Product management was awful leading to a lost opportunity because there was already a major customer but management had unclear vision and could not deliver a product in an agile fashion to the customer for rapid iteration. As someone mention, Riverbed does not have time to wait 5 years for a product to mature internally and then release. They should have released a minimum viable product sooner by being more focused.
The end-of-year 2017 SKO was a fiasco, rather than showing that SDP could bootstrap in minutes plain x86 platforms located at various remote branch offices into an environment where steel connect, steel head and other Riverbed products can be orchestrated. They showed a face recognition demo that left the sales force confused on what is the value add they will sell. The main value was supposed to be: "Riverbed will take care of orchestrating any set of Riverbed products and third parties products at SCALE for you and you just need a x86 server". This is very attractive to service providers with large installation base and big enterprise too. Riverbed would ship software rather than hardware and could sell more products as customers could license the products on live demand.
The engineering team was left to make a lot of decisions which worked well technically but might not be what the main customer wanted right now. It was not the fault of the engineers who gave more than their dued efforts to try to make the product a success.
A lot of the senior management jumped ship to other companies as they realize that the product was not gaining traction at consumer events and the lower engineers were left on a sinking ship. It was a disgraceful behaviour similar to captains leaving passengers on a sinking ship. I sat at one of the demo in MWC and it just did not make sense the way the demo was run.
All in all, bad product management lead to wasted resources for a struggling company and putting the engineers in a difficult financial position after the project was sc-apped.