You know your overall NPS and your churn rate, but not where in the experience things go wrong. Survey scores sit with the insights team, tickets with support and usage logs with product, so nobody sees the whole path. Customer journey analytics joins those sources stage by stage, so you can see which moments cost you loyalty and revenue. It is one part of the wider practice covered in our guide to customer experience analytics.
Key takeaways
- Customer journey analytics measures satisfaction and behavior at each stage, such as purchase, onboarding, use, support and renewal.
- Assign every data source to a stage: touchpoint surveys, support contacts, usage and transactions, and operational times.
- At each stage, measure satisfaction and drop-off, then compare segments.
- The moments that matter are the stages whose experience best predicts overall loyalty or churn, not simply the stages with the lowest scores.
- If you cannot link customers across sources, aggregate by stage and month.
What is customer journey analytics?
Customer journey analytics measures what customers experience and do at each stage of their relationship with you, from awareness to renewal. Its job is to find where customers struggle and to size the problem: how many drop off, and how much loyalty or revenue the friction costs.
It is easy to confuse with two neighbors.
| Journey mapping | Web and app analytics | Customer journey analytics | |
|---|---|---|---|
| What it produces | A qualitative picture of stages, touchpoints and emotions | Clicks, sessions and conversion paths | Satisfaction, drop-off and loyalty measured stage by stage |
| Main data | Workshops, interviews, observation | Behavioral clickstream | Touchpoint surveys plus behavioral and operational data |
| Answers | What the journey looks like | What people did on the site or app | Where the journey breaks and what it costs |
A journey map gives you the stage names, and web analytics is one input. Journey analytics puts numbers behind both, including stages that never touch a screen, such as a delivery or a phone call.
What data do you need at each journey stage?
Most companies already hold this data in separate systems. Typical sources by stage:
| Stage | Touchpoint surveys | Support contacts | Usage and transactions | Operational times |
|---|---|---|---|---|
| Awareness and consideration | Brand tracker, post-visit survey | Pre-sales questions | Web and app visits, trial sign-ups | Response time to inquiries |
| Purchase | Post-purchase survey | Order and payment tickets | Orders, abandoned checkouts | Order-to-delivery time |
| Onboarding | Post-setup survey | Setup and how-to tickets | Setup completion, first use | Time from sign-up to first use |
| Use | Relationship survey, reviews | Product and fault tickets | Active use, repeat purchase | Downtime, time to fix |
| Support | Post-contact survey | Ticket themes, repeat contacts | Contacts per customer | Time to resolve |
| Renewal | Pre-renewal relationship survey | Cancellation contacts | Renewals, downgrades, cancellations | Time to process changes |
Your CRM adds segments such as plan, region and tenure. For how each score is defined, see our guide to customer experience metrics.
How do you do customer journey analysis?
- Define the stages. Use five to seven stages your teams already recognize, named from the customer’s side (“getting set up”, not “provisioning”).
- Assign each data source to a stage. Every survey, ticket category, usage event and operational timer belongs to one stage.
- Measure satisfaction and drop-off at each stage. Satisfaction comes from the touchpoint survey. Drop-off is the number of customers who reach a stage but not the next one.
- Compare segments. Split each stage by plan, channel, region or tenure. A stage that looks fine overall may be failing one group.
- Find the moments that matter. For each stage, compare the overall NPS, or the churn rate, of customers who rated that stage low with those who rated it high. The stages with the widest gaps are the ones whose experience predicts loyalty.
Step 5 is where customer journey analysis earns its place: a low stage score is not a priority if it does not change how customers feel about you overall. To move from loyalty to actual losses, customer churn analysis shows how to find who leaves and why.
What does a worked example look like?
The numbers below are illustrative, not from a real company. A subscription business surveys 1,000 customers once a year. Each rates purchase, onboarding and everyday use on a 1 to 5 satisfaction scale, and the 400 who contacted support in the past six months also rate support. Everyone answers the 0 to 10 recommend question. “Low” means a rating of 1 to 3; “high” means 4 or 5.
| Stage | Base | Rated high | NPS, rated low | NPS, rated high | Gap (points) |
|---|---|---|---|---|---|
| Purchase | 1,000 | 82% | −5 | +29 | 34 |
| Onboarding | 1,000 | 61% | −22 | +52 | 74 |
| Everyday use | 1,000 | 74% | −20 | +38 | 58 |
| Support | 400 | 70% | −30 | +30 | 60 |
NPS is the percentage of promoters (9 or 10) minus the percentage of detractors (0 to 6). For onboarding, 390 customers rated it low: 78 promoters (20%) and 164 detractors (42%), so their NPS is 20 − 42 = −22. The 610 who rated it high include 378 promoters (62%) and 61 detractors (10%), so 62 − 10 = +52. Together that is 456 promoters and 225 detractors out of 1,000, which gives the overall NPS of 45.6 − 22.5 = +23.
The onboarding gap is 52 − (−22) = 74 points, the widest of the four stages. Onboarding also has the most low raters, 39% against 18% for purchase and 26% for everyday use.
Behavioral data tells the same story for a cohort of 1,200 customers who signed up in one quarter.
| Step | Customers | Lost | Drop-off |
|---|---|---|---|
| Signed up | 1,200 | ||
| Completed setup | 1,020 | 180 | 180 ÷ 1,200 = 15.0% |
| Used it weekly in month two | 870 | 150 | 150 ÷ 1,020 = 14.7% |
| Renewed after the first term | 810 | 60 | 60 ÷ 870 = 6.9% |
The largest single loss, 180 customers, happens before setup is finished. If a customer who completes setup is worth $600 a year (also illustrative), that is up to 180 × $600 = $108,000 a year of revenue that never starts. Support has a wide gap too, but onboarding combines the widest loyalty gap, the most unhappy customers and the biggest drop-off. Look there first, starting with setup tickets and post-setup comments.
How do you run journey analytics when you cannot link customers?
Often the survey is anonymous or your systems share no customer ID, so you cannot follow one person from survey to renewal. You can still aggregate each source by stage and month and lay the series side by side.
| Month (illustrative) | Onboarding satisfaction (% rating 4–5) | Setup completion | Setup tickets per 100 sign-ups |
|---|---|---|---|
| January | 64% | 86% | 12 |
| February | 63% | 85% | 13 |
| March | 55% | 79% | 21 |
| April | 58% | 81% | 18 |
All three measures got worse together in March. That does not prove cause, but something changed in onboarding that month, such as a release or a new process. Collect at least a year of months before trusting a pattern, and remember that month-level movement describes the group, not individual customers.
What are the common mistakes?
- Treating the journey map as the analysis. A map names the stages but does not show which one costs you customers.
- Measuring only digital stages. Clickstream misses delivery, store and phone experiences.
- Ranking stages by score alone. The lowest-scoring stage may not drive loyalty. Rank by the gap in NPS or churn between low and high raters, together with how many customers rate the stage low.
- Ignoring who is missing. Customers who left during onboarding never answer the annual survey, so later stages look better than the journey really is. Pair survey scores with drop-off counts.
- Reading small segments as findings. A gap among 40 customers can move several points by chance. Check the base before you act.
How do you put customer journey analytics to work?
The hard part is getting survey scores, tickets, usage and operational times into one place each quarter. You can import or connect survey, customer and operational data into one governed layer instead of rebuilding that join by hand. From there, see what drives satisfaction and loyalty by segment and journey and act on issues before they become churn. Once you know the moment that matters, a voice of customer analysis of comments at that stage shows what to change.
For your next step, pick one journey and one outcome, such as first-year renewal. List five to seven stages and the data you hold for each, then run the low-versus-high comparison on your latest relationship survey.