Off Center
Conversations about contact center metrics often cause controversy – especially if you are having that conversation with me, and especially if I’ve been drinking at a conference cocktail reception. Drinking brings out the bold truth in me, and the bold truth (ok, my bold opinion) about metrics tends to stir things up amongst contact center folk. It’s one of the main reasons why I’m rarely invited back to industry events.

But I’m not looking to make friends – according to my Facebook page, I already have four. If what I write here results in me receiving hate mail or having things thrown at me by detractors, so be it – as long as the customer care industry as a whole improves.

In the interest of time and space, I’m not going to expound on what I feel are the most critical contact center metrics (Service Level, C-Sat, FCR, Contact Quality, Forecast Accuracy, Adherence to Schedule, and E-Sat). Instead, I’m going to discuss three popular metrics – those that have been coveted in our industry since its inception  – that I feel need to be clarified and brought back down to earth once and for all. These metrics certainly have their place – just not at the top of the heap.

1) Average Speed of Answer

No doubt that tracking how many seconds on average each caller is forced to wait before reaching a live agent is important. Then why isn’t ASA listed among my metrics A-list? Simple: It gets trumped by service level – a mightier metric that gives managers key information that already includes an ASA component.

ASA is like the unstable cousin or brother of service level – closely related to SL but much more likely to cause problems, mess things up, or get drunk and start a fight at a family wedding. What ASA lacks is the “X% in Y seconds” attribute of its blood relative, thus it does not always give the most accurate representation of what callers experience when trying to reach an agent in the contact center. ASA is a straight average, and many managers when seeing an ASA figure – say, 30 seconds – make the mistake of assuming that this figure is indicative of how long most customers waited before getting to an agent. However, the reality is that many callers get connected much more quickly than 30 seconds while many other customers may get connected after much more than 30 seconds.

So if you are looking to more accurately gauge what is happening to individual callers and not just hang your hat on a sometimes unreliable average – one that could conceal a host of irate callers who have access to social media and know how to use it – then it’s best to keep service level as your main accessibility KPI.

However, there is no need to disown ASA. It is still a useful metric in its own right – it can help in calculating trunk load and serves as a solid back-up measure to service level (since it comes from the same set of data as SL). Just remember that it can be dangerous to fully embrace this somewhat unstable metric or to invite it to important family functions.

2) Abandonment

When I say abandonment, I’m talking about the accessibility metric, not the feeling that contact center managers experience after asking senior management for a budget increase.

A contact center’s abandonment rate is the percentage of callers who hang up before their call is answered by a live agent. This data can be obtained directly from ACD reports. The most common formula used for calculating abandonment is:

Total # of calls abandoned ÷ (total # of calls abandoned + total # of calls answered)

While abandonment is an important metric and one that should be tracked on an interval basis by all contact centers, it is not an ideal measure of accessibility. True, high abandonment rates usually indicate a problem with the center’s staffing levels – but not always; sometimes conditions beyond the contact center’s control result in higher than usual abandonment rates, and sometimes a large group of callers are just cranky and impatient because their football team lost the day before.

Another problem with relying primarily on abandonment as an accessibility measure is that a low abandonment rate – while certainly the goal in any contact center – does not automatically signify smooth operating. Again, different conditions often affect how long a caller will wait in queue before hanging up. For example, on a typical day, a contact center may achieve a respectable abandonment rate of, say, 1.5% and everything is fine with regard to service level and customer satisfaction. However, the next day let’s say that the company runs a “limited time” online promotion (“Call now and receive a…”); the center will likely receive a flood of calls and few customers will hang up because they have an incentive to wait – for long periods of time. So abandonment may be less than even 1%, but does this mean that service level is good, costs are contained, and that customers aren’t getting perturbed? Nope. Nope. And nope. 

3) Average Handle Time

Back in the pre-historic era of call centers, when cost reduction/efficiency was the only real objective, AHT ruled the land of metrics. But customer contact has evolved rapidly over the past couple of decades; and in the best-run organizations, so have the performance metrics. Top contact centers today, while still keeping an eye on AHT, focus more on quality, call resolution and customer satisfaction. It’s the Age of the Customer Experience. And though it’s still important to control costs and be efficient – just as it is in any business or industry – winning organizations realize that they stand to make a lot more money if they focus on their customers first to create lasting loyalty and advocacy. 

Contact centers should certainly not do away with such traditional productivity metrics as AHT and number of calls handled per hour/shift. After all, the center needs to have an idea about how many calls a typical agent is handling and how long those calls are lasting to help pinpoint any scheduling adjustments and/or special training/coaching that might be required. But the best contact center managers don’t set strict “hit this target or else” AHT objectives. For one, they recognize that many factors influencing handle times are out agents’ direct control – e.g., call complexity, caller mood, electric shocks from cheap headsets, etc. And secondly, these managers know that if they do set and rate performance based on a strict AHT target, agents will do what it takes to hit that target, which may include rushing calls, sloppily inputting data, and yelling “fire” before hanging up on customers and running out the building.   

Instead, managers in leading contact centers set acceptable ranges for AHT and only really focus on the numbers when the center’s or an agent’s AHT falls far outside the normal range, which typically indicates a problem with staffing levels, agent adherence or agent skills/knowledge.

Progressive managers view AHT reduction ultimately as their responsibility, not that of their agents. They realize that it’s up to them to make sure that the center’s forecasting and scheduling processes are sound and that agents receive the coaching and training needed to handle calls efficiently and effectively. What smart managers do hold agents accountable for is adherence to schedule. Agents can, for the most part, control when they are in their seats ready to take calls. And when they are where they should be, things like AHT and other productivity metrics tend to take care of themselves. No whip-cracking nor threats of physical harm to agents or their family are typically necessary.

9/16/2010 10:58:43 pm

Enjoyed (and agree) with the metric assessment. Will be delivering a session for Benchmark Portal's Call Talk radio in the near future and have some ideas for the 30 minute talk, including:
Social media, is it really social?
Metric madness (a blatant plagiarism of your stuff)
Dealing with vendor schmooze
If you have some time over the next week, a conversation about the Call Talk session would be appreciated (but not compensated). Call 508-541-8889 if interested. Thanks.

Robert Last
9/16/2010 11:23:04 pm

9/16/2010 11:42:44 pm

Thanks for the comment, JR, and for your interest in involving me in the Call Talk Radio event. Sounds fun and intriguing.

I'll give you a call next week to discuss. And I won't even charge you!



9/17/2010 05:30:41 am

Great article, Greg...and once again we agree...particularly on the note about combining alcohol with metrics providing a volatile and passionate conversation!

9/17/2010 05:36:02 am

Metrics are only numbers. The only numbers that count are the ones important to the customer.

I don't use average handle time because it's more important to the customer to get the resolution on the first call regardless of how long it takes.

We don't put a lot of emphasis on abandon percent based on our largest customer segement - who doesn't care about this metric.

It all goes back to what's important to our customer.

9/17/2010 06:55:12 am

Thanks Michelle! Just remember -- friends don't let friends drink and forecast/schedule.

9/17/2010 06:59:15 am

Excellent input, Catherine. Sounds like you are running a truly customer-centric contact center. I love it!

I wish you and your team much success; judging by what you've written, you're already there.



9/21/2010 01:45:44 am

Greg, good article. As you allude to, the issue happens when these metrics are measured in a silo. This does not get you the 'big picture' on what is happening to the customer experience. Higher AHT could in fact correlate to a better customer experience as an example. Rather than knocking these metrics off their pedestals, they should be used to augment other metrics and paint a more complete picture of the customer experience.

9/21/2010 02:02:00 am

You are absolutely right, Omar. It's great to see that you -- and several others who have commented -- fully understand the shortcomings and pitfalls associated with the popular metrics I discussed in this post.

Thanks for your input!



7/24/2017 03:19:36 am

Nice post. Thank you.


Leave a Reply.