support-analytics

Use this skill when measuring CSAT, NPS, resolution time, deflection rates, or analyzing support trends. Triggers on CSAT, NPS, resolution time, deflection rate, support metrics, trend analysis, support reporting, and any task requiring customer support data analysis or reporting.

What is support-analytics?

Quick Start

Open your terminal or command prompt
Run: npx skills add AbsolutelySkilled/AbsolutelySkilled --skill support-analytics
Start your AI coding agent (Claude Code, Cursor, Gemini CLI, or any supported agent)
The support-analytics skill is now active and ready to use

Overview Files

support-analytics

support-analytics is a production-ready AI agent skill for claude-code, gemini-cli, openai-codex. Measuring CSAT, NPS, resolution time, deflection rates, or analyzing support trends.

Quick Facts

Field	Value
Category	operations
Version	0.1.0
Platforms	claude-code, gemini-cli, openai-codex
License	MIT

How to Install

Make sure you have Node.js installed on your machine.
Run the following command in your terminal:

npx skills add AbsolutelySkilled/AbsolutelySkilled --skill support-analytics

The support-analytics skill is now available in your AI coding agent (Claude Code, Gemini CLI, OpenAI Codex, etc.).

Overview

Support analytics turns raw ticket data into operational intelligence. The goal is not to generate reports - it is to change behavior. Whether measuring how satisfied customers are after an interaction, how quickly issues are resolved, or how often customers find answers without contacting support, every metric should connect to a decision. This skill covers the full analytics lifecycle: what to measure, how to measure it, and how to act on what you find.

Platforms

claude-code
gemini-cli
openai-codex

Related Skills

Pair support-analytics with these complementary skills:

Frequently Asked Questions

What is support-analytics?

How do I install support-analytics?

Run npx skills add AbsolutelySkilled/AbsolutelySkilled --skill support-analytics in your terminal. The skill will be immediately available in your AI coding agent.

What AI agents support support-analytics?

This skill works with claude-code, gemini-cli, openai-codex. Install it once and use it across any supported AI coding agent.

Maintainers

@maddhruv

Generated from AbsolutelySkilled

SKILL.md

Support Analytics

When to use this skill

Trigger this skill when the user:

Wants to set up or improve a CSAT or NPS measurement program
Needs to track, report on, or reduce resolution time or first-contact resolution
Asks about deflection rate or self-service effectiveness
Wants to analyze support ticket trends, topic clusters, or volume forecasting
Needs to build a support dashboard for an executive, team lead, or agent
Is creating a support metrics framework or KPI hierarchy
Asks about survey design, response rate improvement, or score interpretation
Needs to segment support data by channel, tier, topic, or agent

Do NOT trigger this skill for:

Product analytics or funnel metrics (use analytics-engineering instead)
Infrastructure monitoring, SLOs, or error rate tracking (use backend-engineering instead)

Key principles

Measure what matters, not what's easy - Ticket volume is easy to count but rarely actionable on its own. Focus on metrics that reveal customer experience and operational efficiency: CSAT, resolution time, and deflection rate expose the health of your support operation far more than raw volume does.
Benchmarks are starting points, not goals - Industry benchmarks give you a calibration point, not a finish line. A CSAT of 85% may be excellent for a complex enterprise product and unacceptable for a consumer app. Compare to your own historical trend first; compare to benchmarks second.
Trends matter more than snapshots - A single week's CSAT score means almost nothing. A 12-week trend that is declining 1 point per week means something is systematically wrong. Always show time-series data alongside point-in-time figures. Week-over-week and month-over-month comparisons prevent overreaction to normal variance.
Segment by channel, tier, and topic - Aggregate scores hide the story. A CSAT of 82% overall might mask a chat score of 91% and an email score of 68%. Segmenting by channel, customer tier, product area, and ticket topic reveals where to invest and what is working.
Close the loop - insights to action - An analytics program that produces dashboards no one acts on is a cost center. Every metric should own a DRI (directly responsible individual), a target, and a process for escalating when the target is missed. The cadence is: measure, review, decide, act, re-measure.

Core concepts

Satisfaction metrics

CSAT (Customer Satisfaction Score) - A post-interaction rating, typically 1-5 stars or a thumbs up/down, sent immediately after a ticket closes. Measures satisfaction with a specific support interaction, not the product overall. The score is the percentage of positive responses out of total responses received.

NPS (Net Promoter Score) - A relationship-level survey asking "How likely are you to recommend us to a colleague?" on a 0-10 scale. Promoters (9-10) minus Detractors (0-6) equals the NPS. Transactional NPS (tNPS) is sent after support interactions to capture loyalty impact from a specific resolution.

CES (Customer Effort Score) - Measures how easy it was to get help: "How much effort did you personally have to put forth to handle your request?" Low effort correlates with reduced churn more reliably than high satisfaction does.

Operational metrics

First Contact Resolution (FCR) - The percentage of tickets resolved on the first reply without the customer needing to follow up. High FCR is the single strongest predictor of high CSAT. Improving FCR reduces cost and improves satisfaction simultaneously.

Resolution Time - The elapsed time from ticket creation to resolution. Report as median (p50) and p90 to capture both typical experience and worst-case outliers. Segment by ticket priority, channel, and topic - a blanket average hides whether P1 bugs are being prioritized over billing questions.

Handle Time - Agent-active time spent on a ticket (not elapsed clock time). Useful for capacity planning and identifying where agents need tooling or training improvements.

Reopen Rate - Percentage of resolved tickets reopened by the customer. A high reopen rate indicates resolutions are incomplete or unclear, or that the underlying issue is recurring.

Self-service metrics

Deflection Rate - The percentage of potential support contacts handled by self-service (docs, chatbot, FAQ) without reaching a human. Calculated as deflections / (deflections + human contacts). Hard to measure precisely - proxy methods include doc views before ticket submission and chatbot resolution rates.

Article Effectiveness - For knowledge bases: the percentage of doc views that end without a support ticket being submitted. Track alongside search-with-no-results counts to identify content gaps.

Containment Rate - For chatbots and IVR: the percentage of sessions that reach a resolution without escalating to a human. A session can be contained but still leave the customer unsatisfied - always pair with a satisfaction signal.

Quality metrics

QA Score - Internal quality assurance review of ticket handling: tone, accuracy, policy adherence, completeness. Typically sampled (5-10% of tickets) and scored on a rubric. Correlates with CSAT but catches issues that surveys miss such as correct but cold responses.

Agent CSAT - CSAT segmented by individual agent. Useful for coaching, not for ranking. Agents on complex ticket queues will have lower scores than agents on simple billing questions - normalize by ticket type before comparing agents.

Common tasks

Set up a metrics framework - KPI hierarchy

Build a three-tier hierarchy: strategic, operational, and diagnostic.

Tier	Audience	Cadence	Examples
Strategic	Leadership	Monthly / Quarterly	NPS, CSAT trend, cost-per-ticket, deflection rate
Operational	Support managers	Weekly	FCR, median resolution time, reopen rate, volume by channel
Diagnostic	Team leads, agents	Daily	Queue depth, SLA breach rate, handle time, QA score

Start by identifying who reads each metric and what decision it drives. If no one owns the decision triggered by a metric, do not track it yet.

Steps:

List current pain points from support team retrospectives
Map each pain point to a metric category (satisfaction, operational, quality)
Define the measurement method and data source for each metric
Assign a DRI and a target for each metric
Build the minimal dashboard needed to surface all three tiers

Measure and improve CSAT - survey design and analysis

Survey design checklist:

Send within 1 hour of ticket close - response rate drops sharply after 24 hours
Keep to 1-2 questions: the rating plus one optional free-text follow-up
Use a consistent scale - do not mix 5-star with thumbs up/down across touchpoints
Personalize the subject line with the agent's name and ticket topic

Calculation:

CSAT = (4-star + 5-star responses) / total responses * 100

Analysis steps:

Segment by channel, agent, ticket category, and customer tier
Tag all 1-2 star responses within 24 hours - look for patterns in verbatim feedback
Build a weekly trend chart with 4-week moving average to smooth noise
Create a detractor recovery workflow: manager outreach within 24 hours for any 1-star

Improving response rate:

Subject line "How did [Agent Name] do?" outperforms generic phrasing
Mobile-optimized survey - most customers open on phone
Remove login requirement - anonymous responses get 2-3x higher response rate

Implement NPS program - collection and segmentation

Collection strategy:

Send after significant support interactions (not every ticket)
Trigger rules: send after complex tickets, P1 resolutions, or any escalation closed
Suppress repeat surveys: do not survey the same customer more than once every 90 days

Calculation:

NPS = Promoters% - Detractors%

Example: 60% promoters, 15% detractors, 25% passives
NPS = 60 - 15 = 45

Segmentation framework:

Segment	Score	Action
Promoters	9-10	Case studies, referral asks, community invites
Passives	7-8	Identify friction - most at risk of churn on next negative event
Detractors	0-6	Close-the-loop call within 48 hours; flag to CSM if enterprise tier

Segment NPS by customer tier, product area, support channel, and account age. New customers tend to score differently than long-tenured accounts.

Track and optimize resolution time

Measurement setup:

Track created_at to resolved_at in your ticketing system
Report median (p50) and 90th percentile (p90) - averages mask outlier drag
Exclude pending-customer time from elapsed calculation (clock pauses when waiting on customer)

SLA framework:

Priority	Target Resolution	Alert At
P1 - Service down	4 hours	2 hours
P2 - Major feature broken	24 hours	16 hours
P3 - Minor issue / workaround available	72 hours	48 hours
P4 - Question / enhancement	7 days	5 days

Root cause analysis for high resolution time:

Identify the top 10% slowest tickets in a period
Tag reasons: awaiting escalation, waiting on engineering, reassigned, unclear ask
Quantify each reason as a percentage of slow tickets
Prioritize fixes by volume x impact - routing logic and escalation paths are typically top two

A declining resolution time with a rising reopen rate means agents are closing tickets prematurely. Always track both together.

Measure deflection rate - self-service effectiveness

Proxy measurement methods (direct deflection is rarely measurable):

Doc-to-ticket ratio - Track customers who viewed a help article and then submitted a ticket within 30 minutes. Low ratio means effective docs.
Chatbot containment - % of chatbot sessions that reach resolution without escalating to a human. Target 40-60% for most support types.
Search abandonment - In your help center, track searches that end without a page view. High abandonment signals a content gap.
Before/after experiment - Publish a new article on a common topic, compare ticket volume for that topic over the next 30 days vs prior 30 days.

Improving deflection:

Run monthly content gap analysis: top 20 ticket topics vs help center coverage
Add article links to auto-acknowledgment emails for common categories
Implement a post-submission deflection prompt: show matching articles after ticket submit

Analyze support trends - topic clustering and forecasting

Topic clustering workflow:

Export ticket titles and first customer messages for a 30-90 day window
Group tickets by existing tags first - identify gaps where >10% have no tag
Use keyword frequency on untagged tickets to surface emerging topics
Update your taxonomy - aim for 80%+ of tickets tagged to a specific topic
Review top 10 topics weekly; track volume trend, CSAT, and resolution time per topic

Volume forecasting:

Use 12 weeks of weekly ticket volume as baseline
Apply seasonal adjustment for known events (product launches, billing cycles, holidays)
4-week trailing average with +20% buffer as capacity target
Flag any week where volume exceeds forecast by >30% as an anomaly requiring investigation

Trend signals to monitor:

New topic appearing in top 10 that was not there last month - possible product regression
CSAT drop on a specific topic without volume change - agent knowledge gap or policy confusion
Resolution time increase on one channel only - tooling or routing issue

Build support dashboards - by audience

Executive dashboard (monthly business review):

Panel	Metric	Visualization
Customer Sentiment	CSAT 12-month trend + NPS	Line chart with benchmark line
Efficiency	Cost per ticket, deflection rate	KPI card + trend sparkline
Volume	Total contacts by channel	Stacked bar, MoM comparison
Highlights	Top 3 topic drivers, worst-performing category	Table

Manager dashboard (weekly ops review):

Panel	Metric	Visualization
Volume	Tickets opened/closed, backlog	Area chart
Quality	CSAT by channel, reopen rate	Bar chart
Speed	Median + p90 resolution time vs SLA	Gauge + trend
Team	FCR by agent, QA scores	Table with conditional formatting

Agent dashboard (daily view):

Personal queue: open tickets, SLA risk, oldest unresolved
Personal CSAT for last 30 days (not ranked against peers)
Today's handle time vs personal average

Gotchas

CSAT surveys sent more than 24 hours after ticket close get response bias - Surveys sent days after resolution disproportionately capture customers who had extreme experiences (very positive or very negative) because neutral customers have moved on. Automate delivery within 1 hour of ticket close to get a representative sample.
FCR self-reporting by agents inflates the metric - If agents mark tickets as "resolved first contact" manually, they will mark optimistically. FCR should be measured by the ticketing system based on whether the customer reopened or submitted a new ticket on the same topic within 72 hours, not by agent judgment.
Chatbot containment rate hides frustrated escalation paths - If customers cannot find the escalation button, your containment rate looks great while your CSAT tanks. Always pair containment rate with a post-deflection CSAT signal (even a thumbs up/down) to distinguish genuinely resolved sessions from abandoned ones.
Normalizing agent CSAT by ticket type requires a large sample - Comparing agents with statistical significance requires at minimum 30 surveys per agent per segment. Trying to normalize by ticket type with small sample sizes produces rankings that are noise, not signal. Use QA score for coaching with small agent pools instead.
Volume forecasting without seasonality adjustments leads to understaffing - Applying a flat growth rate to weekly volume ignores known spikes (product launches, billing cycle dates, end-of-fiscal-year surges). Build a seasonal adjustment factor by comparing the same week across prior years before making staffing decisions.

Anti-patterns

Anti-pattern	Why it's wrong	What to do instead
Tracking CSAT average without response rate	A 95% CSAT from 3% response rate is meaningless - response bias distorts the score	Always report response rate alongside CSAT; investigate if below 15%
Comparing agent CSAT without normalizing by ticket type	Agents on billing queues outscore agents on complex bug reports by default	Segment CSAT by ticket category before comparing agents; use for coaching only
Reporting resolution time as an average	Averages are pulled high by a small number of outliers, masking the typical experience	Use median (p50) as primary; add p90 to surface worst-case
Measuring deflection rate from chatbot containment alone	Bots can block escalation paths, yielding high containment and low satisfaction	Pair containment with post-deflection CSAT; 0 escalations + low satisfaction is a false positive
Building dashboards without a decision owner	Dashboards created without a defined reviewer become shelfware	Identify the decision each dashboard drives before building; assign a weekly reviewer
Chasing benchmark NPS without context	A software company and a logistics provider should not share the same NPS target	Set targets relative to your own historical trend and competitive cohort, not generic benchmarks

References

For detailed content on specific topics, read the relevant file from references/:

references/metrics-benchmarks.md - Industry benchmarks for CSAT, NPS, resolution time, and deflection rate by company size and vertical

Only load a references file if the current task requires deep detail on that topic.

References

benchmarks.md

Support Analytics - Industry Benchmarks

Benchmarks for customer support metrics segmented by company size, vertical, and channel. Use these as starting points for target-setting, not as absolute standards. Every benchmark should be validated against your own historical data.

CSAT benchmarks

By industry vertical

Vertical	Median CSAT	Top quartile	Notes
SaaS / Software	78%	88%+	Higher for B2B than B2C
E-commerce	80%	90%+	Driven by shipping and returns experience
Financial services	75%	85%+	Heavily regulated, slower resolution expected
Healthcare	72%	82%+	Complex issues, privacy constraints
Telecommunications	68%	78%+	Historically lowest due to billing complexity
Hospitality / Travel	76%	86%+	High emotional stakes, seasonality

By support channel

Channel	Typical CSAT	Why
Live chat	82-88%	Fast, convenient, in-context
Phone	78-85%	Personal but wait times hurt scores
Email / ticket	72-80%	Slow, lacks real-time back-and-forth
Social media	65-75%	Public complaints, often escalated issues
Self-service (when successful)	85-92%	Instant, no waiting, user feels empowered

By company stage

Stage	Typical CSAT	Context
Early startup (< 50 customers)	90%+	Founders answering tickets, high-touch
Growth (50-500 customers)	78-85%	Scaling pains, first hires learning
Scale (500-5000 customers)	75-82%	Process-driven, specialization emerging
Enterprise (5000+ customers)	72-80%	Tiered support, more complex issues

NPS benchmarks

By industry

Vertical	Median NPS	Top quartile	World class
SaaS / Software	+30	+50	+70
E-commerce	+40	+60	+75
Financial services	+20	+40	+60
Healthcare	+15	+35	+55
Telecommunications	+5	+25	+45

NPS interpretation scale

NPS range	Interpretation
-100 to 0	Critical - systemic issues, high churn risk
0 to +20	Below average - significant improvement needed
+20 to +50	Good - competitive but room for improvement
+50 to +70	Excellent - strong loyalty and word-of-mouth
+70 to +100	World class - rare, typically category leaders

Resolution time benchmarks

First Response Time (FRT) by channel

Channel	Good	Average	Poor
Live chat	< 30 seconds	1-3 minutes	> 5 minutes
Phone	< 60 seconds	2-5 minutes	> 10 minutes
Email / ticket	< 4 hours	4-12 hours	> 24 hours
Social media	< 1 hour	1-4 hours	> 8 hours

Average Resolution Time by complexity

Complexity	Good	Average	Poor
Simple (how-to, password reset)	< 2 hours	2-8 hours	> 24 hours
Medium (config issue, bug report)	< 8 hours	8-24 hours	> 48 hours
Complex (integration, data issue)	< 24 hours	1-3 days	> 5 days
Critical (outage, data loss)	< 4 hours	4-8 hours	> 12 hours

Other efficiency benchmarks

Metric	Good	Average	Needs attention
One-touch resolution rate	> 60%	40-60%	< 40%
Reopen rate	< 5%	5-10%	> 10%
Tickets per agent per day	8-15	15-25	> 25 (burnout risk)
Agent utilization	60-75%	75-85%	> 85% (no buffer)

Deflection rate benchmarks

By self-service channel

Channel	Good	Average	Poor
Knowledge base / help center	30-45%	15-30%	< 15%
AI chatbot	25-40%	10-25%	< 10%
Community forum	15-25%	5-15%	< 5%
In-app contextual help	35-50%	20-35%	< 20%

Overall deflection by company maturity

Maturity	Typical rate	What drives it
No self-service investment	5-10%	Incidental (users Googling)
Basic help center	15-25%	Static articles for top issues
Optimized self-service	25-40%	Data-driven content, chatbot, in-app help
AI-augmented	35-50%+	ML-powered search, generative answers, proactive help

Cost benchmarks

Cost per ticket by channel

Channel	Low	Median	High
Phone	$6	$10	$16
Email / ticket	$3	$5	$8
Live chat	$2	$4	$6
AI chatbot (automated)	$0.10	$0.50	$1.00
Self-service (knowledge base)	$0.05	$0.25	$0.50

Support cost as percentage of revenue

Company stage	Typical range	Context
Early stage	15-25%	Small team, high per-customer cost
Growth	8-15%	Scaling with process, some automation
Scale	5-10%	Mature processes, deflection working
Enterprise	3-7%	Tiered support, heavy self-service

How to use these benchmarks

Start with your own baseline - measure your current metrics for 4-8 weeks before comparing to benchmarks
Find your closest peer group - match by vertical, company size, and support channel mix
Set targets relative to your baseline - aim for 10-15% improvement per quarter, not an immediate jump to "top quartile"
Benchmark internally first - compare teams, channels, and segments within your org before looking externally
Update annually - industry benchmarks shift as tooling and customer expectations evolve

metric-formulas.md

Support Metrics - Formula Reference

Complete formula reference for all support analytics metrics, including statistical considerations for reliable measurement.

CSAT formulas

Basic CSAT score

CSAT % = (count of 4 and 5 responses) / (total responses) * 100

Weighted CSAT (when response volumes differ across segments)

Weighted CSAT = SUM(segment_csat * segment_ticket_volume) / total_ticket_volume

Use weighted CSAT when comparing across segments with very different volumes. A segment with 10 responses and 100% CSAT should not dominate reporting.

CSAT confidence interval

For a given sample size and CSAT score, the 95% confidence interval is:

margin_of_error = 1.96 * sqrt((csat * (1 - csat)) / n)

Example:
  CSAT = 0.82 (82%), n = 200 responses
  margin = 1.96 * sqrt(0.82 * 0.18 / 200) = 1.96 * 0.0272 = 0.053
  95% CI: 76.7% to 87.3%

Minimum sample sizes for reliable CSAT:

Desired margin of error	Required responses (at 80% CSAT)
+/- 10%	62
+/- 5%	246
+/- 3%	683
+/- 1%	6,147

If your weekly response count falls below 62, report CSAT as directional only, not as a reliable metric.

Response rate and bias

Response Rate = total_responses / total_resolved_tickets * 100

Response rates below 10% introduce severe selection bias - dissatisfied customers are more likely to respond, skewing CSAT downward. Conversely, if only prompted users respond, CSAT may skew upward.

Corrective strategies:

Target 15-25% response rate as the reliability threshold
Compare respondent demographics to full ticket population
Use stratified sampling if certain segments under-respond

NPS formulas

Standard NPS

NPS = (% Promoters) - (% Detractors)

Where:
  Promoters:  score 9 or 10
  Passives:   score 7 or 8
  Detractors: score 0 through 6

Range: -100 to +100

NPS margin of error

NPS standard error = sqrt((p_promoter * (1 - p_promoter) + p_detractor * (1 - p_detractor)
                          + 2 * p_promoter * p_detractor) / n)

95% CI = NPS +/- 1.96 * standard_error

Example:
  n = 300, Promoters = 40%, Detractors = 20%, NPS = +20
  SE = sqrt((0.40 * 0.60 + 0.20 * 0.80 + 2 * 0.40 * 0.20) / 300)
     = sqrt((0.24 + 0.16 + 0.16) / 300) = sqrt(0.001867) = 0.0432
  95% CI: +20 +/- 8.5 points -> range [+11.5, +28.5]

Minimum sample sizes for NPS:

Desired margin of error	Required responses
+/- 10 points	~100
+/- 5 points	~400
+/- 3 points	~1,100

NPS trend significance

To determine if an NPS change between two periods is statistically significant:

z = (NPS_2 - NPS_1) / sqrt(SE_1^2 + SE_2^2)

If |z| > 1.96, the change is significant at p < 0.05

Resolution time formulas

Percentile-based reporting

Always report resolution time as percentiles, not averages. Averages are skewed by outliers (one 30-day ticket destroys a weekly average).

Key percentiles:
  p50 (median):  Typical customer experience
  p75:           Start of the "long tail" - where most frustration lives
  p90:           Worst common experience
  p95:           Near-worst case, useful for SLA compliance

SQL:
  SELECT
    PERCENTILE_CONT(0.50) WITHIN GROUP (ORDER BY resolution_hours) AS p50,
    PERCENTILE_CONT(0.75) WITHIN GROUP (ORDER BY resolution_hours) AS p75,
    PERCENTILE_CONT(0.90) WITHIN GROUP (ORDER BY resolution_hours) AS p90,
    PERCENTILE_CONT(0.95) WITHIN GROUP (ORDER BY resolution_hours) AS p95
  FROM tickets
  WHERE resolved_at IS NOT NULL
    AND resolved_at >= NOW() - INTERVAL '28 days';

Business hours adjustment

Raw resolution time includes nights and weekends. For teams with defined business hours, calculate business-hours resolution time:

Business Hours Resolution = total_elapsed_time - non_business_hours_in_range

Non-business hours calculation:
  For each calendar day in the range:
    If weekday: subtract hours outside 9am-6pm (15h per day)
    If weekend: subtract full 24h
    If holiday: subtract full 24h

SLA compliance rate

SLA Compliance % = (tickets resolved within SLA target) / (total tickets) * 100

Report per priority tier:
  P1 SLA compliance: tickets_resolved_under_4h / total_p1_tickets * 100
  P2 SLA compliance: tickets_resolved_under_8h / total_p2_tickets * 100
  ...

Deflection rate formulas

Standard deflection rate

Deflection Rate = deflected / (deflected + not_deflected) * 100

Where:
  deflected     = self-service views with NO ticket created within 24h
  not_deflected = self-service views WITH ticket created within 24h

Content effectiveness score

For individual help articles or chatbot flows:

Article Effectiveness = 1 - (tickets_created_after_view / total_article_views)

Rank articles by:
  1. Total views (high traffic = high impact)
  2. Effectiveness score (low score = needs improvement)
  3. Impact = views * (1 - effectiveness) = potential tickets saved if improved

Deflection ROI

Monthly deflection savings =
  deflected_interactions * (avg_ticket_cost - avg_self_service_cost)

Annual ROI of deflection improvement =
  (new_deflection_rate - old_deflection_rate) * monthly_contacts * 12
  * (avg_ticket_cost - avg_self_service_cost)

Trend analysis formulas

Trailing average (smoothing)

4-week trailing average:
  avg_t = (value_t + value_t-1 + value_t-2 + value_t-3) / 4

Exponential moving average (more weight on recent data):
  EMA_t = alpha * value_t + (1 - alpha) * EMA_t-1
  Where alpha = 2 / (N + 1), N = number of periods

Week-over-week change detection

WoW change % = (this_week - last_week) / last_week * 100

Alert if:
  |WoW change| > 2 * standard_deviation_of_weekly_changes

This catches unusual spikes or drops relative to normal variation.

Seasonality decomposition

For weekly data with known seasonal patterns:
  seasonal_index_w = avg_value_for_week_w / overall_avg

Deseasonalized value:
  adjusted_t = actual_t / seasonal_index_for_that_week

Use deseasonalized values for trend detection to avoid false alarms from
predictable seasonal spikes (e.g., post-holiday ticket surges).

Sampling guidelines

When ticket volume is too high for 100% survey coverage:

Stratified sampling approach:
  1. Define strata: channel x priority x product_area
  2. Calculate required sample per stratum (min 30 for statistical validity)
  3. Sample proportionally or use minimum allocation per stratum

Simple random sampling minimum:
  For 95% confidence, +/- 5% margin on a population of:
    500 tickets/week:   217 surveys needed
    1,000 tickets/week: 278 surveys needed
    5,000 tickets/week: 357 surveys needed
    10,000+ tickets/week: 370 surveys needed (diminishing returns)

metrics-benchmarks.md

Support Metrics - Industry Benchmarks

Benchmarks give you a calibration point, not a target. Always establish your own baseline over 4-8 weeks before comparing externally. Match your peer group by vertical, company size, and support channel mix before drawing conclusions.

CSAT benchmarks

By industry vertical

Vertical	Median CSAT	Top quartile	Notes
SaaS / Software	78%	88%+	Higher for B2B than B2C due to relationship depth
E-commerce	80%	90%+	Driven primarily by shipping and returns experience
Financial services	75%	85%+	Slower resolution expected; regulatory complexity
Healthcare	72%	82%+	Complex issues, privacy constraints inflate effort
Telecommunications	68%	78%+	Historically lowest - billing complexity and churn
Hospitality / Travel	76%	86%+	High emotional stakes, strong seasonality effects

By support channel

Channel	Typical CSAT	Driver
Live chat	82-88%	Fast, convenient, in-context responses
Phone	78-85%	Personal but wait times hurt scores
Email / ticket	72-80%	Slow, lacks real-time back-and-forth
Social media	65-75%	Public complaints, often pre-escalated issues
Self-service (successful)	85-92%	Instant resolution; user feels empowered

By company stage

Stage	Typical CSAT	Context
Early startup (< 50 customers)	90%+	Founders answering tickets, very high-touch
Growth (50-500 customers)	78-85%	First hires learning; scaling pains visible
Scale (500-5,000 customers)	75-82%	Process-driven; specialization emerging
Enterprise (5,000+ customers)	72-80%	Tiered support; issues are more complex

Response rate benchmarks

Response rate	Interpretation
> 25%	Excellent - survey timing and delivery are healthy
15-25%	Good - standard for most SaaS support programs
10-15%	Acceptable - watch for response bias in low scorers
< 10%	Investigate - survey fatigue, wrong channel, or delivery issue

NPS benchmarks

By industry vertical

Vertical	Median NPS	Top quartile	World class
SaaS / Software	+30	+50	+70
E-commerce	+40	+60	+75
Financial services	+20	+40	+60
Healthcare	+15	+35	+55
Telecommunications	+5	+25	+45

NPS interpretation scale

NPS range	Interpretation
-100 to 0	Critical - systemic issues, high churn risk
0 to +20	Below average - significant improvement needed
+20 to +50	Good - competitive but meaningful room to improve
+50 to +70	Excellent - strong loyalty and word-of-mouth
+70 to +100	World class - rare, typically category leaders

Transactional NPS (tNPS) vs relationship NPS

Type	When to send	What it measures	Cadence
Relationship NPS	Periodic, no trigger event	Overall brand loyalty	Quarterly
Transactional NPS (tNPS)	After significant support interaction	Loyalty impact of specific resolution	Per qualifying interaction

tNPS scores run 10-15 points higher than relationship NPS for the same company because the survey primes the customer to think about a specific resolution. Do not compare them directly.

Resolution time benchmarks

First Response Time (FRT) by channel

Channel	Good	Average	Needs attention
Live chat	< 30 seconds	1-3 minutes	> 5 minutes
Phone (hold time)	< 60 seconds	2-5 minutes	> 10 minutes
Email / ticket	< 4 hours	4-12 hours	> 24 hours
Social media	< 1 hour	1-4 hours	> 8 hours

Average Resolution Time by complexity tier

Complexity	Good	Average	Needs attention
Simple (how-to, password reset)	< 2 hours	2-8 hours	> 24 hours
Medium (config issue, bug report)	< 8 hours	8-24 hours	> 48 hours
Complex (integration, data issue)	< 24 hours	1-3 days	> 5 days
Critical (outage, data loss)	< 4 hours	4-8 hours	> 12 hours

Efficiency metric benchmarks

Metric	Good	Average	Needs attention
First Contact Resolution (FCR)	> 70%	50-70%	< 50%
One-touch resolution rate	> 60%	40-60%	< 40%
Reopen rate	< 5%	5-10%	> 10%
Tickets per agent per day	8-15	15-25	> 25 (burnout risk)
Agent utilization	60-75%	75-85%	> 85% (no buffer for quality)

Deflection rate benchmarks

By self-service channel

Channel	Good	Average	Needs attention
Knowledge base / help center	30-45%	15-30%	< 15%
AI chatbot	25-40%	10-25%	< 10%
Community forum	15-25%	5-15%	< 5%
In-app contextual help	35-50%	20-35%	< 20%

Overall deflection by self-service maturity

Maturity level	Typical rate	What drives it
No self-service investment	5-10%	Incidental (users searching externally)
Basic help center	15-25%	Static articles for top issues
Optimized self-service	25-40%	Data-driven content, chatbot, in-app help
AI-augmented self-service	35-55%+	ML-powered search, generative answers

Warning: deflection rates above 50% often indicate the product has UX or reliability issues that self-service is papering over. Investigate with CSAT and reopen rate before celebrating a high deflection number.

Cost benchmarks

Cost per contact by channel

Channel	Low	Median	High
Phone	$6	$10	$16
Email / ticket	$3	$5	$8
Live chat	$2	$4	$6
AI chatbot (automated)	$0.10	$0.50	$1.00
Self-service (knowledge base)	$0.05	$0.25	$0.50

Figures reflect US/Western Europe SaaS companies. Costs vary significantly by geography, labor market, and tooling stack.

Support cost as a percentage of revenue

Company stage	Typical range	Context
Early stage	15-25%	Small team, high per-customer cost
Growth	8-15%	Scaling with process; some automation in place
Scale	5-10%	Mature processes; deflection working
Enterprise	3-7%	Tiered support; heavy self-service investment

How to use these benchmarks

Establish your own baseline first - measure for 4-8 weeks before comparing
Match your peer group - filter by vertical, company size, and channel mix
Set incremental targets - aim for 10-15% improvement per quarter, not an immediate jump to top quartile
Prioritize internal comparisons - channel vs channel, team vs team, and week vs week tells you more than external benchmarks do
Revisit annually - benchmarks shift as tooling and customer expectations evolve

Frequently Asked Questions

What is support-analytics?

How do I install support-analytics?

Run npx skills add AbsolutelySkilled/AbsolutelySkilled --skill support-analytics in your terminal. The skill will be immediately available in your AI coding agent.

What AI agents support support-analytics?

support-analytics works with claude-code, gemini-cli, openai-codex. Install it once and use it across any supported AI coding agent.

Is support-analytics free?

Yes, support-analytics is completely free and open source under the MIT license. Install it with a single command and start using it immediately.

What is the difference between support-analytics and similar tools?

support-analytics is an AI agent skill that teaches your coding agent specialized operations knowledge. Unlike standalone tools, it integrates directly into claude-code, gemini-cli, openai-codex and other AI agents.

Can I use support-analytics with Cursor or Windsurf?

support-analytics works with any AI coding agent that supports the skills protocol, including Claude Code, Cursor, Windsurf, GitHub Copilot, Gemini CLI, and 40+ more.

support-analytics

What is support-analytics?

Quick Start

support-analytics

Quick Facts

How to Install

Overview

Tags

Platforms

Related Skills

Frequently Asked Questions

What is support-analytics?

How do I install support-analytics?

What AI agents support support-analytics?

Maintainers

SKILL.md

Support Analytics

When to use this skill

Key principles

Core concepts

Satisfaction metrics

Operational metrics

Self-service metrics

Quality metrics

Common tasks

Set up a metrics framework - KPI hierarchy

Measure and improve CSAT - survey design and analysis

Implement NPS program - collection and segmentation

Track and optimize resolution time

Measure deflection rate - self-service effectiveness

Analyze support trends - topic clustering and forecasting

Build support dashboards - by audience

Gotchas

Anti-patterns

References

References

benchmarks.md

Support Analytics - Industry Benchmarks

CSAT benchmarks

By industry vertical

By support channel

By company stage

NPS benchmarks

By industry

NPS interpretation scale

Resolution time benchmarks

First Response Time (FRT) by channel

Average Resolution Time by complexity

Other efficiency benchmarks

Deflection rate benchmarks

By self-service channel

Overall deflection by company maturity

Cost benchmarks

Cost per ticket by channel

Support cost as percentage of revenue

How to use these benchmarks

metric-formulas.md

Support Metrics - Formula Reference

CSAT formulas

Basic CSAT score

Weighted CSAT (when response volumes differ across segments)

CSAT confidence interval

Response rate and bias

NPS formulas

Standard NPS

NPS margin of error

NPS trend significance

Resolution time formulas

Percentile-based reporting

Business hours adjustment

SLA compliance rate

Deflection rate formulas

Standard deflection rate

Content effectiveness score

Deflection ROI

Trend analysis formulas

Trailing average (smoothing)

Week-over-week change detection

Seasonality decomposition

Sampling guidelines