Approach any Riverbed competitor and ask them why you should consider their product, and they'll likely give you the answer, "…because our product works just like Riverbed, but it's much cheaper…"
At first this claim might seem non-controversial. Certainly, the "it's cheaper" aspect may not be difficult to prove. But the first part of that claim–"it works just like Riverbed"–might take some digging to understand. How can you really tell if the competitor product is as effective as Riverbed? Here are five questions that I believe are key to answering that question:
1) Does the competitor offer layer-7 application-specific optimization capabilities for all of your important apps? To distract from weak or non-existent application-specific optimization capabilities, Riverbed competitors often draw attention to their compression and byte-level data deduplication features of their product. But the reality is that just about every WAN optimization product can perform data compression. What competitor products universally lack is the breadth and diversity of Riverbed's layer-7 optimization capabilities. The list is quite extensive–SMBv2 (important if you use Windows 7), encrypted MAPI (important for Exchange 2010), Citrix ICA and CGP, Oracle E-business suite (for both JRE and JInitiator), SSL (for encrypted web applications), SRDF (for EMC customers), NFS (for Unix environments), and a number of others. These capabilities are necessary not only to address latency and protocol chattiness issues, but also data encoding or encryption performed by the application that can hinder the effectiveness of data deduplication algorithms.
2) Does the product use a universal data store or a per-peer data store? I've commented in past blogs about the importance of the universal data store used in Riverbed Steelhead products. The reality is that a WAN optimization deployment will not scale with the per-peer data store used by most competitors, and this has been demonstrated in numerous failed WAN optimization deployment attempts, including by many current Riverbed customers who previously attempted to use a different vendor's product. If you have a large enterprise infrastructure with tens or hundreds of networked remote locations, my suggestion is to not be tempted to believe that you will be the exceptional customer who is skilled-enough to avoid the scaling issues that have afflicted so many other customers.
3) Is the vendor's products useful for all my requirements? One vendor may claim they excel in Disaster Recovery applications. Another might claim their product works better for ICA thin client traffic. Other vendors specialize in QoS. Still another might claim they work well in branch office consolidation environments, although they don't have a mobile software client. The reality is that purchasing from any of these vendors will lead to a fragmented multi-vendor strategy where each vendor product focuses on a narrow set of WAN optimization requirements. This leads to wasteful and inefficient allocation of management, support, and operational resources. Riverbed is the only WAN optimization vendor that can comprehensively meet all of your WAN optimization needs, including for QoS, cloud, mobile, data center, branch office, and virtual appliance requirements.
4) Are other customers happy with the vendor's products? Competitors are quite eager to make claims of parity or even superiority to Riverbed, but often they are less willing to back up their claims by letting you talk to their other customers. My suggestion is that before you spend your resources in evaluating and testing any vendor's products, that you first demand to talk to other customers who are happy with those products. While this very reasonable request might seem to be easy-to-fulfill, the reality is that for many vendors it often is not.
5) Why is the competitor product cheaper in the first place? Vendors such as Cisco are known for premium prices. Cisco's core switching and routing products are not necessarily superior to those from competitors, but customers are accustomed to paying higher prices for them merely because they are purchasing from the dominant equipment vendor. But in the WAN optimization market, things are different, where Cisco claims their products are the same as Riverbed, except cheaper. Somehow, there seems to be an inconsistency to this logic. I believe the key question is this: if Cisco WAAS products were truly equivalent to those from Riverbed, then would they be cheaper?