WFM guideQuality & CX

Contact centre CSAT improvement

Q: What are the biggest drivers of low CSAT in a contact centre?

The biggest drivers of low CSAT in a contact centre consistently across research and operational data are: (1) issue not resolved on the first contact — unresolved contacts score CSAT 25–40% lower than resolved ones, making FCR the single most predictive variable of CSAT outcome; (2) perceived agent effort — customers who feel the agent did not try hard enough or was dismissive give low CSAT even when the issue is technically resolved; (3) wait time before reaching an agent — long queue wait times lower the CSAT score even when the contact itself goes well, as customers arrive to the interaction already dissatisfied; (4) having to call back or contact multiple times — repeat contacts generate very low CSAT because customers feel the organisation failed them; (5) agent knowledge gaps — customers who feel the agent did not know what they were talking about score low even when the agent tried to help; (6) outcome disagreement — customers who were told something they did not want to hear (claim declined, policy not waived, unable to escalate) will give low CSAT regardless of how well the agent communicated the outcome, but strong empathy reduces the CSAT impact of a bad outcome.

A CSAT score tells you how many customers were satisfied. It does not tell you why the dissatisfied ones scored low, which contacts failed, or what would move the score. That requires driver analysis, root cause investigation, and a structured improvement programme — not just a better survey.

The AHT-CSAT conflict: the most common mistake in contact centre optimisation

AHT pressure vs. CSAT outcome — what the data shows

Scenario: Centre targets AHT reduction from 6.5min to 5.5min

Unit cost per contact: ↓ 12%
Contacts per agent hour: ↑ 14%
Agent coaching on speed: Active
FCR rate: ↓ 8pp
CSAT score: ↓ 6pp
Repeat contact rate: ↑ 15%
Total cost including repeats: ↑ 4%

AHT reduction programmes that do not control for FCR frequently produce this pattern: unit cost falls, total cost rises, CSAT falls. The mechanism is that agents shorten handle time by resolving contacts less thoroughly — rushing to avoid the AHT overage flag — which increases repeat contacts.

The rule is: AHT is a safe target when agents are spending time inefficiently (excessive hold, ACW, off-topic conversation). AHT is a dangerous target when agents are spending time on resolution — explaining, reassuring, documenting accurately. Measure FCR and CSAT alongside AHT before and after any AHT reduction initiative.

CSAT driver analysis: what moves the score

A CSAT driver analysis identifies which variables predict CSAT outcome for your specific contact centre. The drivers below are consistent across research — but their relative weighting varies by sector, customer profile, and contact type. Run the correlation analysis on your own data before committing to a prioritised improvement plan.

First contact resolution (FCR)

Impact: Very high

CC controllability

Mostly — depends on agent knowledge, authority, and system access

Lever

Agent training, authority levels, knowledge base quality

Measurement

FCR rate; repeat contact rate within 7 days by agent

Agent effort and empathy

Impact: High

CC controllability

Yes — agent behaviour is trainable and coachable

Lever

QA calibration, coaching, scripting for empathy moments, recognition

Measurement

Empathy score on QA forms; correlation with CSAT by agent

Pre-contact wait time

Impact: High

CC controllability

Partially — WFM can reduce wait time; demand cannot always be controlled

Lever

Staffing level, queue messaging, callback offers, self-service deflection

Measurement

ASA; abandonment rate; CSAT by wait time band

Repeat contacts / having to call back

Impact: Very high

CC controllability

Mostly — root cause analysis of repeat contacts drives FCR improvements

Lever

Root cause analysis of repeats; follow-up promise fulfilment; proactive outbound

Measurement

Repeat contact rate; promise fulfilment rate

Agent knowledge and confidence

Impact: High

CC controllability

Yes — knowledge base quality, training, and onboarding time

Lever

Knowledge base design, training programme, buddy support for new agents

Measurement

Knowledge assessment scores; transfer rate; escalation rate

Handling time (AHT) pressure

Impact: Ambiguous

CC controllability

Yes — target-setting and coaching choices are management decisions

Lever

Review AHT targets vs. FCR and CSAT data; avoid rewarding speed at cost of resolution

Measurement

CSAT by AHT band; FCR by AHT band

Outcome (was the customer given what they wanted?)

Impact: High

CC controllability

Partially — empathy delivery is trainable; outcome policy is upstream

Lever

Empathy coaching for difficult calls; agent authority to offer goodwill; escalation path

Measurement

CSAT by outcome type; empathy QA score on declined contacts

Designing the improvement programme: five steps

Segment CSAT to find where the low scores concentrate

Average CSAT scores hide the problem. Segment by contact type, team, channel, agent, time of day, and outcome. A centre averaging 76% CSAT may have specific contact types (complaint, escalation, complex product) running at 45% and high-volume simple contacts at 85%. The improvement programme targets the low-scoring segments, not the average.

Identify the driver(s) for each low-scoring segment

For each low-scoring contact segment, run a driver correlation: Is FCR lower? Is repeat contact rate higher? Is the average wait time before this contact type higher than average? Is QA empathy score lower on these contacts? Is agent knowledge less consistent? The driver analysis tells you what to fix — not the CSAT score itself.

Separate controllable from non-controllable drivers

Some CSAT drivers are outside the contact centre's control: a product that does not work well, a price that customers consider unfair, a policy that is inflexible. Document these and escalate to the relevant business owner — they are not contact centre improvement actions. Focus the programme on the drivers the contact centre can move: agent behaviour, knowledge, FCR, and wait time.

Design specific interventions for each controllable driver

FCR: identify root cause of repeats (knowledge gap, authority gap, system gap) and address at source. Empathy: targeted coaching using call recordings as examples; calibrate QA empathy scoring; recognise strong empathy examples. Wait time: WFM staffing adjustment or callback implementation. Knowledge: knowledge base update for the failing contact type; refresher training for agents with lower knowledge scores.

Measure, report, and iterate at the driver level — not just the score

Track FCR, empathy QA scores, repeat contact rate, and wait time by segment alongside CSAT. If the CSAT score is not moving but FCR has improved, the intervention is working but there may be another driver — investigate further. If both FCR and CSAT have moved, the programme is working — sustain and extend. Review at 4-week intervals minimum.

The FCR-CSAT relationship: the most powerful lever

First contact resolution is the single strongest predictor of CSAT across all contact centre research. A contact that is not resolved on the first attempt creates a repeat contact — which has a double negative impact: the customer is dissatisfied (reducing CSAT for that contact), and they call again (increasing costs and reducing CSAT for the repeat contact too).

25–40pp

CSAT gap between resolved and unresolved first contacts

30–45pp

CSAT gap between first-contact customers and those requiring a second contact

1 pp FCR

improvement is typically associated with 0.5–0.8pp improvement in CSAT score

FCR improvement targeting for CSAT: If the centre is running 65% FCR and targeting an 8pp CSAT improvement, FCR improvement is the most direct lever. A 15pp FCR improvement (from 65% to 80%) would typically be associated with a 7–12pp CSAT improvement — without changing empathy delivery, wait times, or any other variable. Run the FCR-CSAT correlation on your own data first; the relationship is strong but not identical across all contact types.

CSAT improvement questions

What are the biggest drivers of low CSAT in a contact centre?

The most consistent drivers of low CSAT across contact centre research: (1) Issue not resolved on first contact — unresolved contacts score 25–40pp lower than resolved ones, making FCR the strongest single predictor; (2) Perceived agent effort — customers who feel the agent did not try hard enough give low CSAT even when technically resolved; (3) Pre-contact wait time — long queue times lower CSAT before the contact even begins; (4) Having to call back — repeat contacts score 30–45pp lower than first contacts; (5) Agent knowledge gaps — customers who feel the agent did not know the answer score low; (6) AHT pressure reducing resolution quality — agents rushing give lower-quality resolutions and lower CSAT; (7) Bad outcome — customers who were told something they did not want still score low, but strong empathy reduces the CSAT impact of a bad outcome.