Schedule Adherence & Conformance, Part III: Target Setting & Implementation

In Part I: Introduction & Definitions, we established an understanding of metrics used in the contact center that assess actual fulfilment of the best-laid plans of forecasting and scheduling; in Part II: Impacts to the Contact Center, we discussed the importance of adherence and conformance in very basic terms--how much money do I stand to lose without measuring adherence and conformance?

This final instalment suggests ways to set targets appropriately, and provides some ideas as to how to implement new adherence and conformance accountabilities for your organization without alienating employees or operations management personnel.

Target-Setting: The Mathematical Approach

Setting a target for conformance is easy, because, as we discussed in Part I, conformance is quantitative in that it is only concerned with the total productive time delivered by an Agent versus the total amount of productive time scheduled.  For that reason, conformance expectations of your Agents can be the same across all lines of business or skill groups, if you choose to approach it that way.

Example: Assuming 7 hours of productive time in an 8-hour schedule (i.e. two 15m breaks and one 30m lunch), if you decide to allow your agents 15 minutes of unscheduled exception, or non-conformance, your conformance target would be around 96%.  It is advisable to include whatever percent of non-conformance you will tolerate to your budget factors, or 'shrinkage' calculations for hiring plans and capacity planning (I'll be starting a series in the near future on effective capacity planning, where we'll discuss this in deeper detail).

Adherence, on the other hand, cannot be set arbitrarily across your entire organization if different functions are performed by different skill groups, or teams--and it isn't as easy as setting targets for conformance.  The reason for this is because the average handling time (AHT) of contacts handled by different groups create varying impacts to achievable adherence.

Example: Agent 'A' works in a skill group that has an AHT of approximately 15 minutes, and Agent 'B' works in a skill group whose contact duration averages around 5 minutes.  They both have their first 15-minute break scheduled at 10:00, and just as they are about to log-off to take their breaks at 10:00, both are offered a call to handle.

Assuming Agent 'A' takes 15 minutes to finish their call, they are going to be out of adherence against schedule by a total of 30 minutes--15 minutes from 10:00 to 10:15 when they were supposed to be on break, and 15 minutes from 10:15 to 10:30 when they were finally able to take their break, but were scheduled to be 'open', or logged in.

Agent 'B', on the other hand, is out of adherence much less; they are out of adherence for a total of 10 minutes--5 minutes from 10:00 to 10:05 when they were supposed to be on their break, and 5 minutes from 10:15 to 10:20 while they were finishing up their break, but should have been logged in.  See the diagram below for a comparison between Agents:

The orange-shaded areas are where Agents were out of adherence.

But does this double-dip circumstance consistently present itself, and does it always happen to the severity used in the example?  The answer is 'no', of course, so what I suggest to my clients and companies I have worked for is to take the AHT for any given group, and multiply that by the scheduled exceptions that exist on a daily basis--typically two breaks and a lunch, or three exceptions.  This splits the difference between unadusted adherence thresholds, and ones adjusted for the absolute worst-case scenarios.  If we apply this to the Agent 'A' example above, that particular Agent would be allowed to be out of adherence a total of 45 minutes out of 480, resulting in an adherence score of approximately 91%.  If we applied this to Agent 'B', the Agent would be allowed 15 minutes out of adherence, resulting in an adherence score of approximately 97%.

Ultimately, this approach should be used as a guide, or a starting point, and ideal targets should be set based on a current-state assessment, and what senior management is willing to account for in capacity and hiring plan development.


More on Target Setting & The Importance of Numbers

Workforce management software allows the management team to manually adjust an Agent’s schedules, so the final schedule adherence report does not indicate they were out of adherence.  Some workforce management tools have the capacity to automate these adjustments once Agents submit them and gain approval for them to exist.  The degree to which you approve of schedule exceptions and/or adjust schedules to reflect them are up to your discretion, but I typically recommend that only exceptions that exist because an Agent was assisting a customer should be considered for adjustment in the schedule.  All other exceptions should be considered non-adherence to schedule.

How you decide to wrangle with updating exceptions will impact where you set your targets, as well as how your Agents perform.  Whether you are heavy-handed with your schedule adjustments (resulting in higher targets and higher actual results), or you allow for everything to show as out of adherence, the relationship between your target and actual performance will remain the same, so as far as your Agents and Supervisors are concerned, what the target actually is becomes inconsequential.  Adherence and conformance should be reduced to a simple "Yes/No" or "Pass/Fail" circumstance for your Agents and Supervisors.  All that matters is whether or not your Agents met the target.  For example, if your adherence target is 90%, and Agent 'A' has an adherence score of 85%, and Agent 'B' has a score of 80%, they both failed to satisfy their targets.  They are both considered equal in performance.  

The only people who should be concerned with the actual numerical results for adherence and conformance are your Workforce Management team and senior management, who will be using this data to assess interval shrinkage levels and budget factors for capacity and hiring plans.


Implementation & Avoiding Shell-Shock

If implemented too aggressively, adherence and conformance measurements have the potential to be huge game-changers in your contact center culture.  Move too fast, or too hard, and at the least, you'll be negatively influencing the level of service provided to customers by Agents' rushing through calls, and exercising poor diligence.  On a worst-case basis, you could alienate your Agent population so badly that you create an attrition problem for your contact center--a proposition which we know is high in cost, and arduous to repair.

What I suggest is to start measuring current-state adherence and conformance.   How are your Agents currently performing, and how far off of your desired or most-appropriate targets (established via the mathematical approach provided above) are they?  Depending on on the variance, you may want to consider establishing ramp plans with achievable targets that help your organization step up to your desired standards in, as an example, a 60-90 day period.

You should also consider the frequency in which performance is assessed.  I recommend measuring adherence and conformance on a daily basis, by Agent to establish trends and to provide indicators, but as far as accountability for Agents and Supervisors against a target, there should be a larger measuring period in order to establish a true representation of behavior.  I typically recommend Supervisors review adherence and conformance performance with Agents on a rolling two-week basis during coaching sessions, with formal scorecard or performance review data collected either on a monthly or quarterly basis.

This allows a center to avoid being paralyzed by daily results which can swing drastically in either direction depending on each individual Agent and circumstance, but also still provides a framework for improvement and quality fulfilment of schedules.

This concludes the three-part series on Adherence and Conformance.   I am hoping you would be kind enough to offer differing perspectives or more importantly, augment what I have provided via your comments for the benefits of the readers.  I would love to hear what you have to say!  Good luck in your endeavors, and stay tuned for more prophecies.