Over-Sparked?by Dave Michels in Telecom
Last week Cisco celebrated Spark’s birthday with a series of announcements. Cisco Spark was introduced just a year ago as Project Square. It was it was christened Cisco Spark at Enterprise Connect last March, and last week became the keystone to Cisco Collaboration – at least in terms of vision.
Amazon started as a bookstore, Google started with search – products often grow and expand, but Spark’s story is unusual. Spark was a platform for enterprise messaging, but its role is to be much broader. If Cisco is successful with the vision it outlined last week, Spark is will be its solution for UCaaS, Web Conferencing, Telepresence, and with its API emphasis a collaboration platform.
It’s a noteworthy vision, but is it possible? Product development, in general and especially for cloud-scale services, takes time. The bigger problem here is that a lot of what Cisco wants to develop has to first be invented.
Cisco intends for Spark to be a global UCaaS service – sort of. Spark is a Cisco hosted service – it cannot be purchased and private or dedicated implementations are not in the cards. Yet, Cisco intends all PSTN services to be provided by partners. To my knowledge no other UCaaS offer supports in-region carriers like this. For example, if a company has offices in three countries – they could conceivably have one virtual implementation of Spark (hosted by Cisco) with three in-region partners providing PSTN. Details on how the providers technically integrate and bill were a bit sketchy. The service is expected in Q1.
Another example pertains to security. The choice with cloud-delivered services has been that content is either secure or searchable. Client-side and network/transport encryption are easy – server-based encryption is the problem. I know of several companies that ‘tolerate’ cloud services, but disallow them for sensitive information. A few certs can satisfy some requirements, but if it’s super sensitive data than prem wins.
Cisco claims to have solved this, and later this year Spark will support search with customer controlled, encrypted data. Not even Cisco, court order or not, will be able to access the information stored on its servers.
Those are two big firsts. There’s lots of other firsts within the Spark vision. Enterprise messaging is still being defined, so lots o firsts there. Cloud-based analytics aren’t exactly new, but Cisco’s extensive use of them is potentially new. I haven’t seen provisioning via QR codes before. There’s also the automated testing processes and Cisco’s ability to roll-out new code anytime without interrupting call states.
Cisco’s vision for Spark is impressive, but then visions are supposed to be broader than current offerings. The tricky part will be turning the vision into a reliable, scalable, global service. Keep in mind, this isn’t a nimble startup, but a networking behemoth with a pedigree in hardware.
The good news is it won’t take long to find out. Cisco believes everything I’ve described above will be generally available in 2016.