Service Level

What 80/20 means and other common targets

80/20 means 80% of calls answered within 20 seconds. It is the traditional voice standard and still widely used for customer service lines, but there is nothing magic about it. Common alternatives:

• 80/30 or 80/60 for lines where customers accept a short wait and cost pressure is high.
• 90/10 or 90/15 for emergency, premium or high-value lines.
• Chat: often 80% in 30 to 60 seconds.
• Email and messaging: usually measured as a response time, such as 90% replied within 4 or 24 hours, rather than seconds.

The right target is a business decision: what wait time customers will accept before satisfaction and abandonment suffer, against what the business is willing to pay in staff. Set it per queue, not one target for the whole center.

Formula variations, and why they matter

Centers handle abandoned calls in different ways, and each choice changes the number:

1. Answered within threshold ÷ calls offered. Counts every abandoned call against you. Strictest.
2. Answered within threshold ÷ (answered + abandoned after threshold). Excludes callers who hung up quickly. Widely used, and the version shown in the formula above.
3. Answered within threshold ÷ calls answered. Ignores abandons entirely. Flattering, and it hides real problems.
4. (Answered within threshold + abandoned within threshold) ÷ calls offered. Counts quick abandons as successes.

None of these is wrong as long as everyone knows which one is used. It matters most in outsourcing, where a service level agreement ties money to the number. A provider and a client using different formulas can disagree by several points on the same day's data.

How staffing drives service level

Service level is mostly a staffing outcome. Workforce management forecasts call volume and handle time for each 15 or 30 minute interval, then uses Erlang C or simulation to calculate how many agents need to be logged in and available to hit the target.

The relationship is not linear. Using Erlang C, a queue receiving 300 calls per half hour at 300 seconds AHT needs about 57 agents to reach 80/20. With 55 agents it falls to roughly 72%, and with 53 to about 53%. Each agent removed near the target hurts more than the last. The reverse is also true: pushing from 80/20 to 90/10 often needs more extra staff than the gain seems worth.

On top of the Erlang number, planners add shrinkage (breaks, training, coaching, meetings, absence), commonly 30% to 35%, to get the scheduled headcount. Undercount shrinkage and the plan is short before the day starts.

Service level vs average speed of answer and abandonment

Service level, average speed of answer and abandonment rate all measure accessibility, but they tell you different things.

• Service level shows the share of callers who got a fast answer. It says nothing about the 20% who did not.
• ASA is the average wait. A center can hit 80/20 while the remaining 20% wait 10 minutes, and ASA will reveal it.
• Abandonment rate shows how many callers gave up.

Report all three together, by interval. A daily 80% service level can hide a lunchtime half hour at 45%, and those customers do not care about the daily average.

What to do when service level drops

Find the cause before you react. The usual suspects, in roughly the order to check:

• Volume above forecast: a billing error, outage or marketing campaign. Talk to the business and update the forecast.
• Handle time above plan: new agents, a process change or a system issue. See average handle time.
• Fewer agents than scheduled: absence, poor adherence, unplanned meetings or training.
• Not enough agents hired: missed hiring targets or a class lost to training attrition shows up weeks later as a structural gap.

Intraday fixes include moving breaks and coaching sessions, offering voluntary overtime, pulling agents from offline work, moving cross-skilled agents between queues and using callback options so customers do not wait on hold. Structural gaps need a capacity plan and a hiring plan. Overtime can cover a week; it cannot cover a quarter.

Common service level mistakes

• Holding agents accountable for service level. It is a staffing and scheduling outcome. Agents control adherence and handle time, so measure those.
• Daily averages only. Plan and report by interval.
• Chasing the number late in the day. Overstaffing the evening to make up for a missed morning wastes money and does not help the morning callers.
• Undocumented formulas. Write down the formula, threshold and abandon rules for each queue.