Suppose we have a call centre with two separate departments:  Sales and Service.

Sales take 5,000 calls per week with an AHT of 405 seconds, while Service take 20,000 calls per week with an AHT of 271 seconds.

Sales have been performing merrily at a conversion level of 25%, but are under pressure to do better.  Marketing decide to change the proposition so that more customers will buy; nothing will be done to drive more calls into the sales centre so there should be no adverse reaction either to Sales or Service in call answering terms.

The campaign is a success, with the Sales agents converting at 35% – a full 10% higher than the previous week.

Unfortunately, however, the service levels in both centres has taken an unexpected turn for the worse and an investigation reveals the following:

• Sales AHT has leapt from 405 to 447, even though the agents haven’t changed what they do!
• Service call volume has increased from 20,000 to 21,000 along with an AHT increase of 5 seconds on every call!

Had the individual call type propensities and AHT structures been understood, the impact of higher conversion would have been managed and the call centres prepared.

Let’s take a closer look.

Sales calls fall into two broad types:  Successes (conversions) and failures (non-converted calls).  Successful calls are the result of careful needs analysis, targeted sales pitch, finding the right product, closing the deal, mandatory verbal contractual phrasing, and the agent placing the order and taking a payment.  Failed calls take less time, unless you’re asking your agents to chase every call for maximum conversion.

In this call centre, an average converted call has an AHT of 720 seconds while fails generally last just 300 seconds.

At a conversion rate of 25%, AHT = 25% x 720 + 75% x 300 = 405 seconds.

When conversion increases to 35%, AHT = 35% x 720 + 65% x 300 = 447 seconds.  Had the breakdown of Sales AHT been understood better, the AHT increase could have been planned for, even if only with overtime.

It also happens that post-sales support is required on 25% of all sales, and that the AHT of these calls is 360 seconds rather than 270 seconds for all other calls.  So,

calls x conversion x post-sales support = 5,000 x 25% x 25% = 313 calls.

The increase in Sales unfortunately led to an increase in post-sales support, rocketing the requirement to 75% of all sales requiring assistance.  This leads to

5,000 x 35% x 75% = 1,333 calls.

This increase of 1,000 explains the increase from 20,000 to 21,000 calls.

The AHT, which was

((313 x 360) + (19687 x 270)) / (313 + 19687) =271 seconds

Now becomes

((1313 x 360) + (19687 x 270)) / (1313 + 19687) =276 seconds.

Again, knowledge of the breakdown of AHT by call types and propensity-based call forecasting enable a full understanding of what happened; smart planners would have seen it coming.