When it comes to telemarketing--or the teleservices or ARM (News
- Alert) industries--one thing is clear: the more actual humans contact center agents can actually make contact with, the better chance they have of making sales and similar connections that translate into cash flow, the lifeblood of every business. But what does a low AMD (News
- Alert)--Answering Machine Detection--error rate have to do with that? As it turns out, a lot. In fact, to merit every company that has a telemarketing department to ask itself: what are we doing to reduce our AMD error rate?
There are two key points at which an AMD error rate can affect the bottom line: one is in the form of employee compensation, and the other is in the form of standard costs that aren't recouped by sales activities. Without an AMD system in place, agents take their chances with the business' money, making calls to customers and, if they only get an answering machine, leave a message with same. While that does have some effect in keeping a business name front-of-mind for the contact target, the effect is limited, and certainly not as valuable as a direct contact. An AMD system reduces the number of answering machines an agent comes in contact with, allowing the bulk of contacts to be solid, live-person responses that have the best chance of resulting in sales.
Meanwhile, while agents are grappling with answering machines, they're also incurring other costs, the standard costs of doing business: software licenses, power consumption, managers, heating costs in the winter and a variety of other costs involved in just being able to stay in business. The less the agents sell, the less there is in resources available to cover these costs.
An AMD system, meanwhile, reduces the false-positive that is an answering machine contact, allowing agents to focus on making contact with actual people in a bid to sell, while the machines handle the machines, making contact, leaving messages where available and carrying on. But like most any system, an error rate exists that allows some answering machines to go through as though they were actual human contacts, and the business therefore needs to look to minimize that number. The lower the AMD error rate, the less time is spent on sub-optimal activities, and the more time is spent on activities that are more likely to recoup costs and drive profit.
The next question becomes one of how to accomplish such a task, and the answer seems to be in the algorithms. While most algorithms will cover things like customer abandonment and connect rates to determine the rate of calling, a standard that's much more universally used is that of number of agents signed in that the AMD should be dialing for. This in turn affects the pacing at which the AMD proceeds through a contact list, removing agents who are dealing with an answering machine system in error from the proceedings until the call is completed. This drop in pacing in turn provides a lower error rate by handling fewer calls while an agent is occupied.
Keeping the error rate in an AMD system low commonly results in improvements in the bottom line by virtue of keeping agents functioning in the field they're most likely to see profitable returns. It's not the only way to improve a bottom line, of course, but it's certainly part of the picture that needs to be addressed for the best chance at success overall.
Edited by Rachel Ramsey