Introduction

In Voice of the Customer (VoC) analysis using Speech Analytics tools, the quality of transcription and sentiment analysis directly depends on the content of the recorded call. A common question is whether it is possible to automatically skip the hold time during this processing.

Below, we explain why this information is included in the analysis and how you can optimize your recordings to obtain more accurate results.

Why is "Hold" included in the analysis?

After a detailed technical review, we confirm that it is not feasible to omit the hold time within VoC analysis.

This is because the Speech Analytics engine processes the call's audio file as a whole. Any sound, message, or music recorded during the hold time is part of the technical communication and is captured by transcription and data analysis tools.

The Impact of Verbal Hold Messages

If your hold music includes advertisements, informational announcements, or pre-recorded voices, the Speech Analytics AI will attempt to:

1. Transcribe those messages as if they were part of the conversation between the agent and the customer.
2. Analyze the sentiment of those announcements, which can skew the actual results of the interaction.
3. Count keywords that do not belong to the case resolution, affecting the accuracy of the reports.

Optimization Recommendations

To ensure that your VoC metrics accurately reflect the customer experience and agent management, we suggest applying the following recommendations:

1. Use of Instrumental Music (Track Type): If it is essential to maintain hold time, set up background music that is exclusively instrumental. Avoid tracks that contain voices, song lyrics, or commercial announcements. Since it is only music, the analysis tool will detect the absence of human voice and will not generate unnecessary transcriptions in that segment.
2. Reduce Hold Times: The best practice for excellent customer experience (CX) and clean data analysis is to minimize hold times. Training agents in the agile use of internal tools helps calls flow without interruptions that need to be analyzed.
3. Avoid Repetitive Verbal Messages: If your telephony platform allows messages while the customer waits, ensure they do not interfere with the quality of the audio captured for subsequent analysis.

Conclusion

VoC analysis is a powerful tool for understanding your customers. By properly configuring your hold environments with tracks that contain no verbal content, you ensure that the data obtained is 100% relevant, eliminating the informational "noise" generated by hold messages.