In the definition of predictive dialing, we said that "the role of the smart software in the dialer is to achieve low wait times by overdialing, while keeping abandoned calls down to acceptable levels at the same time". The search process in the predictive dialing method is used to get this balance right. Once Oceanic® has completed the search process, it then runs the campaign, showing calculated average wait times and abandoned calls.
On quite a few campaigns you will find that in the first hour or two the achieved abandoned calls, measured as a percentage, may fluctuate a little around the target you set. As the campaign progresses you will then find that the difference between actual and target narrows, until in virtually all cases the difference (if there is one) is negligible. And so it should be in real life too, with a good dialer and a sorted calling list
This variance will be most marked when the throughput of calls is low, for example for small numbers of agents, e.g. five agents. Received wisdom is that predictive dialers cannot function effectively at such low agent levels, and that reductions in wait times between calls can only be achieved at the expense of wild fluctuations in the abandoned call rate. This is sometimes true, but try running a campaign with five agents, fairly short talk times and a low live call rate. Not only are achieved abandoned calls likely to hone in on the target quickly, but the reduction in wait times between calls, compared with auto preview, may be significant.