Idea Transcript
2.
Introduction
Speech analytics burst into the commercial market in 2003 and has rapidly become a “household name” in contact centers. Enterprises and companies of all sizes appreciate the power of speech (as well as text) analytics. They may not know how it should and can be applied, but they get the concept and appreciate the power and potential of an application that enables them to convert phone conversations into structured data that can be analyzed and used for the benefit of the organization (and hopefully) for the customer. DMG Remains Bullish on Speech Analytics DMG has been bullish on speech analytics since these applications entered the market. We appreciate and have been discussing the value and benefits of applying speech (and text) analytics inside and outside of the contact center. (Speech analytics is very helpful and beneficial for contact centers, but its true value is as an enterprise analytics solution.) We’re very pleased to see that in our 2015 Speech Analytics Customer Satisfaction Study, the top uses were different from past years. The top three uses of speech analytics are improving agent and operational effectiveness and efficiency, understanding the customer experience/capturing the voice of the customer, and improving sales effectiveness. The first of these uses is contact-center-oriented, which makes sense; this department typically owns and operates the application, as it’s where conversations are captured. The second top use in 2015, understanding the customer journey and capturing customer insights, is a much broader concept, which can deliver insights and findings to contact center managers as well as to many other departments in an enterprise. As delivering an outstanding customer experience is the top enterprise servicing objective for 2015 (based on a worldwide DMG benchmark study), it’s good to see that companies are doing more than just talking about this essential goal. They are using speech analytics to help them listen to their customers and are applying the data to improve the customer experience. Increasing sales effectiveness is the third top use of speech analytics in 2015. This is a highly quantifiable and impactful use of speech (and text) analytics, as customers frequently share their ideas and recommendations. These suggestions have historically been given to contact center reps, who are dedicated to delivering a great service experience or closing a sale, but are not typically interested in (or motivated to) collect these ideas (nor do most agents have a tool for gathering and sharing this information with their department managers or the company’s product management department).
-4-
© DMG Consulting LLC
Along with these uses are many others that benefit contact centers and, increasingly, other departments in the enterprise. Of course, one of the top uses remains the identification of the underlying reasons why customers (or prospects) call, if the application is being used by a customer service group. And for collections departments, speech analytics is being used primarily for script adherence and regulatory compliance. However, even for the most standard and common uses of speech analytics; the analysis is becoming significantly more targeted, which allows insights to be more actionable. Making Speech Analytics More Actionable DMG has been emphasizing the importance of making results actionable since speech analytics solutions were first introduced in the commercial world. Speech analytics has tremendous potential to function as change agent for both contact centers and other enterprise departments, but for this to happen, the findings need to be identified on a timely basis, there must be a way to tie the insights to specific issues or departments, and the results have to be actionable. Telling a department manager that they have a problem is a good first step, but accompanying this alert with a targeted action plan is a key step in transitioning speech analytics from a “nice to have” application for companies with deep pockets to one that is a “must have” because it is considered mission-critical. The vendors are finally listening to their customers, and some are starting to deliver speech analytics solutions that are tied to other analytics tools and performance management or next-best-action solutions that are prescriptive and identify the specific actions needed to address the opportunities. This means that some of the speech analytics vendors are finally starting to appreciate that these applications are useful by themselves, but their value increases greatly when they are combined and used with other applications. Where Speech/Text Analytics Fits into the Broader Customer Analytics Ecosystem Speech (and text) analytics solutions perform a unique function – identifying the meaning and insights contained within spoken (and written) conversations and discussions. (They also identify customer sentiment, but this is just another dimension of identifying the actual issue.) But consider the implications if, for example, someone leaves a message for an energy company regarding a gas leak, but the issue isn’t identified until a couple hours later or the next day – the information is not going to be useful if the leak has already caused an explosion. So, transforming the metadata into actionable insights on a real-time basis, along with the ability to use the findings for post-call strategic analysis, is the way to go. The question that the market has been trying to address is how to get there.
-5-
© DMG Consulting LLC
According to DMG’s strategic roadmap, predictive analytics will be instrumental in the future of contact centers. (We also think that it’s time for the name “contact center” to change, as these organizations are positioned to do so much more than service, sales, collections, etc.) In the future, contact centers will play an essential role in helping organizations deliver an outstanding, personalized service/sales/collections experience that is cost effective. While this sounds similar to the primary objective for many contact centers today, the practices and actions will be different. The major change is that the contact center of the future will be enabled with tools and data, and will also be empowered to make decisions for the organization, instead of just “following orders.” Of course, since people need guidelines and rules, the structure will be provided by a variety of predictive and prescriptive analytics solutions as well as a variety of operations solutions. Speech and text analytics will provide real-time and historical feeds into the contact center analytics ecosystem. Real-time speech will be used to automate the verification process, voice biometrics. (This process will also improve security and reduce the risk and cost of fraud.) Real-time speech analytics will be used for personality-based analysis and matching to enhance customer and agent satisfaction and improve productivity. (Getting interactions to the most appropriate resource will also allow for a more customized service/sales experience, which will also increase revenue.) Real-time speech analytics will be used to drive real-time guidance so that reps know precisely what to do. This will be followed by next-best-action recommendations, driven by a variety of predictive and prescriptive analytical capabilities. But this is just the beginning of the process, as there is a great deal that cannot be done in real time, which is why there is still a need for post-call speech and text analytics, as well as inputs from many other customer-facing applications, operational systems, customer relationship management, marketing, sales and many other solutions. While this sounds complex, it’s a gradual process, and is consistent with the direction of the market for a number of years. The emerging customer journey analytics solutions are an example of the new customer experience analytics framework. Speech analytics is a high-value application with great potential to deliver quantifiable strategic and tactical benefits. But there are still significant challenges that are impeding the success of these solutions. The most significant remain the shortage of experienced analytical resources to implement and manage these solutions. The second is the lack of maturity and functionality in some of the applications, which greatly limits what they can do. However, over time, these issues will be addressed, and speech analytics will be perceived as mission-critical. The story for text analytics is different. Text analytics has been around much longer than speech analytics, yet adoption continues to be very slow. Text analytics is a requirement for any company with a social media program,
-6-
© DMG Consulting LLC
although, it seems as if most of them don’t yet know this. The lack of interest and willingness to invest in text analytics will be slowly overcome as these capabilities are embedded in third-party solutions. Speech and text analytics are valuable on a stand-alone basis, but when combined with other applications, their contributions become even greater. Speech and text analytics functionality will be an essential component of many of the analytical solutions and frameworks that are emerging, and will power the contact center of the future.
-7-
© DMG Consulting LLC
21.2
Detailed Survey Findings and Analysis
Product Satisfaction by Category Survey participants were asked to rate their satisfaction with their vendor’s speech analytics solution for 21 key components, including: ease of use/configuration/maintenance, ease of integration with third-party applications, ease of creating searches and queries, accuracy and business relevance of results, ability to fine-tune results, call analysis capabilities, text analytics capabilities, emotion detection/sentiment analysis capabilities, ability to conduct automated root cause analysis, ability to conduct discovery and surface previously unknown issues or new and breaking trends, ability to correlate seemingly unrelated issues, ability to serve as an early warning system to identify issues before they escalate and impact a large number of customers, ability to make results actionable, real-time capabilities, reporting and dashboards, ability to support analytics-enabled quality assurance processes, ability to improve quality and agent coaching, ability to improve compliance with regulatory requirements, ability to detect fraud, ability to reduce customer effort, and ability to improve customer engagement. It’s important to note that not all of the vendors offer the same number and type of functional modules, nor are all of the survey participants utilizing every module that their vendors offer. For this reason, DMG included N/A as a response option in the product satisfaction section of the survey. A detailed breakdown of individual customer ratings by vendor for each product category can be seen in Figures 93 to 113. Scores that are not based on an average of three scores have been footnoted, as applicable. Figure 92 shows that most (43.7%) of the average satisfaction ratings for the 21 product categories surveyed fell into the highly satisfied (4.0 to 4.66) range, 37.3% fell into the satisfied range (3.0 to 3.66), 10.3% of the scores were completely satisfied (5.0), and 5.6% were only somewhat satisfied (2.33 to 2.66). 3.2% of the responses were N/A, as not all of the vendors offer the same number and type of functional modules, nor are all of the survey participants utilizing every module that their vendors offer. The 21 product categories in order of rank based on customer satisfaction are: ability to improve quality and agent coaching (4.44), ability to improve compliance with regulatory requirements (4.33), ability to support analytics-enabled quality assurance processes (4.28), ease of creating searches and queries (4.22), call analysis capabilities (4.16), ability to make results actionable (4.08), ability to improve customer engagement (4.06), ability to conduct automated root cause analysis (3.92), ability to conduct discovery and surface previously unknown issues or new and breaking trends (3.91), accuracy and business relevance of results and ability to correlate seemingly unrelated issues and ability to reduce customer effort, (each at 3.89), real-time capabilities (3.86), ease of use/configuration/maintenance, ability to fine-tune results and text
- 382 -
© DMG Consulting LLC
analytics capabilities (each at 3.83), ease of integration with third-party applications and ability to serve as an early warning system to identify issues before they escalate and impact a large number of customers (each at 3.72), ability to detect fraud (3.50), reporting and dashboards (3.16), and emotion detection/sentiment analysis capabilities (3.14). The most significant range of scores, 2.50, occurred in the areas of ease of integration with third-party applications and ability to detect fraud. Seven other areas with a wide range, 2.0, included text analytics capabilities, emotion detection/sentiment analysis capabilities, ability to conduct automated root cause analysis, ability to serve as an early warning system to identify issues before they escalate and impact a large number of customers, real-time capabilities, ability to reduce customer effort and ability to improve customer engagement. The wide ranges suggest important disparities among the vendor capabilities in these areas. CallMiner took the lead in the category of product satisfaction with an overall average score of 4.50. CallMiner achieved/shared the top score in 15 of the 21 product component categories. CallMiner achieved a perfect score, 5.0 for ease of integration with third-party applications, real-time capabilities, ability to support analytics-enabled quality assurance processes, ability to improve quality and agent coaching, ability to improve compliance with regulatory requirements, ability to detect fraud, ability to reduce customer effort, and ability to improve customer engagement. CallMiner earned/shared the top rank, 4.66 (highly satisfied), in 7 of the remaining categories: ease of use/configuration/maintenance, ease of creating searches and queries, ability to fine-tune results, call analysis capabilities, ability to conduct discovery and surface previously unknown issues or new and breaking trends, ability to correlate seemingly unrelated issues, and ability to make results actionable. NICE came in second with an average score for product satisfaction of 4.32. NICE achieved/shared the top score, 4.66 (highly satisfied), in seven categories, including a perfect score of 5.0 for ease of use/configuration/ maintenance, ease of creating searches and queries, accuracy and business relevance of results, ability to fine-tune results, call analysis capabilities, ability to conduct discovery and surface previously unknown issues or new and breaking trends, and ability to make results actionable. Genesys, with an average product score of 3.88, rounded out the top three vendors. Genesys achieved/shared the top rank in 5 categories, including a perfect score, 5.0, for text analytics capabilities, and highly satisfied scores for call analysis capabilities (4.66), emotion detection/sentiment analysis capabilities (4.0), ability to serve as an early warning system to identify issues before they
- 383 -
© DMG Consulting LLC
escalate and impact a large number of customers (4.66), and reporting and dashboards (4.0). inContact had an average score for product satisfaction of 3.78, including a perfect score of 5.0 in 4 areas: ability to conduct automated root cause analysis, ability to support analytics-enabled quality assurance processes, ability to improve quality and agent coaching, and ability to improve compliance with regulatory requirements. CallMiner took the lead in the category of product satisfaction with an overall average score of 4.50. CallMiner achieved/shared the top score in 15 of the 21 product component categories. CallMiner achieved a perfect score of 5.0 for ease of integration with third-party applications, real-time capabilities, ability to support analytics-enabled quality assurance processes, ability to improve quality and agent coaching, ability to improve compliance with regulatory requirements, ability to detect fraud, ability to reduce customer effort, and ability to improve customer engagement. Verint achieved an average product score of 3.53. This included highly satisfied rankings of 4.0 across eight categories: ease of creating searches and queries, accuracy and business relevance of results, call analysis capabilities, ability to make results actionable, ability to improve quality and agent coaching, ability to improve compliance with regulatory requirements, ability to reduce customer effort, and ability to improve customer engagement. Avaya earned an average product score of 3.29. This included highly satisfied rankings of 4.0 in two areas: ease of creating searches and queries and ability to correlate seemingly unrelated issues. Figure 92 provides the average satisfaction ratings by vendor for each of the 21 product categories surveyed.
- 384 -
© DMG Consulting LLC
Figure 92: Product Satisfaction Ratings, by Category 1 Vendor Ease of use/configuration/ maintenance
Avaya 3.66
CallMiner 4.66
Genesys
inContact
NICE
Verint
Category Average
3.33
3.66
4.66
3.0
3.83
1.66
3.33
4.00
4.0
2.50
3.72
2.50
Range
Ease of integration with third-party applications
3.50
Ease of creating searches and queries
4.00
4.66
3.66
4.33
4.66
4.0
4.22
1.00
Accuracy and business relevance of results
3.33
4.33
4.00
3.0
4.66
4.0
3.89
1.66
Ability to fine-tune results
3.33
4.66
3.33
3.66
4.66
3.33
3.83
1.33
Call analysis capabilities
3.33
4.66
4.66
3.66
4.66
4.0
4.16
1.33
N/A
3.83
2.00
3.33
3.66
3.14
2.00
Text analytics capabilities
3
N/A
5.00
2
5.00
2
N/A
3.33
4.00
2
2.0
3.00
2
3
3.50
3
3
Emotion detection/sentiment analysis capabilities
2.50
Ability to conduct automated root cause analysis
3.33
4.33
3.00
5.00
2
4.33
3.50
3
3.92
2.00
Ability to conduct discovery and surface previously unknown issues or new and breaking trends
3.66
4.66
4.00
3.00
3
4.66
3.50
3
3.91
1.66
Ability to correlate seemingly unrelated issues
4.00
4.66
4.00
3.00
3
4.33
3.33
3.89
1.66
Ability to serve as an early warning system to identify issues before they escalate and impact a large number of customers
3.33
4.33
4.66
3.00
3
4.33
2.66
3.72
2.00
Ability to make results actionable
3.66
4.66
4.00
3.5
4.66
4.00
4.08
1.16
3.86
2.00
Real-time capabilities
3.00
3
3
5.00
2
3.66
- 385 -
2
3
4.00
3
4.00
3
3.50
3
© DMG Consulting LLC
Figure 92: Product Satisfaction Ratings, by Category 1 Vendor Reporting and dashboards
Avaya
Range
3.16
1.34
CallMiner
Genesys
NICE
Verint
2.66
2.66
4.00
3.00
3
3.66
3.00
Ability to support analytics-enabled quality assurance processes
3.33
5.00
3
4.00
5.00
2
4.66
3.66
4.28
1.67
Ability to improve quality and agent coaching
3.33
5.00
2
4.66
5.00
2
4.66
4.00
4.44
1.67
Ability to improve compliance with regulatory requirements
3.33
5.00
3
4.00
5.00
2
4.66
4.00
4.33
1.67
Ability to detect fraud
2.50
5.00
2
3.50
2.50
Ability to reduce customer effort
3.00
5.00
3 3
3
Ability to improve customer engagement
3.00
5.00
Vendor Average
3.29
4.50
3.00
3
inContact
Category Average
N/A
4.00
1
3.00
2
3
4.00
2
4.33
4.00
3.89
2.00
4.00
4.00
3
4.33
4.00
4.06
2.00
3.88
3.78
4.32
3.53
--
--
3.00
Notes: 1. Not all of the vendors offer the same number and type of functional modules, nor are all of the survey participants utilizing every module that their vendors offer. A detailed breakdown of individual customer ratings by vendor for each product category can be seen in Figures 93 to 113. 2. Score based on one rating. 3. Score based on an average of two ratings. Source: DMG Consulting LLC, June 2015
- 386 -
© DMG Consulting LLC