What you'll learn: How to map end-to-end customer journeys and identify where voice interactions create the most value.
Key takeaways:
- A customer journey is the complete sequence of interactions from first contact through resolution. Voice touchpoints are moments where phone interaction creates or destroys value.
- Map the real process, not the ideal one. Document how work actually happens today, including workarounds, tribal knowledge, and what can go wrong.
- Industry patterns differ. Ecommerce journeys center on order status and returns. Healthcare journeys center on scheduling and information. Financial services journeys center on account access and transactions.
- Journey mapping surfaces automation opportunities that isolated call analysis misses. A single call type may span multiple journey stages with different requirements.
- The goal shapes which journey stages matter most. Cost-focused programs prioritize high-volume stages. CX-focused programs prioritize high-friction stages. Revenue-focused programs prioritize conversion stages.
Nadia ran solutions engineering for a company that sold voice automation to enterprises. A logistics network had hired her team to find the right first use case. The network connects thousands of independent truck drivers with shipping carriers. Fifteen operators handled the phones, drowning in calls.
Nadia's team had done their homework. Certificate of Insurance verification accounted for 90% of inbound calls. Operators spent hours calling insurance providers, confirming coverage, and logging results. The pitch was straightforward. Automate COI verification. Save ten headcounts. The math was clean.
The VP of Product listened, then said something Nadia didn't expect. "That's a huge opportunity. It would save us millions in risk and labor. But that's not why you're here."
He explained that the company's Driver Tracking App was a value-added service they sold to shippers. Every driver who failed to install the app lost revenue. Drivers applied, got verified, and then stalled at the installation step. They didn't know how to install it. They got frustrated. They gave up. The company was losing subscription fees every hour a driver sat uninstalled.
The COI problem was real. But the revenue problem was urgent.
Nadia had mapped the wrong journey because she started from the wrong goal.
What a customer journey is
A customer journey is the complete lifecycle a customer moves through with your organization. Not just the moments when they pick up a phone. It includes every stage from first contact through ongoing service, spanning digital, in-person, and voice interactions.
Most organizations have three to five major journeys involving voice. The logistics company had three. Fleet Onboarding ran from contract through driver verification to go-live. Insurance and Compliance covered COI requests, collection, verification, and renewals. B2B Support handled inbound calls, ID verification, intent capture, and routing.
Each journey is an end-to-end process, not a phone call type. That distinction matters. A phone call is a moment inside a journey. Mapping the full journey reveals where that moment sits, what happens before and after it, and whether it's the right place to intervene.
This chapter teaches you how to map those journeys. The next chapter teaches you how to find hotspots within them and turn them into scored use cases.
The journey mapping process
Start broad. Don't think about automation yet. Just understand the flow.
Nadia's team sketched the Driver Onboarding lifecycle on a whiteboard. Four stages. Intake, where the driver applies. Verification, where documents get checked, and COI gets validated. Set up, where the driver installs the tracking app. Activation, where the driver is cleared to haul.
The volume was in Verification. Thousands of COI calls per month, each one following the same pattern. But because the VP's goal was revenue, not cost, Nadia looked at the journey differently. The value was in Setup, where drivers stalled, gave up, and took subscription revenue with them.
Mapping the full journey made that visible. If Nadia had started with call data alone, she would have seen the COI volume and stopped there.
For each stage in every journey, capture five things. What the customer does at this stage is the actions they take. What the business does behind the scenes is the process running in the background. Which systems are involved, including CRM, EHR, TMS, billing, and scheduling? Where the pain concentrates, meaning what slows down, breaks, or frustrates. And the volume and cost signals, meaning how many interactions happen per month and how long they take.
Industry journey examples
The journey framework applies across verticals, but the shape of the journeys differs.
In healthcare, a multi-location provider typically maps four journeys. Patient Acquisition from outreach through enrollment. Patient Onboarding covering ID verification, records collection, insurance, and first appointment. Ongoing Care spanning scheduling, check-ins, referrals, and billing. And Re-engagement for outreach, re-enrollment, and resumption of care plan. Most call volume lives in Ongoing Care, but the highest-friction voice touchpoints often sit in Onboarding.
In e-commerce, the journeys follow the purchase cycle. Pre-purchase with browsing, questions, and product comparison. Purchase with checkout, payment, and confirmation. Post-purchase with tracking, modifications, and returns. And Loyalty with rewards, repurchase, and referrals.
In financial services, journeys tend to be longer and more regulated. Account Acquisition from inquiry through approval and onboarding. Account Servicing covers balance checks, transfers, disputes, and address changes. Claims and Collections with filing, status, follow-ups, and payments. And Retention spanning renewal, complaints, offers, and win-back.
In marketplace businesses, you map both supply-side and demand-side journeys. A staffing marketplace maps Candidate Intake from application through screening and placement alongside Client Service covering job posting, matching, scheduling, and billing. Both sides generate voice interactions, often at different stages and with different urgency.
Equally important is deciding what to exclude. In healthcare, clinical care delivery, emergency triage, and care transitions involve too much clinical judgment for voice agents. In financial services, active fraud investigation and complex dispute resolution are poor fits. Name your exclusions early. Write them down. When someone asks, "Can the agent also handle X?" you have an answer ready.
Mapping the real process, not the ideal one
Map how things work today, not how the process diagram says they should work.
At the logistics company, the official COI verification process said operators call the insurance company, verify the policy, and log the result. The real process involved operators keeping personal spreadsheets of direct phone numbers for insurance adjusters, timing their calls to avoid peak hold times, and using a shared Slack channel to warn each other when a carrier's phone system was down.
This gap between official process and actual process shows up everywhere. For every journey stage, capture three layers.
How it's done today. Not the SOP, but the actual scripts, systems, and steps your operators follow. Sample 50 to 200 calls across your top call drivers. Tag what the caller is trying to do, what steps the agent takes, and where the process breaks.
Where humans improvise. The workarounds, personal notes, shortcuts, and tribal knowledge that make the process work despite the official documentation. If your best operators succeed because they know which insurance adjuster picks up fastest on Tuesday mornings, that knowledge needs to be captured and formalized before a voice agent can replicate the outcome.
What goes wrong? System outages, exceptions, edge cases, and the situations where even experienced operators aren't sure what to do. Internal documentation that conflicts. Policies that different teams interpret differently.
A concrete way to capture this is a current-state flow for each candidate use case.
For order status at an e-commerce company, the flow might look like this.
Step 1: The caller reaches the IVR and selects the order support option.
Step 2: The agent authenticates the caller using an email address or order number.
Step 3: The agent opens the order management system and locates the order.
Step 4: The agent reads the status, ETA, and tracking information.
Step 5: If the caller wants to modify the delivery, the agent checks whether the order has shipped.
Step 6: If not shipped, the agent makes the change. If shipped, the agent explains the limitation and offers alternatives.
Step 7: The agent confirms the resolution and ends the call.
That's the happy path. The real process includes exceptions. No order found because the caller used a different email. The order shows delivered, but the caller says it never arrived. The system is slow, and the agent fills the silence while screens load. The caller wants a refund, which requires a different queue. Each exception is a decision point your voice agent will need to handle or escalate.
Pull these flows from reality, not from process documents. The workarounds reveal the pain. Tribal knowledge reveals what the AI needs to know. The shortcuts reveal where the process is broken enough that even humans don't follow it.
Identifying voice touchpoints
With your journeys mapped, identify where voice interactions currently happen and where they should happen.
Voice touchpoints fall into three categories. Places where customers call today, which you can find in your IVR paths, queue data, and contact reason codes. Places where customers should be able to call but can't, such as after-hours gaps, channels that force web self-service when a conversation would be faster, or stages where customers abandon because they can't get help. And places where the business should be calling customers but isn't, including outbound opportunities like appointment reminders, onboarding follow-ups, payment collection, or the logistics company's driver installation calls.
For each touchpoint, note the entry channel (phone, IVR path, queue), the current handling team, the monthly volume, and the average handle time. This inventory becomes the raw material for the hotspot analysis in the next chapter.
Journey mapping by goal
Your primary goal from Chapter 3 shapes what you look for as you map.
If your goal is to cut costs, you're looking for the highest-volume, most repetitive stages. Where do operators spend time on tasks that follow the same script? Where are the handle times longest? Where does after-hours demand pile up unserved?
If your goal is fixing customer experience, you're looking for the stages that generate the most complaints, the longest wait times, and the most transfers. Where do customers give up? Where do they keep calling back repeatedly for the same issue?
If your goal is driving revenue, you're looking for conversion moments, renewal touchpoints, and collection opportunities. Where does revenue leak? Where do customers drop off before completing a high-value action?
Nadia learned this the hard way. If she had been mapping for cost, she would have zeroed in on Insurance and Compliance, where 90% of call volume lived. Because the VP was mapping for revenue, she zeroed in on Fleet Onboarding, specifically the Setup stage, where subscription revenue was leaking. Same company, same data, different map. The goal determines what you see.
Start wide. Map the end-to-end lifecycle. Include multiple journeys. Exclude what shouldn't be automated. Then let your goal guide you to the stages that matter most.