Benchmarking Against Oceanic

Here's what you do for dialers using just an overdial algorithm:

1. Pick a campaign being run by a dialer and replicate its data using the Campaign Wizard. Make sure that the dialing practices being observed (see Oceanic® Performance v. Dialer Performance) are the same, and there are no 'unaccounted for' abandoned calls.
2. Run the same campaign in Oceanic®, and compare the average agent talk time per hour, with what you are getting in real life.
3. Keep adjusting the overdial control downwards from 100%, until the two figures for average talk time per hour are the same.

You will then have a percentage measure of efficiency for the dialer, under those campaign conditions.

If your wait time definition is the same as that used in Oceanic®, then you could use this instead of average talk time per agent hour, to equilibrate the two campaigns.

Will dialers be equally efficient under all dialing conditions? It all depends on the strength and flexibility of the dialing algorithm(s) at work. Under most dialing conditions the efficiency of a dialer in terms of the measure we have suggested should be reasonably constant. However, high no answer/ answering machine levels and/ or low call abandonment targets can be tough for a dialer to handle, and efficiency ratings may drop. And don't forget that any dialing inefficiencies forced on the dialer, because of distortions in the calling list, will be included within this assessment. So ensure that the calling list is sorted first, to eliminate any obvious sources of distortion.

If you don't want to go through this benchmarking exercise, but are convinced that you have a great dialer, then you may want a rule of thumb on how to adjust Oceanic®'s performance to mirror what your dialer would do. If you regularly sort your calling list data, but still want to allow for the impact of possible calling list distortions, then as a guideline only we would suggest that you set the overdial control at 90%.

What about if my dialer is making use of both predial and overdial algorithms?

Follow the same procedure as above, but be sure to enter the right predial interval value on the Extended Properties page. This value(s) will have been specified/ set somewhere in your host application. You may have a series of different values for different agents, some applying to wrap, and possibly some to talk only, all on the same campaign. Simply take the average of these values, across all agents.

Your dialing vendor may have a different name to the one that we use; no matter, the principle should be the same.

Then as before, keep adjusting the overdial control, until you have equilibrated the two campaigns, and you will have an efficiency measure for your dialer, on a scale of 1 to 100, again under the ruling campaign conditions.

"You must be joking. We can and do achieve much better performance than Oceanic® shows."

Perhaps you have taken our advice on slugging performance too much to heart. If that's not the case, and you are dialing with no reduction in any either of the predictive dialing controls, then check the dialing practices set out in Oceanic® Performance v. Dialer Performance, especially the abandoned call delay, when there is no agent available. For example, we have used an abandoned call delay of 1 second only in the examples and the demonstration. If you run them with a figure of 5 seconds, then you'll get better agent talk times per hour than those we have documented in the topics for both these subjects. So in assessing Oceanic® performance make sure that you are comparing like with like.

And if you still think that you are getting, or can get, much better performance than Oceanic® shows, unslugged, and with the dialing rules the same, then you're almost certainly not comparing like with like.