What Is Your Call Center’s Conversion Rate?7 May, 2010 By: Greg Sarnow Response
Is your call center’s conversion rate a moving target? Have you ever drilled down to see how your call center defines conversion? Surprisingly, there are many factors that impact conversion that you probably don’t know about.
Call Scrubbing and Conversion
Experienced DR marketers have learned over the years that call centers “scrub” calls to make their numbers look better. Test calls, wrong numbers, short calls, abandoned calls — the list goes on and on.
Call centers deduct these calls because, from their perspective, they are not sales opportunities. But because the agents make the decision, marketers need to be aware that other factors may play a part in deductions.
You, the marketer, want to see the bottom line. In this scenario, all you can really be sure of is that so many calls hit the switch and so many calls became orders.
Do call center agents select the call’s disposition accurately? Every call center has a different set of call descriptors on which the agent bases the decision. Even more interesting, an agent’s disposition of a call will positively or negatively affect their bonus, while giving you an incomplete picture of the call center’s performance.
Does that really happen? Yes, every day!
The flip side of the coin is a slippery slope as well. Quite a few consumers dial the call center for non-sales reasons, such as to reach customer service or get product information. A newer phenomenon is people who dial the call center on their cell phones just to have the phone number recorded in their call history.
Call centers always contend that measuring conversion should be based on qualified leads, not raw call volume. They believe conversion should be measured beyond just the switch/order relationship.
In the past few years, some marketers have changed to a “unique ANI” (automatic number identification) order conversion metric. Say 100 calls hit the switch and 95 were unique ANIs (meaning calls came from 95 unique phone numbers, and five were from people who called, hung up, and called back again). If 55 people ordered, measuring by unique ANI, the conversion rate would be 57.9 percent.
Call centers listen to calls for quality control, legal compliance and to improve results. They will eliminate calls from the results that could never be sales calls. But out of these 95 unique ANIs, if only 85 of those calls were truly selling opportunities, then the 55 orders would result in a 64.7 percent conversion.
What’s a marketer to do? You can’t listen to every call, but you can look at the percentage of calls that are considered non-sales calls. Define these carefully and don’t forget to add “doesn’t speak English or Spanish” to the list (depending on customers’ preferred language).
If the non-sales call percentage is more than 20 percent, a red flag should go up. The higher this number, the more likely that your call center has poor quality assurance (QA), and agent fraud is occurring.
At the end of the day, marketers evaluate call center conversion by how it impacts media cost-per-order (CPO) and the media efficiency ratio (MER). But when call centers have strong QA programs in place, there are fewer cancelled orders and fewer continuity disputes.
Apples to Apples?
When comparing call centers, compare the cost-per-call first. If one call center has a lower cost-per-call (with sufficient media spend to get accurate data), it means they are scrubbing fewer calls out of the results. Their conversion rate may be lower, but they might have better media results. After cost-per-call, then look at unique ANIs and conversion rates from agencies being tested.
After conversion, look at the non-sales call percentage. If higher than 20 percent, start listening in on non-sales calls to see if they are dispositioned correctly.
Beyond conversion, average order value and abandon rate, upsell percentages and telemarketing cost-per-order are all important metrics to consider when comparing call centers. With greater regulatory issues, compliance is fast becoming as important as other metrics.