This is a worthy conversation to have with a Cisco TME from CCBU. Here are a couple of issues I can think off of the top of my head with CCX:
- Every call in IVR treatement or on hold waiting for an agent must be pulled across the WAN to the CCX server. With UCCE/CVP the call can be processed on the local branch router using VoiceXML and then transferred directly to the agent. This saves a great deal of WAN bandwidth in a "spread across the country" model unless you have big centralized SIP trunks for all calls.
- If the customer is starting at 300 agents you're already at the far upper limits of CCX. It may be able to support 300-400 agents but only with a long string of caveats. The two most prominent:
- Either number assumes that the callers will be on hold for three (3) minutes on average. Averages lower than that generate more CTI activity than the MIVR engine can handle.
- That scale is only reachable for inbound voice. Every other feature on the platform (e.g. outbound voice, email, chat, supervisor login, etc) top out WAY before 300 agents.
- CCX runs in an active/standby model with no way to explicitly say that "node one always is the preferred node". They can failover and all of your calls are routing to a different CCX node, even after the outage is over.
- The queues are not multi-channel. In otherwords if you want to do email and voice calls, they are in different, uncoordinated queues.
There are dozens of differences. UCCE always gets new feature years before it is ported to CCX. In fact, the feature backlog for CCX is years long. My generic advice for a customer 300+ agents would be Packaged UCCE; however, there are enough differences to warrant a well-considered qualification discussion with Cisco and the customer.
Please remember to rate helpful responses and identify helpful or correct answers.
