The Customer Intelligence Playbook

Calculates the net sentiment across an entire dataset — the percentage of positive vs. negative vs. neutral feedback. This is your baseline. Every other sentiment analysis is a slice of this number.

Best for — Establishing a starting point. Tracking macro-level shifts over time.

“Across all our feedback channels, are customers getting happier or unhappier?”

Breaks down sentiment scores by feedback source — survey, app store review, support ticket, social media. Reveals which channels surface unhappy customers and which attract positive ones.

Best for — Understanding channel bias. A 4.2 CSAT on email and a 2.1 on Twitter are very different stories.

“Why does our NPS survey show +42 but our app store reviews are 3.1?”

Segments sentiment by themes extracted from the feedback. Customers might love your product but hate your billing. Topic-level sentiment reveals where positive and negative clusters form.

Best for — Pinpointing which aspects of the experience drive the overall score.

“Which features generate the most negative sentiment?”

Identifies the variables (time, channel, product line, customer segment) that most strongly influence sentiment shifts. Moves from “what is the sentiment” to “what causes it to change.”

Best for — Action planning. You can't fix sentiment — you fix the thing driving it.

“What changed between Q2 and Q3 that caused sentiment to drop 8 points?”

Tracks sentiment over time at granular intervals (daily, weekly, monthly). Identifies inflection points, seasonal patterns, and the impact of specific events (product launches, outages, PR crises).

Best for — Monitoring. Detecting early warning signals before they become trends.

“Did our pricing announcement last week move sentiment?”

Categorises each piece of feedback into predefined classes (Bug Report, Feature Request, Praise, Complaint, Question, Churn Signal). The taxonomy is governed — the AI assigns labels from your approved list, never invents new ones.

Best for — Triage. Routing feedback to the right team (engineering, product, support, marketing).

“What percentage of our support tickets are actually feature requests?”

Identifies the most frequent and salient terms across the dataset. Goes beyond word frequency — the algorithm weights terms by uniqueness and context, surfacing what's distinctive rather than just what's common.

Best for — Quick orientation. Understanding the vocabulary customers use to describe your product.

“What words do customers use most when describing our checkout process?”

Discovers latent themes across the dataset without predefined categories. The algorithm groups semantically related feedback into topics you didn't know existed. Different from classification: classification assigns known labels; topic modeling discovers new ones.

Best for — Exploratory analysis. Finding emergent issues you weren't looking for.

“What are customers talking about that we haven't asked about?”

Extracts, groups, and ranks feature requests based on frequency, sentiment intensity, and customer segment value. Output: a prioritised backlog sourced directly from customer voice.

Best for — Product roadmap input. Backlog grooming with evidence, not anecdotes.

“Which feature requests come from our highest-value customers?”

Applies the JTBD framework to identify the underlying “job” the customer is hiring your product to do. Moves beyond feature requests to understand the motivation and outcome the customer seeks.

Best for — Strategic product direction. Understanding the “why” behind the “what.”

“What job are customers trying to accomplish when they use our reporting tool?”

Extracts explicit and implicit needs from feedback. Explicit: “I need export to Excel.” Implicit: “I copy-paste the data into spreadsheets” (implied need: native export). The AI surfaces both.

Best for — Comprehensive needs discovery. Complementary to JTBD.

“What do customers need that they haven't directly asked for?”

Identifies specific points in the customer journey where customers experience difficulty, confusion, or frustration. Maps each friction point to the stage and touchpoint where it occurs.

Best for — Journey optimisation. Prioritising fixes by customer impact.

“Where in the onboarding flow do customers get stuck?”

Surfaces pain points that customers describe indirectly — through tone changes, repeated workarounds, comparison with competitors, or expressed resignation (“I've given up on…”). Catches issues that direct complaint analysis misses.

Best for — Uncovering hidden issues. Finding the “silent churn” signals.

“What problems are customers dealing with that they've stopped complaining about?”

Traces an identified issue (e.g., “checkout abandonment is up 15%”) through the feedback to identify the underlying cause. Correlates multiple signals — sentiment shifts, keyword frequency changes, new topic emergence — to pinpoint the origin.

Best for — Post-incident analysis. Explaining the “why” behind a metric change.

“Why did our CSAT drop 0.4 points in the last two weeks?”

Compares key metrics (CSAT, sentiment, resolution rate) across customer service and engagement channels — email, phone, chat, social media, in-app. Identifies which channels perform well and which drag the average down.

Best for — Channel investment decisions. “Should we add more chat agents or improve self-service?”

“Which support channel has the lowest customer effort?”

Identifies critical moments in the customer journey — the touchpoints that disproportionately influence overall satisfaction and loyalty. These are the make-or-break interactions where extra investment pays off.

Best for — Prioritisation. Focusing experience improvement budget on high-leverage moments.

“Which three interactions determine whether a customer becomes a promoter?”

Scores each touchpoint on effectiveness — does it accomplish what it's supposed to? Measures both customer experience quality and functional outcome success at each touchpoint.

Best for — Operational assessment. “Is this touchpoint doing its job?”

“How effective is our post-purchase email sequence at driving repeat purchases?”

Pinpoints where customers abandon a journey — onboarding flows, checkout processes, feature adoption sequences. Correlates feedback themes with behavioural data to explain why.

Best for — Funnel optimisation. Reducing churn at critical stages.

“At what step in onboarding do we lose 40% of new customers, and why?”

Analyses the customer journey stage-by-stage, aggregating sentiment, pain points, and needs for each stage. Produces a health score per stage — a journey heatmap.

Best for — Strategic overview. Knowing which journey stages need attention.

“Which stage of the customer lifecycle has the worst experience?”

Measures how easy it is for customers to accomplish a task. CES is the strongest predictor of future behaviour — customers who work hard to achieve an outcome are more likely to churn. InsightNarrator scores CES and analyses the verbatims to identify what made the experience hard.

Best for — Service interactions. Transactional feedback. Lower is better.

“What makes our support interactions difficult for customers?”

The workhorse of transactional measurement. CSAT captures satisfaction with a specific interaction. InsightNarrator goes beyond the score: it categorises the verbatims by theme, so you see not just that satisfaction is 82%, but that the 18% who rated poorly all mention “slow response.”

Best for — Post-interaction measurement. Touchpoint-level feedback.

“What drives low CSAT scores in our billing interactions?”

The relationship metric. NPS measures likelihood to recommend — a proxy for loyalty. InsightNarrator calculates the score, then performs deep analysis on the verbatims: promoter themes, detractor themes, passive themes, and what would move each segment.

Best for — Relationship measurement. Benchmarking. Strategic CX health.

“What would turn our passives into promoters?”

Identifies customers or segments showing churn signals in their feedback — declining sentiment, increased complaints, mentions of competitors, disengagement language. Flags at-risk accounts before they cancel.

Best for — Proactive retention. Account management prioritisation.

“Which enterprise accounts are showing early churn signals?”

Estimates CLV based on feedback patterns correlated with retention behaviour. Customers who provide constructive feedback, engage with surveys, and express loyalty intent tend to have higher lifetime value. Links qualitative data to financial outcomes.

Best for — Segmenting customers by value. Prioritising retention investment.

“Which customer segments have the highest predicted lifetime value?”

Identifies the specific factors that drive customer loyalty — not just satisfaction, but the kind of commitment that translates into advocacy and retention. Isolates the variables that matter most from those that are table stakes.

Best for — Loyalty programme design. Identifying what to invest in for maximum retention impact.

“What makes customers recommend us to colleagues?”

Groups customers into behavioural segments based on usage patterns, feedback themes, and demographic data. Reveals natural clusters — power users, at-risk users, dormant users, growing users — each with different needs and value.

Best for — Targeted CX programmes. Tailoring experiences by segment.

“What are our natural customer segments, and how do their needs differ?”

Correlates feature adoption with satisfaction, sentiment, and retention. Identifies which features are “sticky” (drive retention), which are “delight” (drive satisfaction), and which are “hygiene” (expected but don't drive either).

Best for — Product strategy. Feature investment and sunset decisions.

“Which features correlate most strongly with customer retention?”

Traces the paths customers take through your product — where they start, what they do, where they exit. Combined with feedback, reveals not just the flow but the friction points within it.

Best for — UX optimisation. Onboarding flow design.

“What's the most common path from signup to first value, and where does it break?”

Measures activation — the moment a customer reaches the “aha” moment where they experience core product value. Analyses what accelerates activation, what blocks it, and how activated customers differ from non-activated ones in feedback and behaviour.

Best for — Onboarding optimisation. Time-to-value reduction.

“What actions predict whether a new customer will activate within 7 days?”

Computes a composite engagement score based on interaction frequency, feature breadth, feedback participation, and support interaction patterns. Customers with higher engagement scores have higher retention and CLV.

Best for — Account health monitoring. Identifying disengagement before churn.

“Which accounts have seen engagement scores drop more than 20% this quarter?”

Analyses how customers and the market describe your brand — the associations, emotions, and attributes they attach to your name. Reveals whether your intended brand positioning matches actual perception.

Best for — Brand strategy. Positioning audits. Reputation management.

“When customers mention us unprompted, what words do they use?”

Measures your share of the total customer conversation in your category — across reviews, social media, forums, and community discussions. Tracks whether your share is growing or shrinking relative to competitors.

Best for — Market position tracking. Campaign effectiveness measurement.

“What percentage of the conversation in our category mentions us vs. competitors?”

Analyses feedback from customers who chose you (wins) and those who chose a competitor (losses). Extracts the decision factors, objections, and differentiators that shaped their choice.

Best for — Sales enablement. Competitive positioning. Product gaps that lose deals.

“Why do customers who evaluate us and a competitor choose the competitor?”

Compares your sentiment scores against up to five named competitors across the same review sources and time period. Reveals where you lead and where you trail — not on your own survey, but on public perception.

Best for — Benchmarking. Identifying competitive advantages and vulnerabilities.

“How does our review sentiment compare to our three main competitors?”

Builds a feature comparison matrix based on what customers mention — not your marketing claims, but what customers actually value and mention in feedback. Shows where competitors are perceived as stronger or weaker.

Best for — Competitive product strategy. Identifying feature gaps that matter to customers.

“Which features do customers mention most often when comparing us to Competitor X?”

Measures and analyses first contact resolution rate — the percentage of customer issues resolved in a single interaction. Correlates FCR with satisfaction, effort, and churn to quantify its business impact.

Best for — Support team performance. Process improvement.

“What types of issues have the lowest first contact resolution rate?”

Identifies stages in your service delivery process where feedback indicates delays, confusion, or excessive effort. Maps bottlenecks to specific process steps and quantifies their customer impact.

Best for — Process optimisation. Reducing cycle time.

“Where in our support workflow do customers experience the longest delays?”

Analyses the relationship between response times and customer outcomes — satisfaction, sentiment, escalation rates. Identifies the response time thresholds where satisfaction drops sharply.

Best for — SLA optimisation. Staffing and scheduling decisions.

“At what response time does CSAT drop below 80%?”

Measures the rate at which the organisation detects and acts on customer issues — before the customer has to complain. Proactive detection vs. reactive response. Combines feedback monitoring with issue resolution tracking.

Best for — Proactive CX management. Early warning systems.

“What percentage of our issues are detected by us vs. reported by customers?”

Tracks quality metrics over time — defect rates, error mentions, complaint recurrence — to measure whether process changes actually improve the customer experience. Closes the loop between improvement initiatives and customer outcomes.

Best for — Continuous improvement programmes. Demonstrating CX ROI.

“Did our new QA process actually reduce complaint rates?”