What Does “Enterprise-Ready” Mean for Speech Analytics?

109
Filed under - Archived Content

Earlier this year, Forrester Research identified CallMiner as a speech analytics solution with “most advanced, enterprise-ready speech analytics functionality”.

That’s a great description, but what does “enterprise-ready” really mean? After all, doesn’t every “enterprise” have their own set of objectives and requirements?

The answer is of course “yes”, but there are a significant range of features and benefits within speech analytics software solutions that should be considered as common or even baseline for virtually any organisation.

A core speech analytics value proposition is deriving insight from calls by converting unstructured contact centre conversations into structured intelligence.  That’s making big data actionable!  But there are several speech analytics solution components to consider in order to separate an enterprise-ready analytics application from what can amount to as an “IT science experiment”.

Following are five of these considerations:

1. API or Application Complete

Will a speech analytics solution deliver “out-of-the box” value or do you have to build out your solution via an Application Programming Interface (API)? Speech Analytics solutions that offer speech to text insight with a packaged user interface, standard reporting and established integration for audio capture accelerate results and define “Application Complete” with time to deploy speed and ease of use.

2. Time to Insight

Capable speech analytics solutions are far more than mere speech to text transcripts. Categorisation is applied to identify intent, effort and emotion from how callers and agents express themselves with voice.

Enterprise-ready speech analytics solutions offer packaged categorisation for what callers say (semantics) and even how they say it (acoustics) to jump-start awareness from every telephone call.

Automated scoring goes one step further by directing attention to negative hotspots and positive highlights.

3. Self-Service

Being able to easily tailor your speech analytics solution for your specific requirements is a key enterprise-ready “marker”. Accelerating how to understand what callers and agents are saying at scale with packaged categorisation is no doubt great. However, most organisations will want to go to the next level of customisation to automate insight from voice dialogs based upon your organisations products and interaction language.

Enterprise-ready speech applications make this easy with graphical user interfaces that encourage self-service customisation.

4. Security

Dialog insight must stand on a secure foundation of infrastructure, process and application-specific defence to be considered enterprise ready. The Amazon Web Services (AWS) enterprise hosting environment provides a secure baseline of safeguards with compliance measures at scale for a cloud-based speech analytics.

Beyond that, however, an enterprise-ready solution will display rigorous process and application compliance with security standards such as PCI DSS level 1.  Automated redaction of sensitive data within a speech analytics application must be in place for enterprise-ready security.

5. Multichannel

Organisations of all sizes are dealing with channel-of-choice communication demand across their customer base. Calls are certainly critical, but chat, email and more are being used, compounded with channel-hopping tendencies of maybe starting with chat and then following up with a call. Enterprise-ready speech analytics applications deal with channel-of-choice reality with analytics value beyond voice.

Consolidating customer interactions within a common analytics framework means enterprise-ready. Additional value such as automatically mapping the journey for customers who are engaging multiple times for the same issue even when they channel hop takes a speech analytics solution to a new level of enterprise-ready.

The above considerations are just a few examples to think about when deploying a speech analytics solution that’s ready for your enterprise.  Core speech-to-text technology is certainly important, but it does not stand alone as a value point.

How a capable speech analytics solution is packaged with features that support rapid deployment, immediate insight and ease of use contribute significantly to what will be considered an enterprise-ready speech analytics solution.

Following are some specific speech analytics solution components to look for beyond the previously mentioned to ensure you are on the enterprise-ready track for a speech analytics solution:

  • Application Scalability – Do not overlook the importance of how speech analytics can (or cannot) scale for your organisation beyond the basics of Software as a Service. For example, methods for data retention with indexing enable trend comparison over time with search and find speed. Presentation of that data is also critical without bogging down due to multiple simultaneous users.  For enterprise-ready confidence look for speech analytics solutions that deliver scalability tested and proven for large numbers of simultaneous users.
  • Speaker Separation – Make sure you can identify customer and agent speakers in categorised transcription. Additionally, look for categories that apply to agent behaviours such as ownership and empathy help provide insight into how employees encourage customer experience values.
  • Packaged Categorisation – Look for speech analytics solutions that help you get started with packaged categorisation. Bundled key words and phrases, acoustic measures, outliers (such as excessive silence) and search criteria along with scoring will get you jump-started.
  • Automated Scoring – Users can easily focus on interactions where attention is needed, or kudos is in order if automated scoring is available within an easy to use interface. Customisation with category tailoring and score weighting in a self-service model when available will allow tuning for specific requirements.
  • Organic Discovery – Look for a speech analytics solution that can bring attention to the unexpected. Visualisation that identifies trending topics spoken by customers or agents can be invaluable in driving to a root cause.
Author: Guest Author

Published On: 25th Oct 2018 - Last modified: 26th Feb 2019
Read more about - Archived Content

Follow Us on LinkedIn