Lessons Learned From 3Par & Data Domain

In the past two years, we have been privy to two exciting bidding wars that ended in two-billion dollar payouts for both 3Par and Data Domain. Despite the fact that they were in two different ends of the storage market, there is a lot to be learned from both of these companies. The key factor is that both companies focused on the task at hand. They essentially did one thing really well. In the case of 3Par, it was simplifying block storage in the enterprise, and in the case of Data Domain, it was simplifying backup storage. Each company's focus made users' jobs easier, and users voted with their dollars.

Both companies focused on infrastructures that were unconventional. 3Par sold with a Fibre Channel block-storage system when IP and NAS are all the rage in primary storage. Data Domain entered the market with a primarily IP-based disk backup system when Fibre Channel VTLs were all the conventional offering. In other words, you don't have to do what the market is telling you to do, you have to do what makes the most sense for your product's capabilities.

Both companies became identified with a single feature, for the most part, that they rode to the promised land. Neither was the first out with that feature nor were they the first in the overall category. In the case of 3Par, the feature was thin provisioning and in the case of Data Domain it was deduplication. Clearly, they do more than this, but the companies, either by luck or design, realized that becoming the poster child for a particular capability would allow them to focus their limited resources and capture the attention of busy users. It is also interesting to note that those two features had something specific in common: they allowed users to do more with less in a tangible way. We always talk about how customers are being forced to do more with less, but thin provisioning and deduplication actually make that happen. Performance is nice, but saving money and keeping things simple are critical.

When both companies started promoting their technology, the incumbents in the space called those features too risky and dangerous, until of course they brought out or OEM'd the feature themselves. Most importantly, when thin provisioning and deduplication started to become commonplace, 3Par and Data Domain didn't abandon the technologies. They kept promoting how their implementation was better and more mature, keeping their focus instead of deciding to come out with a newer better feature. Both companies had a track record of ignoring the twinkling lights of opportunity and kept focused on the big payoff at the end of the road. 3PAR did enterprise, tier one storage. Data Domain did disk-to-disk backup systems. They were not dabblers in other areas of the market. I see companies all the time that try to be primary storage vendors but will also try to sell their solutions for backup, for archive and anything else that might capture a customer or two.

Another key similarity is that in both cases these were complete systems. While the systems were largely driven by software, that software was delivered with hard drives and controllers. While I like software-only solutions and virtual appliances may change this, clearly customers and companies tend to buy complete systems instead of kits.  While both 3Par and Data Domain are or have been clients of Storage Switzerland, this is really not so much a tribute to them as to their methodology. If you want to be successful in the storage market, you need to provide complete systems based on a single software stack, that are simple to use, even in the enterprise, and you need to stay focused on your area of the market even when times are tough.