Skip to content

Scouting Report: Speech Analytics Enters Its Next Act — Maturity

By Donna Fluss

View this document on the publisher’s website.

The speech analytics market continues on its remarkable journey as it matures, enters middle age, and confronts a variety of new challenges. The primary issues are these: Speech analytics is not yet considered a “must-have” application; analytics-enabled quality assurance (AQA) has not caught on; real-time speech analytics has a limited number of use cases; and text analytics continues to struggle to be noticed.

Each of these market challenges represents substantial opportunity for vendors and end users. Companies that are serious about using speech analytics and treat it as an enterprise-level analytics solution, support it with highly trained business intelligence analysts, and build a change management program to apply its findings are realizing great benefits.

Unfortunately, there are also many hundreds of organizations that are not yet doing so. This typically happens when a company uses a basic and limited speech analytics application that performs only keyword searches, or a company has not made the investment required to build a true enterprise-level speech analytics program to apply findings on a timely basis.

Speech analytics has reached maturity after 14 years in the commercial market. Currently, there are two types of speech analytics solutions. The first group consists of a small number of feature-rich solutions. These sophisticated applications are evolving into business intelligence platforms that use speech- and text-based findings to provide enterprise-level data. The second group of solutions is much larger and includes many that have more rudimentary functionality. They concentrate on identifying keywords and even some key phrases and might come with charts or dashboards to help users find some basic word and phrase patterns. DMG expects this high-value IT sector to continue to evolve and transform into increasingly valuable solutions.


The upside for speech analytics remains great. While adoption has been strong, the addressable market is still large, and there are thousands of contact centers that have not yet adopted speech analytics. Given the maturity of the market, it’s time for a replacement cycle; there is great opportunity for speech analytics vendors to replace existing solutions, particularly now that the current generation of solutions is available in the cloud and addresses both speech- and text-based interactions.

The technical innovation that is coming from leading providers of speech analytics is compelling, and there are substantial differences between current and older applications. From the market’s beginning, these solutions held great potential for companies, but their contributions will increase as their capabilities are incorporated into other applications. This is absolutely the case for customer journey analytics. Speech and text analytics are necessary to understand many aspects of the customer journey and are a great way to get started in building a voice-of-the-customer program. Interaction analytics, another name for speech and text analytics, does not address all touch points in the customer journey, but among its many uses, this tool can provide context and great insights about customer needs and wants through many communications channels.

Speech analytics technology, specifically keyword search and transcription, is being incorporated into a variety of third-party applications, creating a very large opportunity for the engine providers. It is common to find speech analytics as a fully integrated module of sales solutions and marketing automation platforms. Speech analytics engines are a standard component of e-discovery tools and are finding their way into enterprise business intelligence solutions.

A few speech analytics vendors have created packages to address specific business opportunities like customer churn identification, improving sales/collections, first-contact resolution, and other targeted uses. This is a step in the right direction, but such targeted uses are primarily intended to give users a lexicon/library of search terms and phrases to help them get started. The next step is for speech analytics to enter the world of data science and create algorithms to identify a variety of enterprise issues. DMG expects to see these types of capabilities emerge, although they won’t necessarily come from the speech analytics vendors. The point is that there is a lot more that can be done with speech and text analytics, and this sector presents substantial opportunity for both vendors and end users.[vc_column_text css=”.vc_custom_1514932490568{background-color: #e6e6e6 !important;}” el_class=”free-research-subscribe”]


Analytics-enabled quality assurance solutions hold great promise for enterprises, agents, and customers. AQA solutions automate contact center quality assurance/quality management (QA/QM), a process that has not changed substantially in the past 30 years. QA is necessary for many reasons. Enterprises invest a significant amount of resources to monitor and evaluate how well their contact center agents handle interactions and adhere to internal policies and procedures. It’s essential to monitor what agents are doing and for them to know that there is always a chance that management is watching. In addition, with many contact centers, QA is the only way to gain an understanding of why people are contacting them. And QA is often the most effective method for providing timely coaching to agents to improve service quality and increase agent satisfaction and retention.

The current QA process is people-intensive, which is expensive, inherently subjective, and, with most companies, statistically irrelevant because too few interactions are evaluated. While speech analytics—the underlying technology in an AQA program—cannot understand everything agents say or write and does require some human intervention, it can deliver statistically relevant findings and is much more likely to identify issues than the current QA approach. Despite these benefits, few enterprises have adopted AQA. This is because AQA solutions are complex and considered ineffective and untrustworthy. But with the introduction of machine learning, DMG believes these solutions will improve enough that companies will feel comfortable using them.


Artificial intelligence (AI) is the buzziest term in the technology arena. Most IT vendors in various sectors—speech and text analytics vendors among them—are claiming to provide AI-enabled solutions. But few if any vendors offer real AI, although interaction analytics in particular will benefit greatly from it. Imagine a speech or text analytics solution that identifies and quantifies the impact of a new trend without human intervention. Machine learning is starting to be used to accomplish this task, and while these capabilities currently require human intervention, DMG expects a new round of investments in AI that will make significant contributions and improvements in speech and text analytics solutions in the next two to three years.


As always, potential customers should be aware that all speech and text analytics solutions are not equal and require careful evaluation to ensure they have the desired features and capabilities. Speech and text analytics solutions provide quantifiable benefits to companies that invest in the right solutions as well as the resources and best practices to build an effective program. These solutions are highly valuable on a stand-alone basis, and their benefits increase as the technology and findings are incorporated into third-party applications and processes.

Donna Fluss is president of DMG Consulting. For more than two decades she has helped emerging and established companies develop and deliver outstanding customer experiences. A recognized visionary author and speaker, Fluss drives strategic transformation and innovation throughout the service industry. She provides strategic and practical counsel for enterprises, solution providers, and the investment community.