"Mr. Savago is going to log in and monitor the queues but he isn't going to take calls... most of the time. In a pinch, he might take a few to help out though. Is there any way to set him up like that?"
I get this question during every installation of a T-Metrics TM-2000 ACD. It makes sense that a group of agents be managed or supervised by somebody (or a group of somebodies) who can help out when call volumes are high and so this question is often raised as an attempt to make that happen. There are many ways to achieve this sort of setup. You could, for instance, simply have Mr. Savago promise that he won't make himself available unless he wants to help out the agents in the queue. He could simply remain in an unavailable position
status until he is needed and then change it in order to help out. The problem is that end users forget things (don't we all...) and one day Mr. Savago will make himself available and start getting calls when he shouldn't be. Fortunately, the TM-2000 has three tools built-in by default that are specifically designed to help automate this process!
Let us consider a help desk for an example. Suppose this help desk meets the needs of an entire organization nationwide, and is centralized in Hermosa Beach, CA (if you like sushi, and you're ever in Hermosa Beach, go to Paradise Sushi - it will change your life). Suppose also that Mr. Savago, our mild-mannered help desk manager, organizes and supervises a team of 16 engineers who are the folks taking calls for this organization. He has a business procedure in place which places 8 of these engineers into a group he calls "Tier 1 Support". These 8 folks are the first line of defense. There is also a "Tier 2 Support" and a "Tier 3 Support", with four agents each, which exist for two reasons. First, when a Tier 1 technician cannot solve the problem for lack of knowledge or experience, he may pass the caller up to the next tier. Second, when call volumes soar, these other tiers are there to accept calls.
The first, and simplest, of these built-in TM-2000 tools is the skill level. This is the most commonly used method of mitigation when it comes to keeping supervisors behind the front lines. In short, agents who have higher skill levels will receive calls from the ACD before agents who have lower skill levels. Skill levels are assigned to an agent on a per skill basis, so an agent might have one skill level in one skill and another skill level in another skill. Every skill assigned to an agent must have a skill level associated with it. Skill levels range from zero (0) to nine (9). So for our help desk example: Mr. Savago might implement a system whereby his Tier 1 folks have all of the help desk skills at a skill level of 9, his Tier 2 folks a skill level of 8, and his Tier 3 folks a skill level of 7. This setup will cause the server to withhold calls from the Tier 2 folks until all of the Tier 1 folks are already on the phone and another call is delivered into the queue. At that time a Tier 2 technician (if such a technician is logged in and available) will be given the call. Likewise, Tier 3 technicians will not receive calls until Tier 2 technicians are all unavailable. Thus you can put certain agents in a position to receive calls for a certain skill before other agents. It should be noted that an agent with a skill level of 0 in a skill cannot take calls for that skill EVER - even if they are available and there are calls in queue.
The second tool used is the queue delay. This feature must be used in conjunction with the skill level feature that was discussed in the last passage. Queue delays allow you to specify a length of time for the server to hold calls in the queue before sending them to agents with lower skill levels. For our help desk, Mr. Savago might setup a queue delay between skill levels 9 and 8 for sixty seconds. This way, if all eight of the Tier 1 agents are busy with calls and a ninth call is delivered to the queue, the system will wait sixty seconds before delivering the ninth call to the Tier 2 agents. If, during the sixty seconds the server tarries, a Tier 1 person becomes available, the server will send the call immediately to the Tier 1 agent. This gives the Tier 1 folks a period of time to finish a call and take the newly queued call before sending it on to Tier 2.
The third tool is Skill Priorities. This feature was designed to give priorities to certain skills over other skills in queue for the same bank of agents. At first glance, this doesn't seem to useful when building tiers of agents. However, let's assume that for our help desk Mr. Savago has setup a super-secret-hidden option in his IVR menu that allows people who know about it to skip straight to a Tier 2 agent. Now let's suppose that we are in a heavy call volume day and all Tier 1 and Tier 2 agents are on calls. Let's also suppose there are three calls in the queue for the Tier 1 agents that have been waiting more than sixty seconds and one call in queue for Tier 2 that just arrived in queue. The Skill Priority tool allows us to specify that if one of the Tier 2 folks becomes available at this time, the Tier 2 call should be given precedence. This might make sense to Mr. Savago because Tier 2 calls can only go to Tier 2 agents, while the three waiting Tier 1 calls can go to any of the Tier 1 agents or the Tier 2 agents.
Using these tools in conjunction with each other, it is possible to come up with a fairly sophisticated system for routing calls to the appropriate agents within the appropriate scenarios. With certain settings in the state table (IVR) programming it is possible to perform even more checks within your call routing (changing skills, checking for agents logged into other skills, routing according to time of day, etc) but these are the basics. If you'd like to discuss more about these tools or other capabilities of the T-Metrics TM-2000 IP ACD, send me an e-mail at Support@tmetrics.com.
Subscribe