# How contact center platforms define service level and abandon and handle time differently

Canonical: https://warmtransfer.net/knowledge/contact-center-metric-definition-differences

Last verified: 2026-10-14

Amazon Connect computes Abandonment rate against contacts queued, while Genesys Cloud computes Abandon % against offered count[^2][^21]. Service level also treats abandons differently: Genesys Cloud, with abandons included, counts every abandon as missing the target, while the Webex Contact Center Average Service Level KPI counts contacts terminated in queue within the SLA time as within service level[^29][^43].

## Service level

### Genesys Cloud

Genesys Cloud defines service level as conversations answered within the target time divided by the total conversations included in the calculation, and configurable options control which conversations are included[^30]. Contact center analytics settings expose 3 organization-level service level options: Include flowouts, Include short abandons and Include abandons in calculation[^33]. Editing these settings requires the Analytics > Insight Configurations > Edit and Analytics > reportingSettings > Edit permissions[^26].

The formula changes with each option:[^29][^32][^31]

- With abandons included, service level is (tAnswered − nOverSla) / (tAnswered + tAbandon), so every abandon counts as missing the target[^29].
- With short abandons excluded, service level becomes (tAnswered − nOverSla) / (tAnswered + tAbandon − tShortAbandon)[^32].
- With flowouts included, conversations that a flow action removes from the queue before answer count as missing service level, using (tAnswered − nOverSla) / (tAnswered + tFlowOut)[^31].

The Include Callback Virtual Wait in SLA option is off by default[^23]. When it is on, callbacks are judged on callback outcomes (reached, not reached, abandoned), using virtual wait from callback offer to outcome[^23].

### Amazon Connect

The Amazon Connect data lake reference query for service level counts contacts handled with a queue answer time at or under X, then divides by all queued contacts[^11]. Abandoned contacts therefore sit in the denominator and never in the numerator[^11].

The GetMetricData API accepts a SERVICE_LEVEL threshold of any whole number of seconds from 1 to 604800 and requires the comparison LT (less than)[^10]. A single GetMetricData request can include up to 20 SERVICE_LEVEL metrics, so you can pull several thresholds at once[^14].

WarmTransfer's reading of the sources is that a contact answered exactly at the threshold counts as within service level in the data lake sample query (<=) but not under the GetMetricData LT comparison, so the two results can differ slightly at the boundary[^13]. The GetMetricData reference links the SERVICE_LEVEL metric's UI name to Average queue abandon time, and we infer this is a documentation labelling error rather than the metric's meaning[^12].

### Webex Contact Center

In Webex Contact Center, the service level threshold is a per-queue advanced setting[^47]. It sets how long a request can wait in queue before it is flagged as outside service level, and it cannot be set for social channels[^47]. The threshold is capped at 86400 seconds for telephony and chat, 604800 for digital, and 1209600 for email[^46].

The Average Service Level KPI counts a contact as within service level if, within the SLA time, it was connected, opted out of the queue, was dequeued, or was terminated in queue[^43]. We infer that early abandons therefore appear to raise this KPI, which is the opposite of how Genesys Cloud and the Amazon Connect data lake query treat them[^44].

### Across platforms

WarmTransfer's reading of the sources is that the same call traffic can produce different service level percentages on Amazon Connect, Genesys Cloud and Webex Contact Center, because their documented treatment of abandons in the formula differs[^50].

## Abandonment

| Platform | Metric | Calculation |
|---|---|---|
| Amazon Connect | Abandonment rate | (Contacts abandoned / Contacts queued) * 100[^2] |
| Genesys Cloud | Abandon % | (Abandoned count / Offered count) * 100[^21] |

### Amazon Connect

Amazon Connect treats a contact as abandoned when the customer disconnected while in queue, before connecting to an agent[^2]. Contacts queued for callback are not counted as abandoned in the Abandonment rate metric[^5].

Abandonment rate shows 0 when results are grouped by routing profile[^9]. This happens because an abandoned contact never reaches an agent and so has no routing profile[^9].

Abandonment rate has no short-abandon exclusion[^1]. The data lake reference query for Contacts abandoned in X seconds counts contacts with is_abandoned = 1 and a queue time at or under X, which is how you can derive a short-abandon bucket[^1].

### Genesys Cloud

In Genesys Cloud, Abandon counts end users abandoning an interaction in a queue[^21]. Short abandons are abandons disconnected within an interval set by the organization's short abandon setting, which is group A of the abandon interval configuration[^27].

The abandon intervals default to:[^20]

- A (Short): under 6 seconds[^20]
- B: 6-20 seconds[^20]
- C: 20-40 seconds[^20]
- D: 40-60 seconds[^20]
- E: 60-120 seconds[^20]
- F: 120-240 seconds[^20]
- G: over 240 seconds[^20]

The customer short disconnect time is a separate setting[^28]. It is used by the Customer Short Disconnect metric in Flows Performance views, defaults to under 10 seconds, and is independent of the 6-second short abandon interval[^28].

### Webex Contact Center

The Short Call Threshold is a tenant voice setting, in seconds, that decides whether a call is short or abandoned[^45]. It defaults to 0 seconds for a new tenant[^45].

The Total Contacts Abandoned KPI counts contacts terminated in queue before reaching an agent, regardless of handling in earlier queues[^41]. It excludes short calls[^41].

### Five9

Five9 describes abandon rate percentage as the share of inbound calls that customers disconnected before reaching an agent[^19]. Five9 does not publish a formula or threshold on that page[^19].

### Short-abandon defaults compared

This is inferred: Genesys Cloud's short abandon interval defaults to under 6 seconds, while a new Webex Contact Center tenant's Short Call Threshold is 0, so Webex treats no abandons as short until the setting is changed[^49].

## Handle time

### Amazon Connect

The Amazon Connect data lake reference query defines average handle time as the time from contact connection to ACW completion[^4]. The query averages handle_time_ms over handled contacts only[^4].

Average after contact work time averages over every contact record that has either an ACW duration or an agent-connected timestamp[^3]. Connected contacts with no ACW count as zero in that average[^3].

### Genesys Cloud

Genesys Cloud handle time includes contacting, dialing, talk, hold and after-call work time[^24]. Genesys Cloud warns that Avg Talk plus Avg Hold plus Avg ACW does not equal Avg Handle, because only some interactions include holds[^22].

Genesys Cloud uses 30-minute intervals[^25]. Answered is counted in the interval when the agent answered, but Handle is counted in the interval when all of the agent's handle segments, including ACW, finished[^25].

### Webex Contact Center

The Average Handle Time KPI card includes connected time and wrap-up time[^42]. The Sudden Disconnect Threshold decides whether an agent-connected call counts as handled or as ended suddenly, and it defaults to 30 seconds[^48].

## Transfers and contact counting

### Amazon Connect

Amazon Connect models each agent-customer interaction as an independent contact segment with its own contact record[^17]. Each transfer leg is linked to the previous segment by previousContactId[^17].

The contact InitiationMethod field distinguishes TRANSFER from QUEUE_TRANSFER, alongside INBOUND, OUTBOUND, CALLBACK, API and other values[^8]. For TRANSFER and QUEUE_TRANSFER contacts, the InitiationTimestamp is when the transfer was initiated, not when the customer first called[^18]. The data lake reports contacts transferred out of a queue with separate internal and external transferred-out flags[^16].

### Genesys Cloud

In Genesys Cloud, a direct agent-to-agent transfer gives the receiving agent a Handle but no Answered[^34]. A transfer routed through a queue gives the receiving agent both[^34].

### Webex Contact Center

Webex Contact Center Analyzer reports Conference Count as the number of times an agent initiated a conference call[^38]. Analyzer counts contacts handled for an agent as outdial connected count plus connected count[^40].

### Comparing per-call totals

We infer that a transferred call can count as more than one handled contact on a platform that writes one contact record per leg, such as Amazon Connect[^51]. WarmTransfer's reading of the sources is that per-call totals from a migrated platform therefore need stitching by initial contact ID before comparison[^51].

## Pulling Amazon Connect data for comparison

Amazon Connect delivers contact records at least once, so exact counts from the data lake need de-duplication on primary keys[^6]. All data lake timestamps are UTC, and the reference queries bucket contact metrics by disconnect timestamp[^15]. GetMetricData only returns historical metrics for the previous 24 hours, and AWS recommends GetMetricDataV2 for the last 3 months[^7].

## See also

- [Contact center KPI definitions and calculation](https://warmtransfer.net/knowledge/contact-center-kpis)
- [Amazon Connect contact records and streams](https://warmtransfer.net/knowledge/amazon-connect-data)
- [Cross-platform CDR normalization model](https://warmtransfer.net/knowledge/cdr-normalization)
- [Webex Contact Center Analyzer and reporting model](https://warmtransfer.net/knowledge/webex-cc-analyzer)
- [What the Analyzer stock reports publish](https://warmtransfer.net/knowledge/wxcc-analyzer-stock-reports)
- [Troubleshooting Webex Contact Center reporting discrepancies](https://warmtransfer.net/knowledge/wxcc-reporting-discrepancies)
- [Planning a contact center platform migration](https://warmtransfer.net/knowledge/contact-center-migration-planning)

## Applicability

Applies to: Amazon Web Services Amazon Connect, Genesys Cloud, Cisco Webex Contact Center, and Five9 Intelligent CX Platform. Deployments: multi-tenant. Sources checked 2026-10-02.

## What remains uncertain

The Amazon Connect admin guide definition of Service level X, and whether abandons within X count there, is not covered by the sources below. The only documented Amazon Connect treatment is the data lake reference query, which places abandoned contacts in the denominator and never in the numerator[^11].

The Five9 service level formula and its abandon treatment in reports are not covered by the sources below. Five9's abandon rate percentage page does not publish a formula or threshold[^19]. The Five9 Speed of Answer Min Time rule for the ABANDONED column is not covered by the sources below.

The default state of the 3 Genesys Cloud service level options (Include flowouts, Include short abandons and Include abandons in calculation) is not covered by the sources below. Only the separate Include Callback Virtual Wait in SLA option is documented as off by default[^23].

The logic of the Webex Contact Center Analyzer Is Within Service Level field, and how it treats abandons, is not covered by the sources below. Whether Webex Contact Center Connected Duration includes hold is not covered by the sources below. The Average Handle Time KPI card is documented as including connected time and wrap-up time[^42]. How Webex Contact Center counts transfers and child contacts in handled and offered totals is not covered by the sources below.

Whether the Amazon Connect Contacts handled metric counts each transfer leg is not covered by the sources below. We infer only that a transferred call can count as more than one handled contact on a platform that writes one contact record per leg[^51].

## Sources

[^1]: The Amazon Connect data lake reference query for Contacts abandoned in X seconds counts contacts with is_abandoned = 1 and queue time at or under X, which is how a short-abandon bucket can be derived since abandonment rate itself has no short-abandon exclusion. Source: [Reference queries for the Connect Customer data lake](https://docs.aws.amazon.com/connect/latest/adminguide/data-lake-reference-queries.html), Contact and queue metrics > Contacts abandoned in X seconds. Checked 2026-10-02.
[^2]: Amazon Connect computes Abandonment rate as (Contacts abandoned / Contacts queued) * 100, where an abandoned contact is one the customer disconnected while in queue before connecting to an agent. Source: [Metrics definitions in Connect Customer](https://docs.aws.amazon.com/connect/latest/adminguide/metrics-definitions.html), section Abandonment rate, Calculation logic. Checked 2026-10-02.
[^3]: Amazon Connect Average after contact work time averages over every contact record that has either an ACW duration or an agent-connected timestamp, counting connected contacts with no ACW as zero. Source: [Metrics definitions in Connect Customer](https://docs.aws.amazon.com/connect/latest/adminguide/metrics-definitions.html), section Average after contact work time, Calculation logic. Checked 2026-10-02.
[^4]: The Amazon Connect data lake reference query defines average handle time as the time from contact connection to ACW completion, averaging handle_time_ms over handled contacts only. Source: [Reference queries for the Connect Customer data lake](https://docs.aws.amazon.com/connect/latest/adminguide/data-lake-reference-queries.html), Agent performance metrics > Average handle time. Checked 2026-10-02.
[^5]: Amazon Connect does not count contacts queued for callback as abandoned in the Abandonment rate metric. Source: [Metrics definitions in Connect Customer](https://docs.aws.amazon.com/connect/latest/adminguide/metrics-definitions.html), section Abandonment rate, definition paragraph. Checked 2026-10-02.
[^6]: Amazon Connect delivers contact records at least once, so exact counts from the data lake need de-duplication on primary keys. Source: [Reference queries for the Connect Customer data lake](https://docs.aws.amazon.com/connect/latest/adminguide/data-lake-reference-queries.html), Best practices > Deduplication. Checked 2026-10-02.
[^7]: Amazon Connect GetMetricData only returns historical metrics for the previous 24 hours, while AWS recommends GetMetricDataV2 for the last 3 months. Source: [GetMetricData - Amazon Connect API Reference](https://docs.aws.amazon.com/connect/latest/APIReference/API_GetMetricData.html), Note at top of page; Request Body > StartTime. Checked 2026-10-02.
[^8]: Amazon Connect contact InitiationMethod distinguishes TRANSFER from QUEUE_TRANSFER, along with INBOUND, OUTBOUND, CALLBACK, API and other values. Source: [Contact - Amazon Connect API Reference](https://docs.aws.amazon.com/connect/latest/APIReference/API_Contact.html), Contents > InitiationMethod valid values. Checked 2026-10-02.
[^9]: In Amazon Connect, Abandonment rate shows 0 when results are grouped by routing profile, because an abandoned contact never reaches an agent and so has no routing profile. Source: [Metrics definitions in Connect Customer](https://docs.aws.amazon.com/connect/latest/adminguide/metrics-definitions.html), section Abandonment rate, note on Routing Profile grouping. Checked 2026-10-02.
[^10]: The Amazon Connect GetMetricData API accepts a SERVICE_LEVEL threshold of any whole number of seconds from 1 to 604800 and requires the comparison LT (less than). Source: [GetMetricData - Amazon Connect API Reference](https://docs.aws.amazon.com/connect/latest/APIReference/API_GetMetricData.html), Request Body > HistoricalMetrics > SERVICE_LEVEL. Checked 2026-10-02.
[^11]: The Amazon Connect data lake reference query for service level counts contacts that were handled with queue answer time at or under X and divides by all queued contacts, so abandoned contacts sit in the denominator and never in the numerator. Source: [Reference queries for the Connect Customer data lake](https://docs.aws.amazon.com/connect/latest/adminguide/data-lake-reference-queries.html), Contact and queue metrics > Service level (anchor data-lake-rq-service-level). Checked 2026-10-02.
[^12]: The Amazon Connect GetMetricData reference links the SERVICE_LEVEL metric's UI name to Average queue abandon time, which is a documentation labelling error rather than the metric's meaning (inferred). Source: [GetMetricData - Amazon Connect API Reference](https://docs.aws.amazon.com/connect/latest/APIReference/API_GetMetricData.html), Request Body > HistoricalMetrics > SERVICE_LEVEL > UI name. Checked 2026-10-02.
[^13]: A contact answered exactly at the threshold counts as within service level in the Amazon Connect data lake sample query (<=) but not under the GetMetricData LT comparison, so the two can differ slightly at the boundary (inferred). Source: [Reference queries for the Connect Customer data lake](https://docs.aws.amazon.com/connect/latest/adminguide/data-lake-reference-queries.html), Service level query (<= 20000) compared with GetMetricData SERVICE_LEVEL Comparison LT. Checked 2026-10-02.
[^14]: A single Amazon Connect GetMetricData request can include up to 20 SERVICE_LEVEL metrics, so several thresholds can be pulled at once. Source: [GetMetricData - Amazon Connect API Reference](https://docs.aws.amazon.com/connect/latest/APIReference/API_GetMetricData.html), Request Body > HistoricalMetrics > SERVICE_LEVEL. Checked 2026-10-02.
[^15]: All Amazon Connect data lake timestamps are UTC and the reference queries bucket contact metrics by disconnect timestamp. Source: [Reference queries for the Connect Customer data lake](https://docs.aws.amazon.com/connect/latest/adminguide/data-lake-reference-queries.html), Best practices > Time zones; WHERE clauses of contact and queue queries. Checked 2026-10-02.
[^16]: The Amazon Connect data lake reports contacts transferred out of a queue with separate internal and external transferred-out flags. Source: [Reference queries for the Connect Customer data lake](https://docs.aws.amazon.com/connect/latest/adminguide/data-lake-reference-queries.html), Contact and queue metrics > Contacts transferred out. Checked 2026-10-02.
[^17]: Amazon Connect models each agent-customer interaction as an independent contact segment with its own contact record, and a transfer leg is linked to the previous segment by previousContactId. Source: [Model contact transfers and conferencing in Connect Customer](https://docs.aws.amazon.com/connect/latest/adminguide/model-contact-transfers-conferencing.html), introductory bullet list before Workflow for warm transfer. Checked 2026-10-02.
[^18]: For Amazon Connect TRANSFER and QUEUE_TRANSFER contacts the InitiationTimestamp is when the transfer was initiated, not when the customer first called. Source: [Contact - Amazon Connect API Reference](https://docs.aws.amazon.com/connect/latest/APIReference/API_Contact.html), Contents > InitiationTimestamp. Checked 2026-10-02.
[^19]: Five9 describes abandon rate percentage as the share of inbound calls disconnected by customers before reaching an agent, without publishing a formula or threshold on that page. Source: [What is Abandon Rate Percentage](https://www.five9.com/faq/what-is-abandon-rate-percentages-arp), FAQ What is Abandon Rate Percentage, opening paragraph. Checked 2026-10-02.
[^20]: Genesys Cloud abandon intervals default to column A (Short) under 6 seconds, then B 6-20, C 20-40, D 40-60, E 60-120, F 120-240 and G over 240 seconds. Source: [Configure contact center analytics options](https://help.genesys.cloud/articles/configure-contact-center-analytics-options/), section Configure abandon intervals. Checked 2026-10-02.
[^21]: Genesys Cloud computes Abandon % as (Abandoned count / Offered count) * 100, where Abandon counts end users abandoning an interaction in a queue. Source: [Metric definitions](https://help.genesys.cloud/articles/metric-definitions/), metric rows Abandon and Abandon %. Checked 2026-10-02.
[^22]: Genesys Cloud warns that Avg Talk plus Avg Hold plus Avg ACW does not equal Avg Handle because only some interactions include holds. Source: [Metric definitions](https://help.genesys.cloud/articles/metric-definitions/), note near Avg Handle column definitions. Checked 2026-10-02.
[^23]: Genesys Cloud's Include Callback Virtual Wait in SLA option is off by default and, when on, judges callbacks on callback outcomes (reached, not reached, abandoned) using virtual wait from callback offer to outcome. Source: [service level (glossary)](https://help.genesys.cloud/glossary/service-level/), glossary entry service level, Include Callback Virtual Wait in SLA paragraph. Checked 2026-10-02.
[^24]: Genesys Cloud handle time includes contacting, dialing, talk, hold and after-call work time. Source: [handle time - Genesys Cloud Resource Center](https://help.genesys.cloud/glossary/handle-time/), glossary entry handle time. Checked 2026-10-02.
[^25]: Genesys Cloud counts Answered in the interval when the agent answered but counts Handle in the interval when all of the agent's handle segments including ACW finished, using 30-minute intervals. Source: [What is the difference between the answered and handle metrics? - Genesys Cloud Resource Center](https://help.genesys.cloud/faqs/what-is-the-difference-between-the-answered-and-handle-metrics/), FAQ paragraph on timing, 10:25 AM example. Checked 2026-10-02.
[^26]: Editing Genesys Cloud contact center analytics settings requires Analytics > Insight Configurations > Edit and Analytics > reportingSettings > Edit permissions. Source: [Configure contact center analytics options](https://help.genesys.cloud/articles/configure-contact-center-analytics-options/), Prerequisites / permissions block. Checked 2026-10-02.
[^27]: Genesys Cloud defines short abandons as abandons disconnected within an interval set by the organization's short abandon setting, which is group A of the abandon interval configuration. Source: [short abandons - Genesys Cloud Resource Center](https://help.genesys.cloud/glossary/short-abandons/), glossary entry short abandons. Checked 2026-10-02.
[^28]: Genesys Cloud's customer short disconnect time, used by the Customer Short Disconnect metric in Flows Performance views, defaults to under 10 seconds and is separate from the 6-second short abandon interval. Source: [Configure contact center analytics options](https://help.genesys.cloud/articles/configure-contact-center-analytics-options/), section Configure analytics settings, short disconnect time. Checked 2026-10-02.
[^29]: With abandons included, Genesys Cloud service level is (tAnswered − nOverSla) / (tAnswered + tAbandon), so every abandon counts as missing the target. Source: [service level (glossary)](https://help.genesys.cloud/glossary/service-level/), glossary entry service level, formula for Include abandons. Checked 2026-10-02.
[^30]: Genesys Cloud defines service level as conversations answered within the target time divided by the total conversations included in the calculation, with the included set controlled by configurable options. Source: [service level (glossary)](https://help.genesys.cloud/glossary/service-level/), glossary entry service level, opening definition and basic formula. Checked 2026-10-02.
[^31]: When flowouts are included, Genesys Cloud counts conversations removed from the queue by a flow action before answer as missing service level, using (tAnswered − nOverSla) / (tAnswered + tFlowOut). Source: [service level (glossary)](https://help.genesys.cloud/glossary/service-level/), glossary entry service level, flowouts option and formula. Checked 2026-10-02.
[^32]: When short abandons are excluded, Genesys Cloud service level becomes (tAnswered − nOverSla) / (tAnswered + tAbandon − tShortAbandon). Source: [service level (glossary)](https://help.genesys.cloud/glossary/service-level/), glossary entry service level, short abandons option and formula. Checked 2026-10-02.
[^33]: Genesys Cloud exposes three organization-level service level options in contact center analytics settings: Include flowouts, Include short abandons and Include abandons in calculation. Source: [Configure contact center analytics options](https://help.genesys.cloud/articles/configure-contact-center-analytics-options/), section Configure analytics settings, Settings tab > Service Level. Checked 2026-10-02.
[^34]: In Genesys Cloud a direct agent-to-agent transfer gives the receiving agent a Handle but no Answered, while a transfer routed through a queue gives the receiving agent both. Source: [What is the difference between the answered and handle metrics? - Genesys Cloud Resource Center](https://help.genesys.cloud/faqs/what-is-the-difference-between-the-answered-and-handle-metrics/), FAQ paragraph on transfers. Checked 2026-10-02.
[^35]: Webex Contact Center Analyzer's Agent Transfer measure counts inbound contacts an agent transferred to another agent after consult, as the sum of Agent To Agent Transfer Count. Source: [Webex Contact Center Analyzer Stock Reports](https://help.webex.com/en-us/article/t137o3/Webex-Contact-Center-Analyzer-Stock-Reports), Agent Details report, Agent Transfer row. Checked 2026-10-02.
[^36]: The Webex Contact Center Analyzer Agent Details report computes average handle time as (sum of connected duration + sum of wrap-up duration) / sum of connected count, for inbound calls. Source: [Cisco Webex Contact Center Analyzer User Guide](https://help.webex.com/en-us/article/tajemk/Cisco-Webex-Contact-Center-Analyzer-User-Guide), Historical Reports > Agent Details, Average Handle Time row. Checked 2026-10-02.
[^37]: Webex Contact Center AHT can differ between the KPI card, Analyzer Agent Details and the Agent Performance Dashboard because the documented formulas differ in whether hold is named and in the denominator used (inferred). Source: [Webex Contact Center Analyzer Stock Reports](https://help.webex.com/en-us/article/t137o3/Webex-Contact-Center-Analyzer-Stock-Reports), Agent Performance Dashboard AHT compared with cisco-help-tajemk-wxcc-analyzer-user-guide Agent Details AHT and cisco-help-63s0bj-wxcc-kpi-cards AHT. Checked 2026-10-02.
[^38]: Webex Contact Center Analyzer reports Conference Count as the number of times an agent initiated a conference call. Source: [Webex Contact Center Analyzer Stock Reports](https://help.webex.com/en-us/article/t137o3/Webex-Contact-Center-Analyzer-Stock-Reports), Agent Details report, Conference Count row. Checked 2026-10-02.
[^39]: The Webex Contact Center Analyzer Agent Performance Dashboard computes average handle time as (hold + connected + wrap-up durations) / count of contact sessions with termination type Normal. Source: [Webex Contact Center Analyzer Stock Reports](https://help.webex.com/en-us/article/t137o3/Webex-Contact-Center-Analyzer-Stock-Reports), Agent Performance Dashboard, Average Handle Time row. Checked 2026-10-02.
[^40]: Webex Contact Center Analyzer counts contacts handled for an agent as outdial connected count plus connected count. Source: [Cisco Webex Contact Center Analyzer User Guide](https://help.webex.com/en-us/article/tajemk/Cisco-Webex-Contact-Center-Analyzer-User-Guide), Historical Reports > Agent Details, Contacts Handled row. Checked 2026-10-02.
[^41]: The Webex Contact Center Total Contacts Abandoned KPI counts contacts terminated in queue before reaching an agent, regardless of handling in earlier queues, and excludes short calls. Source: [View your contact center KPI cards](https://help.webex.com/en-us/article/63s0bj/View-your-contact-center-KPI-cards), KPI Total Contacts Abandoned. Checked 2026-10-02.
[^42]: The Webex Contact Center Average Handle Time KPI card includes connected time and wrap-up time. Source: [View your contact center KPI cards](https://help.webex.com/en-us/article/63s0bj/View-your-contact-center-KPI-cards), KPI Average Handle Time. Checked 2026-10-02.
[^43]: The Webex Contact Center Average Service Level KPI counts contacts that were connected, opted out of the queue, dequeued or terminated in queue within the SLA time as within service level. Source: [View your contact center KPI cards](https://help.webex.com/en-us/article/63s0bj/View-your-contact-center-KPI-cards), KPI Average Service Level. Checked 2026-10-02.
[^44]: Because the Webex Contact Center service level KPI counts contacts terminated in queue within the SLA as within service level, early abandons appear to raise that KPI, the opposite of Genesys Cloud and the Amazon Connect data lake query (inferred). Source: [View your contact center KPI cards](https://help.webex.com/en-us/article/63s0bj/View-your-contact-center-KPI-cards), KPI Average Service Level read against genesys-glossary-service-level and aws-connect-data-lake-reference-queries Service level. Checked 2026-10-02.
[^45]: Webex Contact Center's Short Call Threshold is a tenant voice setting in seconds that decides whether a call is short or abandoned, and it defaults to 0 seconds for a new tenant. Source: [Set up voice settings for Webex Contact Center](https://help.webex.com/en-us/article/gsmsaeb/Set-up-voice-settings-for-Webex-Contact-Center), Control Hub > Contact Center > Tenant Settings > Voice, Short Call Threshold. Checked 2026-10-02.
[^46]: Webex Contact Center caps the queue service level threshold at 86400 seconds for telephony and chat, 604800 for digital and 1209600 for email. Source: [System limits in Webex Contact Center](https://help.webex.com/en-us/article/n7s6ed7/System-limits-in-Webex-Contact-Center), Contact service queues limits, Service level threshold rows. Checked 2026-10-02.
[^47]: In Webex Contact Center the service level threshold is a per-queue advanced setting giving the time a request can wait in queue before being flagged outside service level, and it cannot be set for social channels. Source: [Create queues and configure routing patterns](https://help.webex.com/en-us/article/ubg5qd/Create-queues-and-configure-routing-patterns), section Advanced Settings, Service level threshold field. Checked 2026-10-02.
[^48]: Webex Contact Center's Sudden Disconnect Threshold, defaulting to 30 seconds, decides whether an agent-connected call counts as handled or as ended suddenly. Source: [Set up voice settings for Webex Contact Center](https://help.webex.com/en-us/article/gsmsaeb/Set-up-voice-settings-for-Webex-Contact-Center), Tenant Settings > Voice, Sudden Disconnect Threshold. Checked 2026-10-02.
[^49]: Short-abandon defaults differ: Genesys Cloud's short abandon interval defaults to under 6 seconds, while a new Webex Contact Center tenant's Short Call Threshold is 0 so no abandons are treated as short until changed (inferred). Source: [Set up voice settings for Webex Contact Center](https://help.webex.com/en-us/article/gsmsaeb/Set-up-voice-settings-for-Webex-Contact-Center), Short Call Threshold default compared with genesys-configure-analytics-options Configure abandon intervals. Checked 2026-10-02.
[^50]: The same call traffic can produce different service level percentages on Amazon Connect, Genesys Cloud and Webex Contact Center because their documented treatment of abandons in the formula differs (inferred). Source: [service level (glossary)](https://help.genesys.cloud/glossary/service-level/), service level formulas compared with aws-connect-data-lake-reference-queries Service level and cisco-help-63s0bj-wxcc-kpi-cards Average Service Level. Checked 2026-10-02.
[^51]: A transferred call can count as more than one handled contact on a platform that writes one contact record per leg, such as Amazon Connect, so per-call totals from a migrated platform need stitching by initial contact ID before comparison (inferred). Source: [Model contact transfers and conferencing in Connect Customer](https://docs.aws.amazon.com/connect/latest/adminguide/model-contact-transfers-conferencing.html), intro bullets on contact segments, read with aws-connect-api-contact-datatype InitialContactId. Checked 2026-10-02.
[^52]: Because Five9 service level drops calls abandoned in queue before the threshold and calls shorter than the minimum time, a Five9 service level figure is not directly comparable with a platform that keeps all offered calls in the denominator (inferred). Source: [Administrator - Setting Key Performance Indicators (KPI)](https://community.five9.com/s/article/Administrator-Setting-Key-Performance-Indicators-KPI?language=en_US), Speed of Answer threshold and Speed of Answer Minimum Time sections. Checked 2026-10-14.