Dialing Under Extreme Conditions
The purpose of Oceanic® is to mimic campaign performances that will be both realistic, as well as acceptable in real life, and hence build understanding and confidence in those concerned with running outbound campaigns. Yet it's a lot harder, in some ways, to build a simulator than an outbound dialer, especially one with predictive dialing algorithms in it. Why? Because we know that some of you out there will put in some extreme values, which will very rarely arise in the real world, to see what happens. We don't want to discourage you, but if you do this, be prepared to wait a minute or so for your output.
Oceanic® is very conscious that you are probably sitting in front of it, tapping your fingers, waiting for it to deliver. So despite its occasional slowness under extreme conditions, it's still trying to keep down its computation time, to deliver you results as quickly as possible.
What are 'extreme conditions'?
Any campaign where the limits you have chosen for your inputs are close to or at the upper or lower limits set in Oceanic®. See Min/ Max Values A typical example is where combined percentages for both answering machines and no answers are 80% on a campaign with 250 agents.
How does the simulator cope?
In all dialing methods except predictive dialing, you will get instant feedback. If you have chosen predictive dialing, then depending on the mix of extreme values you have chosen, you may find that Oceanic® takes quite a few seconds in its search process to get the right result. This is a deliberate choice on Oceanic®'s part, in order to allow it to balance the pace of overdialing for the campaign, with the abandoned call target you set.
How do I relate this to real life?
In real life, if you are running your predictive dialer under extreme conditions, then you can probably afford the fastest Pentium processor. That will speed things up for a start.
Your dialer will have more time to fine-tune its the dialing rate.
You will find too that in dialing under extreme conditions, say forcing average wait times for agents up beyond the 30 second mark, the outputs you will get from doing multiple runs will show quite a lot of variation, with the overall average wait time recorded by Oceanic® moving within a band of up to five seconds, on campaigns with small calling lists. And this is what you can expect in real life.
A One Agent Predictive Dialer?
You'll notice that there is nothing to stop you running a campaign with a single agent. Here's what to expect:
Do a number of runs and take a note of the abandoned call rate you get each time. You will get a lot of variation in the abandoned call rate, but if you consider the average rate over all the runs you do, then you will find that the average rate for all runs will be close to the target you set, even in the first hour or two of the campaign. But it's all a bit academic, since the reduction in wait times you'll get, compared with auto preview, will be minimal. A quick way of verifying this is to check the trunk/ ports usage graph. But under some circumstances you can expect benefit with only five agents. See Achieving the Set Target for Abandoned Calls
Note
Don't be surprised if we tell you that we have already dialed well over one hundred billion of calls with Oceanic®, especially in predictive dialing mode, developing it, testing it, learning from it, and trying to confuse and break it. We have been literally probing everything that moves to be able to generate the very best predictive dialing performance possible, so that you can use Oceanic® as a standard against which to measure your own dialing performance. See also Benchmarking Against Oceanic®.