition brings to an end a bidding war between EMC and NetApp over Data Domain that started with a $1.5 billion offer from NetApp at the end of May. EMC stated that it intends to create a new product division within EMC's storage business that will focus on next-generation data protection, and that the plan is for all of Data Domain's employees to join EMC. 3.0 Data Domain Founded in 2001, Data Domain offers data deduplication solutions, shipping its first product in early 2004. Data deduplication is the activity of dividing a file into smaller "chunks," or segments, of data and comparing each segment to a catalog of chunks that have been previously processed by a backup application, archiving solution or a disk-based appliance. If a segment has been processed before, then a pointer is made to the previous copy of the data instead of storing it again. This process can result in significant disk savings, often in the range of 15 times or higher for backup workloads. Data Domain focused on the backup market, expounding a message of tape replacement, offering appliances that are placed after the backup server that perform "inline" deduplication to remove redundant data before the backup server writes data to the disk. Due to where deduplication occurs (after the backup server), Data Domain's approach is referred to as "target-side" deduplication. While Data Domain's devices could be used for archival data and for primary data through its file interface, Gartner believes that more than 90% of the appliances sold were used for backup data. As an early provider of deduplication, Data Domain has delivered six major software revisions and many processor upgrades, improving performance and capacity and extending its replication capabilities. At the time of the acquisition announcement, Data Domain is a mature provider of Publication Date: 25 August 2009/ID Number: G00170233 Page 4 of 10
deduplication technology, with approximately 825 employees, approximately 3,100 customers and over 8,000 systems deployed. For Data Domain, this acquisition represents the best possible exit strategy. Data Domain had grown very quickly and had enjoyed relatively little competition during most of its history, but that was beginning to change. Beginning in 2008, other major storage vendors, such as EMC and IBM, made acquisitions or OEM agreements in the deduplication market. Additionally, ExaGrid Systems, FalconStor, HP, NEC, Ocarina Networks, Permabit, Quantum, Sepaton and Sun Microsystems entered the market with enhanced disk solutions with deduplication capabilities. Gartner expects that by the end of 2009 every major storage vendor will have some form of competing solution, and that issues such as account control, routes to market, pricing, channel partner saturation and "good enough" solutions would have eventually limited Data Domain's growth. Gartner believes that many current Data Domain channel partners will continue selling Data Domain products. However, Gartner expects that some Data Domain partners may have concerns over EMC's ownership and the likely Dell reselling of the product, and that a subset of Data Domain partners will seek alternative solutions this may particularly be an issue for partners that sell Data Domain and NetApp offerings, and/or compete against EMC and Dell. 4.0 EMC's Previous Deduplication Technology EMC already appeared to be well-positioned for deduplication. It acquired Avamar for deduplication in 2006, completed an OEM agreement in 2008 with Quantum for deduplication appliances and, in 2009, began shipping data reduction technology that EMC called "file deduplication" at no cost in its network-attached storage (NAS) solutions to compete with NetApp. EMC acquired Avamar in November 2006 for client-side deduplication for $165 million in cash. Avamar's solution was an early-to-market deduplication product that was a software combination of deduplication and a backup application that deduplicates data before it is transmitted across the network to a backup server (hence the name "client-side deduplication," as the data is reduced at the source, or client). EMC quickly moved to package the Avamar software into an appliance that had the components preinstalled and integrated. The Avamar Data Store appliance has been a highly successful product, and EMC recently announced that the Avamar portfolio is experiencing a 44% growth rate, making it one of EMC's fastest-growing storage products. Gartner has seen EMC's direct sales force and partners push the Avamar solution in many bids, and customer reactions to the appliance packaging has been favorable. In 2007, EMC announced that it would offer target-side deduplication by 2008. Target-side deduplication performs the data reduction process after the backup server sends data to a disk target, thus any application data, operating system and backup solution usually are supported. In one of the most speculated activities of 2008, EMC announced in June 2008 that despite an OEM agreement and a successful VTL product line based on FalconStor technology, rather than adopting FalconStor's deduplication, EMC was entering in to an OEM agreement with Quantum to license deduplication and VTL software. Similar to the Avamar Data Store, EMC again offered an appliance. The new EMC DL 1500 and 3000 solutions were based on Quantum deduplication and VTL technology with no FalconStor code. However, EMC also delivered a deduplication option to its FalconStor-based DL 4000 VTL whereby a Quantum-based appliance would perform deduplication on data after it was written to the DL 4000 VTL. Why did EMC look for another deduplication partner when its current VTL provider already had the capability? At the time of the decision, the installed base and the number of production implementations with FalconStor's deduplication was much lower than Quantum's installed base. Also, Quantum's architecture completely bundled in deduplication and VTL capabilities, and had Publication Date: 25 August 2009/ID Number: G00170233 Page 5 of 10
a file interface (using Common Internet File System [CIFS] and Network File System [NFS] protocols), whereas FalconStor had not yet offered the same level of integration and capabilities (this has since changed in 2009). Additionally, EMC may have favored Quantum's ability to, at times, perform "in-line" deduplication, whereby data is deduplicated as it is being written to the device, whereas FalconStor provided "post processing" deduplication that writes all data to disk first and then performs deduplication as a background process. Why did EMC acquire Data Domain if EMC already had target-side deduplication solutions in the market? Gartner inquiry regarding the Quantum-based DL solutions indicated challenges with the Quantum approach. Concerns over the speed of the deduplication process when the device performed in-line deduplication, the level of robustness in the replication capability and, most especially, the need to have extra disk capacity available as the solutions would often have to resort to postprocessing deduplication led to some installations where customer expectations were not met. By contrast, Data Domain delivered its first deduplication appliance to market nearly three years ahead of Quantum. Gartner estimates that Data Domain had approximately three and a half times as many customers and just more than five times as many deduplication appliances in production, compared with the combined Quantum and EMC deployments at the time of the acquisition. Additionally, Data Domain offered faster in-line deduplication speeds, a broader range of appliances and more-robust replication, and possessed most of the deduplication "mind share" in the industry. EMC may have also seen the Data Domain buy as a defensive move whereby the technology was kept out of the hands of rival NetApp's direct sales force and partners' hands. 5.0 EMC's Acquisition and Partnership History EMC has relied heavily on acquisitions and OEM agreements to drive its growth in the storage market, particularly in the backup and recovery market. In the past six years, EMC has made approximately a dozen acquisitions and three OEM deals in the backup space. Some of EMC's storage acquisitions have been very large, such as Data General, Legato and now Data Domain (each at more than $1 billion), while other acquisitions have been comparatively smaller, such as Avamar, Berkeley Data Systems, Dantz, Kashya, Iomega and others. EMC's storage software strategy has been to strive to "check the boxes early" for emerging functions and capabilities, and to sell into its storage installed base, leveraging its large sales force and customer account relationships. When early solutions no longer seem appropriate, EMC is willing to change gears to provide more-robust solutions, as was the case for continuous data protection (CDP) in 2005, when EMC was early to offer a solution (RecoverPoint) that was based on an OEM relationship with the now defunct Mendocino Software. EMC replaced Mendocino seven months later with the acquisition of Kashya, which offered superior technology. Likewise, in September 2006, EMC released a data classification product called Infoscape that was based on a combination of existing EMC solutions, but, in April 2007, EMC acquired Tablus for content monitoring and later in 2007 Infoscape was discontinued. Similarly, in 2007, EMC investigated the target-side deduplication feature of its existing VTL supplier, FalconStor, then in 2008 chose to strike an OEM deal with Quantum for the capability, and now, in 2009, has invested in owning the intellectual property via the acquisition of Data Domain. In the last 10 years, EMC has made more than 30 storage technology and company acquisitions. While the merger and acquisition activity has substantially added to the assets that EMC possesses, issues of product overlap, positioning and customer, partner and sales force confusion can arise. Integrating a large number of solutions and/or swapping out technologies can also prove to be time-consuming and expensive. The net result can be the impression that the strategy is more opportunistic than methodical; however, Gartner believes that EMC often looks to be an early mover, or to test the market through partnerships and then to acquire when demand and/or the technology is proven. Publication Date: 25 August 2009/ID Number: G00170233 Page 6 of 10
6.0 EMC and Data Domain Initially, EMC plans no major changes for the Data Domain portfolio or its employees, all of whom are reported to have been offered a position within EMC. Importantly, EMC has also stated that R&D investment in the Data Domain product portfolio will be increased. EMC has further stated that Data Domain will eventually form the cornerstone of a new EMC division within the current storage business, which will focus on next-generation data protection. The exact title of the new division and the products it will comprise are still under evaluation; however, EMC CEO Joe Tucci has publicly said that offerings such as Data Domain, Avamar and the DL VTL portfolio are all strong candidates. According to Tucci, the intent of the new division is to reach $1 billion in revenue by the end of 2010 with a combination of product and related service revenue from Data Domain and EMC's Avamar products alone. Gartner believes that the new division would benefit from a broad portfolio of EMC's backup and recovery offerings, and, therefore, believes that Avamar, Data Domain, EMC's DL systems (FalconStor and Quantumbased), NetWorker and other EMC data protection products are also being considered as part of the portfolio for this new division. Former Data Domain CEO Frank Slootman has been tapped to head the new division and will report to EMC CEO Joe Tucci and to Frank Hauck, head of EMC's overall storage business. Data Domain reported on its last quarterly earnings call that its average sales transaction was approximately $113,000. With an acquisition price of $1.9 billion, excluding maintenance revenue potential and the modest installation charge, roughly 16,800 new transactions at the current average deal size would need to be completed to break even on the deal. Gartner believes that EMC has the potential to dramatically increase the average Data Domain deal size, sell more services and sell more of its storage portfolio. However, it's obvious that many thousands of transactions need to occur to recoup the acquisition cost of the deal. EMC has moved quickly to offer positioning on where its deduplication assets best fit. EMC will need to quickly and effectively communicate the positioning to its field force, partners and, of course, its customers and prospects. Gartner believes that EMC can double and perhaps triple the average Data Domain deal size primarily due to EMC's substantial installed base and its penetration of large enterprise accounts. EMC is a heritage storage company that has expanded into other areas. EMC competes against portfolio vendors, such as HP, IBM and Sun, all of which also happen to sell storage Hitachi Data Systems and NetApp are high-profile exceptions to this. With EMC's significant installed base, access to very large accounts, strong account control and combined 9,300 plus storage field force (EMC had more than 9,000 storage field resources, and Data Domain had over 400), EMC has the routes to market, relationships and brand image to push a great deal of Data Domain solutions. The EMC/Dell relationship only adds to the ability to secure new deals. By owning both Avamar and Data Domain, EMC has acquired two of the most mature deduplication solutions. Data Domain introduced its first deduplication product in late 2002, and Avamar introduced its first deduplication solution in early 2004. Avamar is the market-leading product in client-side deduplication, and Data Domain is the market-leading offering of target-side deduplication. Thus, EMC must now be considered the market leader in deduplication. 7.0 Data Domain Implications From the EMC Acquisition Data Domain's board of directors twice accepted offers from NetApp to acquire Data Domain. Data Domain accepted NetApp's first offer of $1.5 billion in cash and stock, then accepted NetApp's improved offer of $1.8 billion in cash and stock after EMC made on offer for the company. This sent the signal that Data Domain wanted to be acquired by NetApp and not by Publication Date: 25 August 2009/ID Number: G00170233 Page 7 of 10
EMC, which raises the question: "How well will key Data Domain management and employees integrate with EMC?" Data Domain's chief technology officer was a former NetApp employee, and other employees had a lineage that traces back to NetApp. Additionally, Data Domain's three main locations are all near NetApp facilities. Gartner believes that the Data Domain direct sales force is likely to view EMC ownership as preferable to NetApp, as the prospect of larger deal sizes will be enticing. Typically, acquiring companies offers attractive retention packages for key management and engineers, and EMC likely secured several years of Data Domain's top talent, thus reducing the risk of a "brain drain" due to a mass exodus of key individuals. Still, there may be cultural items to work through; however, the fact that Data Domain's CEO will lead the new division at EMC of which Data Domain will be a key piece may appease concerns by former Data Domain employees. EMC has publicly stated that it intends to increase the level of R&D investment in Data Domain solutions, and this should be well-received by the Data Domain team. 8.0 Future EMC Implications Gartner does not believe that Avamar investments or futures are in jeopardy as a result of the Data Domain acquisition. EMC's current sales focus on Avamar solutions is likely to continue, even in light of EMC's recent buy. Avamar is a fast-growing product within EMC, and Avamar possesses advantages over Data Domain for a number of use cases in remote-office, servervirtualized and NAS backups. There will be some overlap for data center backup workloads, and some future Avamar sales are likely to be supplanted by Data Domain products. In 2010, expect EMC to investigate manufacturing changes for Data Domain, including the transition of Data Domain products from current Xyratex disk shelves to EMC-based disk. Gartner believes that by 2010 the EMC-Quantum relationship for disk-based appliances will have terminated. Note that EMC is Quantum's largest tape reseller, and this is likely to continue. Gartner predicts that by 2011 the FalconStor-based VTLs will give way to a Data Domain-based solution, as in the next two years the need to emulate physical tape will be supplanted by the desire for well-integrated advanced disk solutions. EMC's current OEM relationship with Bus- Tech for its mainframe VTL solution, the DLm, could also dissolve, though this has not been decided, in favor of a Data Domain solution. Currently, Data Domain has approximately 25 mainframe installations that utilize a Luminex gateway device to convert mainframe data to an open-system format. Looking further into the future, the new EMC next-generation data protection division could extend not only existing products, but also could develop and bring to market entirely new backup solutions that would capitalize on EMC's many backup and recovery assets. However, rationalizing the various products and integrating several current offerings and/or designing brand new solutions are likely to take some time. Today, deduplication is most frequently deployed for backup activities. In large part, this is due to the highly repetitive nature of the backup process, which captures the same data again and again, leading to redundancies that deduplication is very good at factoring out. Data reduction can also be beneficial for archiving and for primary data (see "Emerging Technology Analysis: Primary Data Deduplication, Storage Software and Hardware Technologies"), and EMC is likely to explore how Data Domain's technology can fit into its future disk, NAS and archiving solutions. By moving beyond backup, EMC will have acquired Data Domain not only as a means of obtaining a fast-growing backup disk appliance, but will also have added hundreds of storage engineers and salespeople and added functionality to its storage portfolio to differentiate itself from competitors such as 3PAR, Compellent, Hitachi Data Systems, HP, IBM and NetApp. Publication Date: 25 August 2009/ID Number: G00170233 Page 8 of 10
RECOMMENDED READING "How EMC's Data Domain Buy Will Affect the Industry:" EMC's high-profile bidding war and ultimate buy of Data Domain has significantly raised the awareness level of deduplication in the industry; in the future, this deal could trigger additional partnerships and/or business development activity. "Data Deduplication Is Poised to Transform Backup and Recovery:" Data deduplication is an advanced form of data compression. The use of this technology is poised to transform the way data is protected, improving backup and recovery times, as well as the economics of disk-based approaches. "Use Data Deduplication to Improve Availability and Lower Cost:" Organizations continue to evaluate data deduplication as they seek new, cost-effective ways to store data. This research examines the benefits of deduplication, focusing on its effect on data availability, and addresses concerns about risk. "Emerging Technology Analysis: Primary Data Deduplication, Storage Software and Hardware Technologies:" Primary data reduction is getting more traction as IT departments look to reduce the ever-expanding footprint of data in their storage environments. "SWOT: NetApp, Worldwide, 2008:" This strength, weakness, opportunity and threat (SWOT) analysis examines the key factors Gartner expects will have the greatest impact on NetApp's strategy and market position in the global, external-controller-based disk storage market today, and during the next 12 months. "Vendor Rating Update: EMC's Information Infrastructure Business:" This research is a 2Q09 update to Gartner's evaluation of EMC's information infrastructure business. "Four Technologies That Midsize Businesses Should Consider in 2009:" This analysis identifies four technologies that have not been widely adopted in the midmarket, but have reached a level of maturity and affordability that makes them good investments for midsize businesses. "IBM Legitimizes Enterprise-Level Deduplication with Diligent Buy:" IBM's acquisition of Diligent Technologies gives significant visibility and legitimacy to data deduplication, which will speed the evolution and adoption of the technology in general. Publication Date: 25 August 2009/ID Number: G00170233 Page 9 of 10
REGIONAL HEADQUARTERS Corporate Headquarters 56 Top Gallant Road Stamford, CT 06902-7700 U.S.A. +1 203 964 0096 European Headquarters Tamesis The Glanty Egham Surrey, TW20 9AW UNITED KINGDOM +44 1784 431611 Asia/Pacific Headquarters Gartner Australasia Pty. Ltd. Level 9, 141 Walker Street North Sydney New South Wales 2060 AUSTRALIA +61 2 9459 4600 Japan Headquarters Gartner Japan Ltd. Aobadai Hills, 6F 7-7, Aobadai, 4-chome Meguro-ku, Tokyo 153-0042 JAPAN +81 3 3481 3670 Latin America Headquarters Gartner do Brazil Av. das Nações Unidas, 12551 9 andar World Trade Center 04578-903 São Paulo SP BRAZIL +55 11 3443 1509 Publication Date: 25 August 2009/ID Number: G00170233 Page 10 of 10