Call Center Metrics & KPIs: 20 examples with formulas

Call center metrics are the measurements that show how well a phone team answers, resolves and converts calls. KPIs are the few metrics tied directly to a business goal. The most tracked include average speed of answer, abandonment rate, average handle time, first call resolution, CSAT and call conversion rate. Measure them on every call, not a sample.

Jump to the four families below: accessibility metrics (KPIs 1 to 6), efficiency metrics (7 to 11), quality metrics (12 to 17) and revenue metrics (18 to 20). Every KPI comes with a definition, a formula, a worked example and the levers that move it.

What Are Call Center Metrics and KPIs?

A call center metric is any number you can count about your calls: how long callers wait, how long conversations last, how many issues get solved, how many calls turn into bookings. A key performance indicator (KPI) is a metric you have chosen to manage the business by, because it connects directly to a goal someone owns. Every KPI is a metric. Most metrics should never become KPIs.

Here is the difference in practice. A team that books appointments by phone might watch dozens of metrics in its phone system, from hold time to transfers to calls per hour. It would manage by only a handful: whether callers get through, whether the call went well, and whether the call ended in a booking. Those few become KPIs, with a target, an owner and a regular review. The rest stay available for diagnosis when a KPI moves.

The same definitions apply whether you run an in-house phone team, a multi-location sales operation or an outsourced BPO program. In a BPO, the KPIs are often written into the client contract, which makes clear definitions and consistent formulas even more important.

What makes a good call center KPI

• Tied to a goal. It moves when revenue, cost or customer experience moves, not just when activity does.
• Measurable per call. You can calculate it from call data, not from guesses or occasional audits.
• Owned by someone. An agent, supervisor or location manager is accountable for it.
• Actionable. When it changes, the team knows what to do next: coach, staff, route or fix a process.

Types of call center KPIs

The 20 KPIs below fall into four families. Accessibility metrics tell you whether callers can reach a person. Efficiency metrics tell you how well agent time is used. Quality metrics tell you whether the conversation went well for the caller. Revenue metrics tell you whether the call created value for the business. Most dashboards cover the first three well. The fourth is where the money is decided.

Accessibility Metrics: Can Callers Reach You?

Every other metric depends on this family. A call that never reaches a person cannot be resolved, scored or converted, so start here.

1. Average Speed of Answer (ASA)

What it measures: The average time a caller waits in the queue before a person answers.

Formula: ASA = total wait time for answered calls ÷ number of answered calls

Example: For example, if a team answers 1,000 calls in a week and those callers waited a combined 30,000 seconds, ASA is 30 seconds.

What moves it: Staffing that matches hourly call volume, and routing that sends each call to the first qualified agent instead of a single overloaded line.

2. Service Level

What it measures: The share of calls answered within a time threshold your team sets.

Formula: Service level = calls answered within the threshold ÷ calls offered × 100

Example: For example, if 1,000 calls are offered and 750 are answered within 30 seconds, service level is 75% within 30 seconds.

Service level is often written as X/Y, meaning X% of calls answered within Y seconds, so the example above would be written 75/30. That is notation, not a recommended target. Set your threshold from your own baseline and what your callers expect.

What moves it: Forecasting and scheduling by hour of day, plus overflow routing when the queue backs up.

3. Call Abandonment Rate

What it measures: The share of callers who hang up before anyone answers.

Formula: Abandonment rate = calls abandoned before answer ÷ calls offered × 100

Example: For example, if 1,000 calls are offered and 50 callers hang up while waiting, the abandonment rate is 5%.

Decide up front whether very short abandons count. Many teams exclude calls that hang up within a few seconds, since those are often misdials. Whatever rule you choose, write it down and apply it every period so the trend stays honest.

What moves it: Shorter waits, accurate hold messaging and a callback option so callers do not have to stay on the line.

4. Blocked Call Rate

What it measures: The share of call attempts that hit a busy signal or could not enter the queue at all.

Formula: Blocked call rate = blocked call attempts ÷ total call attempts × 100

Example: For example, if callers make 2,000 attempts in a month and 40 get a busy signal, the blocked call rate is 2%.

What moves it: Enough lines and trunk capacity for peak hours, and queue settings that accept calls instead of rejecting them. Blocked calls never show up in abandonment numbers, which is why they are easy to miss.

5. Missed Call Rate

What it measures: The share of inbound calls that never reach a live person, whether they went unanswered, to voicemail, hung up or were blocked.

Formula: Missed call rate = inbound calls not answered by a person ÷ total inbound calls × 100

Example: For example, if a location receives 1,000 inbound calls in a month and 150 end in no answer, voicemail or a hangup, the missed call rate is 15%.

This is a bigger problem than most teams assume. Convirza aggregate data across 750+ customers shows that 20-30% of inbound calls never connect, lost to after-hours, overflow, voicemail and hangups. See more in our inbound call benchmarks.

What moves it: Overflow routing, callback workflows and coverage for the hours when nobody is at the desk.

6. After-Hours Call Coverage

What it measures: The share of calls that arrive outside business hours and still get handled by a live agent, an AI voice agent or a timely callback.

Formula: After-hours coverage = after-hours calls answered or handled ÷ after-hours calls received × 100

Example: For example, if 200 calls arrive after closing in a month and 150 are answered live, handled by a voice agent or called back the next morning, after-hours coverage is 75%.

This KPI matters because demand does not keep office hours. Convirza after-hours analysis found that 28% of bookable demand arrives after business hours. If that demand reaches voicemail, it rarely calls back. See how teams handle it in after-hours call handling and why calls get missed after hours.

What moves it: Extended or shifted staffing, AI voice agents for after-hours calls, and a callback process with a clear deadline.

Efficiency Metrics: How Well Is Time Used?

Efficiency metrics show how agent time is spent. They are useful for staffing and cost, and dangerous on their own: push them too hard and agents rush the calls that matter most. Read them next to the quality and revenue families, and see agent performance management for how to coach against them fairly.

7. Average Handle Time (AHT)

What it measures: The average total time an agent spends on one call, from answer through wrap-up.

Formula: AHT = (total talk time + total hold time + total after-call work) ÷ calls handled

Example: For example, if agents handle 500 calls with 2,000 minutes of talk time, 250 minutes of hold and 250 minutes of after-call work, AHT is 2,500 ÷ 500 = 5 minutes.

What moves it: Better knowledge access during the call, fewer transfers and simpler wrap-up steps. Lower is not automatically better: a short call that misses the sale costs more than it saves.

8. After-Call Work (ACW) Time

What it measures: The average time an agent spends on wrap-up after the caller hangs up, such as notes, dispositions and follow-up tasks.

Formula: ACW time = total after-call work time ÷ calls handled

Example: For example, if agents spend 500 minutes on wrap-up across 500 calls, ACW time is 1 minute per call.

What moves it: Automatic call summaries and dispositions, and CRM fields that fill from the call instead of from memory.

9. Average Hold Time

What it measures: The average time callers spend on hold after an agent has answered.

Formula: Average hold time = total hold time ÷ calls placed on hold

Example: For example, if 100 of 500 calls are placed on hold for a combined 150 minutes, average hold time is 1.5 minutes per held call. Divide by all 500 calls instead and the figure is 0.3 minutes. Both are valid; say which denominator you use and keep it consistent.

What moves it: Faster access to answers and approvals, and fewer handoffs that need a supervisor.

10. Agent Occupancy

What it measures: The share of an agent's available, logged-in time spent actually handling calls.

Formula: Occupancy = time handling calls ÷ total logged-in available time × 100

Example: For example, if an agent is logged in and available for 400 minutes and spends 320 of them on calls and wrap-up, occupancy is 80%.

What moves it: Matching schedules to demand. Very low occupancy means idle capacity; very high occupancy leaves no time to recover between calls and shows up later as rushed conversations and turnover.

11. Schedule Adherence

What it measures: How closely agents work the hours and activities they were scheduled for.

Formula: Schedule adherence = time working as scheduled ÷ total scheduled time × 100

Example: For example, if an agent is scheduled for 40 hours in a week and works 38 of them as planned, adherence is 95%.

What moves it: Realistic schedules, visible real-time status and a clear policy for breaks and coaching time.

Quality Metrics: Did the Call Go Well?

Quality metrics describe the caller's side of the conversation: was the problem solved, was it easy, and did the agent do what good looks like on your scorecard.

12. First Call Resolution (FCR)

What it measures: The share of issues resolved on the first contact, with no repeat call needed.

Formula: FCR = issues resolved on the first contact ÷ total issues × 100

Example: For example, if a team handles 1,000 issues in a month and 700 need no follow-up call, FCR is 70%.

You decide what counts as a repeat. Most teams pick a window, such as the same caller calling back about the same issue within a set number of days, and apply it every period. Convirza reports that teams applying AI call scoring to support quality see first-call resolution improve 20-35% within the first 90 days of deployment (company-reported). See how to improve first call resolution.

What moves it: Agent authority to solve problems on the call, accurate routing to the right skill, and coaching on the calls that ended in a repeat.

13. Transfer Rate

What it measures: The share of calls passed from the first agent to someone else.

Formula: Transfer rate = calls transferred ÷ calls handled × 100

Example: For example, if agents handle 1,000 calls and transfer 100 of them, the transfer rate is 10%.

What moves it: Routing that sends callers to the right person the first time, and training that widens what each agent can handle.

14. Customer Satisfaction Score (CSAT)

What it measures: The share of surveyed callers who say they were satisfied with the call.

Formula: CSAT = satisfied responses (for example, 4 or 5 on a 5-point scale) ÷ total responses × 100

Example: For example, if 400 callers answer the post-call survey and 340 choose 4 or 5, CSAT is 85%.

What moves it: Resolution on the first call, shorter waits and agents who confirm the next step before hanging up. Remember that only some callers answer surveys, so CSAT describes the people who replied.

15. Net Promoter Score (NPS)

What it measures: How likely callers are to recommend you, on a 0 to 10 scale.

Formula: NPS = % promoters (scores 9 to 10) minus % detractors (scores 0 to 6)

Example: For example, if 200 callers respond, 100 are promoters (50%) and 40 are detractors (20%), NPS is 50 minus 20 = 30.

What moves it: NPS reflects the whole relationship, not one call, so use it as a long-term signal and pair it with call-level metrics you can coach.

16. Customer Effort Score (CES)

What it measures: How easy callers say it was to get what they needed.

Formula: CES = sum of effort ratings ÷ number of responses, on a stated scale

Example: For example, on a 7-point scale where 7 means very easy, if 300 callers give ratings that add up to 1,500, CES is 5.0.

What moves it: Fewer transfers, fewer repeat explanations and a clear answer on the first call. Always publish the scale and its direction with the score.

17. Call Quality Score

What it measures: How well each call meets the standards on your call scorecard, such as greeting, discovery, handling objections and asking for the booking.

Formula: Call quality score = points earned ÷ points possible on the scorecard × 100, averaged across scored calls

Example: For example, if a scorecard has 50 possible points and a call earns 40, that call scores 80%. Average the scores of every call an agent handled to get the agent's quality score.

Start with a clear quality assurance scorecard for calls and a consistent way to measure call quality. Call scoring applies that scorecard automatically, and call center quality assurance software brings the results into one review workflow.

What moves it: Coaching tied to specific scorecard items, using real calls as examples.

Revenue Metrics: Did the Call Create Value?

These are the KPIs most contact center lists skip, and the ones that matter most when calls carry buying intent. They require knowing what happened inside the conversation, not just how long it lasted.

18. Call Conversion Rate

What it measures: The share of real sales-opportunity calls that ended in a booking or sale.

Formula: Call conversion rate = calls that ended in a booking or sale ÷ calls that were real sales opportunities × 100

Example: For example, if 400 calls in a month are real sales opportunities and 120 end in a booked appointment, the call conversion rate is 30%.

The denominator matters. Counting every call, including existing customers, vendors and wrong numbers, hides how well agents sell. Learn more about appointment booking conversion.

What moves it: Coaching on discovery and asking for the booking, and routing hot leads to your strongest agents.

19. Missed Opportunity Rate

What it measures: The share of sales-opportunity calls that did not convert.

Formula: Missed opportunity rate = sales-opportunity calls that did not convert ÷ sales-opportunity calls × 100

Example: For example, if 400 calls are sales opportunities and 280 end without a booking or sale, the missed opportunity rate is 70%.

It is the mirror of conversion rate, but it earns its own line because it points to a list of specific calls you can review: the caller asked for a price and never got an appointment offer, or the agent never asked for the booking.

What moves it: Follow-up on missed opportunities while the caller is still shopping, and coaching on the patterns those calls share.

20. Cost per Call

What it measures: What it costs to handle one call, across people, technology and overhead.

Formula: Cost per call = total phone operation cost for the period ÷ calls handled

Example: For example, if running the phone team costs $20,000 in a month and agents handle 4,000 calls, cost per call is $5.

Read it next to call conversion rate. A cheaper call that loses the booking is not a saving. See the full definition of cost per call.

What moves it: Better first call resolution, less after-call work and fewer transfers.

Put the four families side by side and a pattern appears. Accessibility, efficiency and quality metrics describe the queue. Revenue metrics describe the business. The operators who grow fastest measure both.

Why Most Call Center Metrics Miss What Happens on the Call

Speed and volume metrics come from the phone system, so they cover every call. Survey metrics cover only the callers who respond. And the conversation itself, where the booking is won or lost, is usually reviewed by hand on a small sample.

The average organization manually reviews less than 4% of inbound calls, so 96% of those conversations go unanalyzed (Convirza analysis of 72,000+ locations). That leaves call conversion rate, missed opportunity rate and call quality score resting on a sample too small to coach from.

The cost shows up as uneven results. Convirza's multi-location benchmark finds 3× performance variance between the best and worst location or agent in multi-location businesses. Without a view of every conversation, you cannot see which locations or agents sit at the bottom, or what the top performers do differently. Read more on why most sales calls go unreviewed.

How to Choose the Right KPIs for Your Call Center

Twenty metrics is a menu, not a dashboard. Pick the few that match your goals, then make them stick.

1. Start from the business goal. If the phone line exists to book appointments, call conversion rate belongs at the top. If it exists to support customers, FCR and CSAT lead.
2. Pick one metric per family. One accessibility, one efficiency, one quality and one revenue metric gives a balanced view without drowning the team in numbers.
3. Set targets from your own baseline. Measure where you are for a few weeks, then set targets against that baseline. Someone else's average reflects their callers, their hours and their definitions, not yours.
4. Match review cadence to the role. Review agent KPIs weekly in coaching sessions and location or team KPIs monthly with managers.
5. Pair every speed metric with a quality or revenue metric. AHT next to call conversion rate, ASA next to FCR. That way nobody can improve one number by hurting the one that pays the bills.
6. Write the definitions down. Record the formula, the denominator and any exclusions for each KPI so the number means the same thing every month.

For the operating habits that sit around these KPIs, see the call center best practices guide.

How to Track Call Center Metrics on Every Call

Your phone system already gives you the accessibility and efficiency numbers. The gap is the conversation. To measure quality and revenue on every call instead of a sample, you need a system that listens to each call and scores it against the criteria you set.

That is the job of Convirza Conversation Analytics™. It scores every conversation with call scoring built on your own criteria and attributes each call to the source that drove it, so call conversion rate, missed opportunity rate and call quality score come from all of your calls. Convirza is conversation intelligence infrastructure, measuring calls for 750+ customers across 72,000+ locations. For a wider look at analyzing call data, see call center analytics.

1. Choose your KPIs using the steps above.
2. Set your scoring criteria so every call is measured the same way.
3. Coach and route from the results, starting with the calls that were real opportunities and did not convert.

Want these numbers outside the dashboard? The Convirza API returns calls with their scores, and the Convirza MCP server lets Claude, ChatGPT or Cursor answer questions like "How are my agents performing this month?" and "Show me missed opportunities from this week." under your own Convirza permissions.

Stop guessing which calls made money. Maximize the value of every call. Start free and see your KPIs measured on every conversation.