In this article I share how to use the True Calls per Hour calculation in the Contact Centre.
When you hear folks talk about Contact Centre productivity they’re usually talking about the Agents
Usually when we hear people talk about productivity they have their finger pointed firmly in the direction of their Agents.
“How can we get our Agents to be more productive?” they ask.
When we ask “What do you mean by productivity?” the most common answer is –
“How can we get the Agents to handle more calls or live chats per hour (day/week)?”
But Quantity Handled per Agent is and always has been a problematic measure
Productivity in a Contact Centre is not about how many calls or chats are handled.
This measure for Service Level based contacts has always been problematic.
There are very real mathematical realities at work that put the number of calls or chats handled outside the direct control of the Agent.
When you stop and look at it, the key factors that drive contact quantity either up (or down) per Agent include:
- The Service Level set (and its resulting Occupancy rate)
- The health of the Forecasting, Staffing, Scheduling & Real Time Management process at the interval level
- The size of the Queue at any given time (known as the Pooling Principle)
- The undeniable mathematics of random contact arrival (which is why we have Erlang C)
For Centres that have sorted this out and no longer target Agents on quantity handled – congratulations.
You’re well on the way to enhancing Agent and Customer Experience.
But let’s pause a moment.
Ok Dan (you might say). Got it. We don’t (or won’t) target Agents on Quantity Handled for Service level based contacts.
But for planning, comparative and high level management purposes, is there some way we can analyze the quantity handled across different shifts, cities and even countries?
Well I’m glad you asked. There is.
Let me show you how.
The True Pizzas per Hour calculation
When I teach this in workshops, I like to use the example of making pizzas in a Pizza Outlet.
See if you can answer the question posed in the picture below for our fictional Pizza Palace company.
What makes this difficult to answer is that our Delhi outlet is ‘busier’ than our Chennai outlet.
Perhaps our Delhi outlet is located on the ground level of a busy mall while our Chennai outlet is a bit off the beaten track in a low traffic area.
But we can’t possibly hold Prachi or Sangeetha accountable for how busy (or not) their outlets were – they’re not in the Sales & Marketing Team.
They were hired to make pizzas.
Got your answer?
Ok – let me show you how we ‘normalize’ the figures:
In order to correctly compare both Prachi and Sangeetha, you take what they actually ‘did’ (in this case how many pizzas they made) and divide that by the Occupancy rate they experienced at that time.
Once you normalize the data as you see above, we can calculate the ‘rate’ at which both of these people are working.
The use of the word ‘rate’ is important (think of a speedometer telling you your rate of speed).
Prachi is working at the rate/speed of 25.3 pizzas per hour. (In other words if her Occupancy rate had been 100% this is how many she would have made).
Sangeetha is working at the rate/speed of 28.3 pizzas per hour. (In other words if her Occupancy rate had been 100% this is how many she would have made).
So now we can compare both of our pizza makers on the same basis because we have factored out the impact of the different Occupancy rates.
But could we have a problem?
Typically at this point in the discussion the topic of ‘Quality’ comes into the picture – hurray Quality!
What we don’t know (or haven’t figured out) yet
What we don’t know in this exercise (at least so far) is the appropriate or best ‘rate’ at which we should be making pizzas.
What is the ‘right’ rate that yields a delicious pizza. Because we want Customers to come back again!
And because higher isn’t always better. (that had to be said)
Studies must be done
Fast food companies are well known for doing very scientific time & motion studies on how many can be ‘done’ and still deliver the required level of quality.
Contact Centres could learn from their example.
It is very likely that Pizza Palace has conducted in depth time and motion studies.
For purposes of this article let’s assume that they discovered that a pizza maker operating at the rate of 22 – 25 pizzas per hour during the lunch hour was ‘in the zone’.
By in the zone they mean that quality standards were achieved without any obvious loss or potential gain in productivity.
Now we can draw some conclusions about our pizza makers in the example.
Prachi is probably ‘doing fine’ – she’s operating at the upper end of the zone. But we should still taste her pizza now and then for quality assurance purposes (otherwise known as monitoring).
On the other hand, Sangeetha is operating outside of the zone – on the high side. We better go monitor her pizza to ensure quality hasn’t been compromised in some way.
Of course, if someone is ‘too fast’ it could be that a) they are in fact working too fast (and thus Quality falls – such as the taste of the pizza) or b) they could have made some kind of process or quality innovation that should be studied and replicated.
In the best Contact Centres…
In the best Contact Centres they don’t target Agents individually on the quantity of contacts handled (for Service Level based contacts – that caveat must always be there).
But when they want to do comparative quantity analyses they use the same normalization technique we used for pizzas.
Some of the conclusions I’ve seen Clients come up with using normalization include:
- We know for Ireland/Germany/Singapore (name your market or city) that on a Saturday afternoon shift the right ‘rate’ of call handling that delivers on quality is about 12 – 15 calls per hour (remember rates will vary across the course of a day)
- We know that our night shift Team calls per hour achievement will always be lower than our day shift Team calls per hour achievement
- We know that our calls per hour rate for Japan will always be lower than our calls per hour rate for India
- We know that if we see variations in the rate we need to explore the underlying reasons (root cause analysis) and not just blame Agents. .
Notice that none of these learnings had to do with targeting individual Agent calls handled.
I mean come on – if you want an Agent to handle a lot of calls all you have to do is understaff.
But that decision comes with a whole host of disasters and that’s why most Centres don’t intentionally understaff even when they know it would raise the actual call handling rate of each Agent.
These examples have everything to do with high level planning and analysis.
If you seek to compare the ‘rate’ of contact handling for different times of day, for different shifts, for different cities or countries – an educated implementation and use of ‘True Calls per Hour’ can help. This applies to Service Level based contacts only!
If in some way ‘quantity’ matters to you – for example the quantity of closures made per Agent or Team – then normalization works well because you’ve factored out the impact of Occupancy variations.
Simply targeting people to achieve a certain number of ‘closes’ is unfair – in the same way that targeting the number of calls or live chats per person is unfair.
Thanks for reading!