& Digital experience
Monitoring the customer journey from IVR to agent applications
When someone calls a bank, an insurer, a healthcare service, an electricity provider, or after-sales support, several systems must work together.
The caller should be able to navigate the telephone menu, reach the intended service, and be heard clearly. The representative needs access to the expected information and responsive applications to handle the caller's request.
Monitoring this journey brings together five practical questions:
IVR navigation
Can the caller navigate the required interactive voice response (IVR) path?
Call routing
Does the call reach the intended destination?
Voice quality
Is voice quality acceptable in both directions?
CTI integration
Does computer-telephony integration (CTI) display the expected customer information?
Agent application performance
Can the representative complete the supporting application steps within acceptable response times?
An IVR may work correctly and a CRM application may be available, yet the handoff between them may fail.
These checks connect the caller's experience with the representative's ability to assist. An IVR may work correctly and a customer relationship management (CRM) application may be available, yet the handoff between them may fail.
Coordinated synthetic scenarios test those interactions, checking expected results and response times. The findings provide evidence of the user experience on both sides of the phone call.
Depending on the user-experience results, infrastructure and application monitoring can help teams investigate further when necessary.
The IVR journey before the agent answers
IVR monitoring tests the availability, functionality, and response times of defined paths through an automated telephone service. A call being answered is the first observable result; subsequent interactions determine whether the caller reaches the service they need.
The journey might require the caller to choose a language, select a department, enter a reference, and request assistance. Each step has an expected result. A missing prompt, an unrecognised selection, or an unexpected return to the main menu can interrupt progression towards the intended destination.
Automated phone calls reproduce these interactions using keypad tones or synthesised speech where appropriate.
Measurements show whether the journey completes and how long each step takes. Recordings of what the monitoring system heard provide evidence when results differ from expectations.
Roadside assistance
For an insurer, the monitored IVR path might lead the caller to roadside assistance.
Appointment booking
For a healthcare appointment line, the monitored path might lead to the booking team.
Each synthetic IVR scenario reflects the opening hours, prompts, and routing expected for that service, checking whether callers can progress towards their intended destination.
& VoIP monitoring
A connected call can still be difficult to understand
A successful connection does not establish that a conversation will be usable.
On a phone call conducted via voice over internet protocol (VoIP), audio can be distorted, interrupted, or difficult to understand even when the call remains connected.
Voice quality monitoring assesses this experience through controlled audio exchanges. A monitoring system sends a known message and records what arrives.
Perceptual speech-quality algorithms compare the received speech with a reference recording to estimate listening quality, expressed as a MOS-based score (Mean Opinion Score).
Representative to caller
Assess the listening quality of the audio received by the caller.
Caller to representative
Assess whether the caller's voice reaches the representative clearly, without distortion or interruption.
Both directions of the exchange are important. The caller may hear the representative clearly while his or her own voice arrives distorted. Bidirectional testing assesses the quality received on each side.
Call setup time
Measures the time required to establish the call.
Speech-quality measurements
Assess the quality of received speech, including MOS-based estimates.
Audio interruptions
Identify interruptions affecting the continuity of the conversation.
Dropped calls
Provide evidence about the stability of the connection.
Call setup time, speech-quality measurements, interruptions, and dropped calls provide additional evidence about connection performance, listening quality, and stability.
Alongside a clear voice connection, the information associated with the phone call must appear on the customer service representative's screen. The link between telephone activity and that information is another part of the journey that monitoring can assess.
Alongside a clear voice connection, the agent needs access to the information required to handle the enquiry. Where an incoming call is configured to open a customer record automatically, monitoring can check whether a record opens, whether it is the expected one, and how quickly it appears.
CTI monitoring: where the call meets the customer record
Computer-telephony integration (CTI) links telephone activity with applications used by representatives.
Depending on the implementation, an incoming call can trigger the display of a customer record or other relevant information, often called a screen pop.
This handoff connects the caller's journey to the representative's application experience. The call may reach its intended destination while the expected record fails to appear, appears late, or displays information that does not match the test scenario.
A call reaching the right representative does not guarantee that the expected customer record appears correctly and on time.
Missing record
The expected customer record does not appear.
Delayed screen pop
The customer record appears later than expected.
Incorrect information
The displayed information does not match the test scenario.
For example, in a bank-support scenario, a test call follows a defined IVR route and supplies an agreed test identifier. When it reaches a dedicated representative account, CTI monitoring checks whether the corresponding test record appears; it also measures the delay from a defined call event.
The same principle applies to an insurance policy record, an electricity customer's account, or after-sales service.
Opening the CRM independently and searching for the record tests the application journey on its own. In contrast, a coordinated voice and application scenario tests the convergence between the two: whether the phone call itself triggers the expected response.
Monitoring the voice connection and the CTI handoff helps verify not only that a call reaches its destination, but also that the representative receives the expected information to handle the enquiry.
applications
CRM and business applications: the agent's side of the call
Once a representative answers, the caller's experience also depends on software they cannot see.
Opening a record, retrieving information, or saving an update can keep both people waiting, even while the telephone connection remains clear.
In a bank, the representative may use a CRM such as Salesforce to consult account or transaction information supplied by other banking systems, within the same interface or through another application.
Similar dependencies exist in healthcare scheduling, electricity account and billing services, and after-sales support.
CRM monitoring and monitoring of other business applications make these steps observable.
Synthetic scenarios reproduce defined actions, check expected results, and measure response times across browser-based applications, desktop clients, or virtual desktop environments.
Browser-based applications
Reproduce defined actions and verify expected results in web-based business applications.
Desktop clients
Monitor defined actions and response times in desktop business applications.
Virtual desktop environments
Check the expected behaviour and response times of applications accessed through virtual desktops.
Results for individual steps help distinguish disruptions. A customer record that appears late following a call reveals a delay in the monitored handoff; a record that appears promptly but responds slowly to subsequent actions reveals a problem at a later stage.
Further telephony, network, or application evidence may be needed to establish the underlying cause.
XConnect: investigating supporting infrastructure
ITRS's XConnect case study illustrates this complementary perspective: infrastructure monitoring across a VoIP company's routers, hardware, and network servers helps teams investigate the systems supporting its services.
What monitoring coverage can reveal
The paths and environments included in monitoring shape what its results can tell service teams. Useful perspectives include:
Different routes to assistance
Daytime enquiries and overnight assistance may follow different IVR paths, with their own prompts, destinations, and expected results.
Differences between environments
Checks from selected networks or locations, and through representative application environments, can reveal variations in availability or response times.
Delays at particular stages
Separate timings for an IVR response, a customer-record display, and a subsequent application action help identify where waiting occurs.
Changes over time
Repeated scenarios can show whether monitored journeys continue to work as expected after an IVR update, a CRM release, or a migration to a new VoIP service.
Each result relates to the journey and conditions tested. Monitoring coverage determines which parts of the experience can be observed.
Each result relates to the journey and conditions tested. Infrastructure and application monitoring can provide complementary evidence to investigate the systems supporting an affected stage.
Coordinated monitoring of CRM and business applications helps identify where delays occur in the representative's journey, while complementary infrastructure and application evidence supports further investigation.
User experience
The technical journey behind human assistance
For the caller, navigating the IVR, reaching a representative, and receiving assistance form one continuous experience.
The representative depends on that same connection, alongside the information and applications needed to handle the enquiry. A disruption on either side can interrupt the conversation.
Navigate the IVR
The caller follows the expected telephone menu and reaches the intended service.
Reach a representative
The telephone connection enables the caller and representative to communicate.
Receive assistance
The representative uses the required information and applications to handle the enquiry.
User experience monitoring makes these connected interactions observable: whether the expected steps complete, what information appears, and how long each stage takes.
Alongside infrastructure and application monitoring, that evidence helps teams investigate how technical behaviour affects the service delivered through the helpline.
Monitoring connects the caller's journey with the representative's application experience, making technical disruptions and response times observable.