The following article is a re-post (with permission) written by Dan Rickwalder – you can find the original HERE Dan is a WFM consultant who works in customer service centres in many industries to solve their WFM issues, you can find out more about his consulting work HERE
- Forecasting-I love Minitab, and SPSS and Forecast Pro and all the other forecasting tools out there. And I often recommend EZForecaster (an excel add in) to my clients. WFM systems also have built-in forecasting tools. Using sound statistical models and processes in forecasting is critical. But because we go to such a deep level (half hour intervals anyone?) forecasting models can’t cope. These tools work great at the monthly or weekly levels but can not deal with intra-week and intra-day patterns unless we dissect them and hand feed them into the software. Add in the typical large and small scale events and drivers that occur and we must manage this information by hand. These tools are definitely not designed to evaluate the call impact of an email blast sent out at 3:30 on a Thursday!
- Cleaning data-The other side of the forecasting process is preparing and cleaning the data. This is another process which can be semi automated, but even most WFM systems require you to manually clean the data. Systems or tools that automatically scrub data can easily smooth out a legitimate spike, still need direction about special days and patterns and cannot cope with the undocumented events we always find out about too late.
- Long Term Planning-At last year’s SWPP conference I saw a company that had automated all the data inputs in their long term plan. Man do I want that!!! It was freakin’ awesome and if I had a lot of groups I would totally use it. But…and its a big one.. new long term planners need to hand enter data for a while. Typically, these plans combine data from many systems in ways that new planners have not considered- even if they were your best scheduler, they are looking at a much more complex picture now. Manually entering the data allows them to explore how the model works, and how the data connects. And this is not to mention the need to manually update future expected changes, forecasts attrition etc.
- Performance Management-By all means automate as much reporting as possible. And provide a scorecard to your supervisors so they don’t have to hunt for data. But automated data does not mean automated coaching. Supervisors need to take time and prepare for coaching sessions: review results, document issues and expectations, and develop a plan for improvement. Too often instead, they walk into a coaching session unprepared and pull up a scorecard to discover if there are any issues. This wastes time and is a disservice to the agent.
- Scheduling Team Meetings-OK for my last one, this is more of a “we shouldn’t have to but we do”. Our WFM systems have great tools to find the best time to have a team meeting, but then we have to go out and find a room and check the supervisor’s calendar in order to complete the meeting. And in many cases we send an automatic notification to an agent but then have to send a separate email for the room number. Come on man! If every two bit online meeting company can integrate into outlook so can you. And if nothing else, use the same tech you used for schedule preferences to allow us to at least highlight blocks in a day that are available and allow for more than just a description field so we can at least assign rooms and include an agenda.
So there you have ’em, five things we still do by hand. We have so many tools today but the fundamentals are the same and somethings just need to get done by a human.