Solving the Hardest Job in Service Desk Management – Part II

In Part I we looked at how we go about working out the numbers of agents a service desk would need, based on the number of calls handled over a given period of time. Using a real-world example of around 450 calls a day, broken down in hourly intervals, we could work out the number of agents needed each hour between 8am and 11pm. This worked out at around 10 staff during peak times. Finally using overlapping shifts, we were able to work out a variable supply of agents during the day, for example fewer agents at the beginning of the day, more agents during the middle of the day, back to fewer agents toward the end of the day. This enabled us to avoid the over and under-staffing you get with a flat supply of 10 agents throughout the day.

Using Erlang C calculation or rather KoolToolz CC-Modeller which is much more user friendly, we can determine the staff supply needed each hour (or any other interval of your choosing) to deploy sufficient agents to handle the number of incoming calls. In this example, the overlapping shifts needed and the number of agents on each shift, each day, is as follows:

4 Agents working 8am-3pm (7 hour) shift
2 Agents working 9am-5pm (8 hour) shift
1 Agent working 10am-6pm (8 hour) shift
1 Agent working 11am-8pm (9 hour) shift
2 Agents working 1pm-9pm (8 hour) shift
3 Agents working 3pm-11pm (8 hour) shift
2 Agents working 5pm-11pm (6 hour) shift
1 Agent working 6pm-11pm (5 hour) shift

Sixteen Agents in total; and using the national mean for wages in the UK, this operation works out at £425k a year in staff costs.

Unfortunately life is not that simple. First we are not dealing with just one day but seven. That means we have to factor in days-off, we also do not expect Agents to work the same shift every day. So we need to rotate the Agents across all the shifts, to even out the drawbacks and the benefits of the shorter shifts, and also days off so they all get a fair share of weekends – or other days in the week. Do not be surprised when you find some Agents preferring week days off to week ends. There are other factors. The direction of rotation (DOR) will make a significant difference to Agents ‘recovery’ as their ‘body clock’ or circadian rhythms are impacted by shift time changes (this is particularly aggravated with ‘night shifts’, but in this example they are not worked). A ‘forward’ or ‘clockwise’ shift rotation is to be preferred. This means Agents shift start times get later as they progress from one shift to another in the shift pattern. The speed of rotation (SOR) also has a significant impact on Agent ‘sentiment’ or how they ‘feel’ about their working conditions. Generally speaking a faster rotation is to be preferred, but too fast and it can be counter productive.

Finally, an Agent’s schedule needs to be published at least 6 months in advance, and ideally a little over 12 months. This enables you to plan proactively and for Agents to plan their lives too. Don’t even think about anything less than 3 months unless you want shift bidding, overtime spiraling out of control, and reactive “agency” fallback hiking up costs as a matter of routine.

We will take a look how we create the Agents shift pattern, and then ultimately an Agent’s schedule for the year – and no, a shift pattern and a schedule are not the same thing. If you have the time you can check out the differences here otherwise read on.

Basically we need to turn this profile of inbound calls:

…into something like this: a structured schedule that will deliver a variable and cost efficient supply of Agents to effectively manage those 450 inbound calls a day.

We will discover we need at least 22 Agents not 16; and while this means staff costs of around £585k the schedule is both efficient and effective. Adopting a simplistic flat supply of 10 agents throughout the day, while remaining effective, efficiencies plummet as staff costs increase a further 36% to a total of £798k. A typical outcome caused by applying a literal Erlang C calculation across a variable time range – don’t do it.

Part III is the final part and will be a video taking you through the conceptual steps of creating the above shift pattern and generating the Agents schedule for a year in real time.

For details on Schedule24 Resource Manager, visit our website.

Clip to Evernote
Posted in Industry Spotlight Tagged with: , , ,
4 comments on “Solving the Hardest Job in Service Desk Management – Part II
  1. Aubert says:

    Is it possible to get a copy of the shift pattern used in this article? It is very similar to what I researched. for our organisation. It would be good if I could talk with someone.

    Many regards,
    Aubert Lanoie

    • Support Tec says:

      Hello Aubert,

      Please send me your contact details to support [@] and I will arrange for someone to contact you direct.

      Chris White

  2. Randy Moore says:

    Thanks for the tips. I’ve been wondering about the Erlang C calcs and how-to apply to shift patterns just did not take the time to search for it. You saved me a great deal of time.
    Randy Moore
    San Diego, CA

  3. Anonymous says:


Leave a Reply

Your email address will not be published. Required fields are marked *