Performance Analytics: Critical Roles, Critical Needs

By Brendan B. Read

Senior Contributing Editor, Customer Interaction Solutions

There has rarely been a more critical time to ensure that contact center performance is optimal thanks to today’s challenging economy that is stressing cost containment and revenue maximization. Performance analytics solutions, when wisely designed and deployed, can play key role in meeting these objectives by identifying opportunities via interactions to cut expenses, increase agent effectiveness and output, retain customers and save sales, and grow revenues.

“Contact center managers are now realizing that they can make better resource allocation decisions based on broader trends with smarter analysis of existing processes, “ says Keith Dawson, Principal Analyst, Frost and Sullivan. “They are using analytics to demonstrate the revenue-generating capabilities of centers, especially when it comes to critical business-wide problems like customer retention and churn. Analytics gives them a chance to say ‘we may be asking for x $ for training, but if use it to identify agents with upsell skills, you'll see an ROI in just a few months.’”

Barak Eilam, corporate vice president and general manager, Interaction Business Applications, NICE Systems says that performance and interaction analytics have shifted from a ‘nice to have’ to a ‘must have’ in contact centers as these become more important to businesses. The tools give firms a quantifiable and structured way to drive higher performance by leveraging the insights in unstructured data, which account for 70 percent to 80 percent of information gathered by centers.

“In today’s dynamic environment, trying to minimize the gap between contact center operations and overall business goals without a structured analytics solution is ‘mission impossible’,” says Eilam.

Oscar Alban, Principal Global Market Consultant, Verint Witness Actionable Solutions, has spent over 20 years in the contact center industry: starting out as agent and eventually managing the 1,200 person inbound/outbound contact center division of a Fortune 500 company. He welcomes performance analytics as it enables centers to be managed more proactively. Equally, if not more importantly, it allows managers to get to the root causes prompting customer complaints and rising call volumes. Business trends, opportunities and deficiencies can be rapidly identified and addressed. So can the needs from other non-customer facing departments that help shape the customer experience such as billing, claims processing and order fulfillment.

“If you talk with some contact center directors today the traditional response to issues is to ‘chase the fire truck,’ meaning reactively follow up on customer complaints,” says Alban. “That’s not going to cut it anymore with the cost and complexities in today multi-channel support environments. With performance analytics companies can take a more proactive customer-centric approach to sales, service and support.”

New and enhanced analytics solutions
There are more performance analytics solutions coming onto the market. TelStrat’s recently released Engage Analyze indexes and audio mines words and phrases buried in calls using a patented phonetic audio search and recognition engine. Unlike older, less efficient speech-to-text approaches, it says phonetic speech search is not dependent on finite dictionary and grammar models which require constant maintenance. This makes it easy to accurately search for new competitors, product names, slang, and other dynamically changing terms.

Phonetic search technology also makes Engage Analyze much quicker than speech-to-text systems. Pre-processing or indexing of content is typically 60-80 times faster than real time, more than an order of magnitude faster than Large Vocabulary Continuous Speech Recognition (LVCSR) speech-to-text systems. Subsequent searches for words or phrases average over 30,000 times and reaching rates up to 80,000 times faster than real time. Engage Analyze enables audio mining up to 100 percent of calls, compared with limited (three to five percent sample rates) in LVCSR speech-to-text systems but without the massive computing power these demand.

UTOPY’s SpeechMiner® V5.2 offers a significant increase in scalability and accuracy allowing the processing of unlimited customer call volume from contact centers. It includes built-in support for new languages like French, Dutch, Italian, German, Chinese (Mandarin and Cantonese) and Japanese. It supports more than 45 languages. The new offering also has built-in industry-specific knowledge library for several verticals including communications, financial services, retail, healthcare, and utilities. Cutting costs with performance analytics
Performance analytics is being deployed effectively to reduce costs. One of the prime areas that companies are using it for is shrinking live agent zero-outs from voice self-service and calls from customers frustrated with websites as this has enormous ROI. Interaction costs with self-service are a fraction of those handled by live agents. For that reason they are also using analytics to determine which other live agent transactions can be automated.

“Firms want to keep people in self service, and in instances where this is successful figure out what made this self-service campaign or capability work so well so that they can do this with the other existing or new programs,” says Kraus.

Daniel Ziv, Vice President, Customer Interaction Analytics, Verint Witness Actionable Solutions, shared a customer example in which a credit card company was experiencing call volume spikes associated with lost passwords. The firm leveraged Verint’s speech analytics solution, and discovered that customers were being locked out of their online accounts because they could not remember the answers to selected ‘challenge’ questions. Consequently, their only option was to call the company’s contact center.

“With a process change that added the flexibility to select a different question, the company was able to reduce thousands of unnecessary calls, thereby simultaneously cutting costs and increasing customer satisfaction,” says Ziv.

Performance analytics is helping firms shrink average handle time (AHT), which helps both companies and customers. Jeff Schlueter, Vice President Marketing and Business Development, Nexidia points out that by carefully analyzing calls opportunities may be found to shorten them.

One key AHT determinant is time spent on hold. One health insurance company used speech analytics to show that calls relating to out-of-plan issues were taking significantly longer than the overall average handle time, much of which was spent with the customer on hold. By modifying the process, solving the issues without making the customer hold, and then notifying the customer with a callback, they significantly reduced the time and cost to handle these issues and also improved customers’ satisfaction.

“Contact centers need to reduce headcounts by shortening calls and driving more calls into the self-service, which will help them provide good customer service at less cost, “ says Schlueter. “Analytics can get the vital information to make that possible.”

Performance analytics is helping to reduce costs in other channels. Ronald Hildebrandt, co-founder of Enkata and its senior vice president of marketing, pointed to a large financial services firm that in 60 percent of retail transactions saw web transactions within two days prior of visiting the branches. It applied analytics to calculate a true cross-channel first contact resolution rate and identified the specific contact types where the Web fell short of retail capabilities, which prompted unnecessary branch visits. After a few minor web updates, the problem was solved and the company improved its customer experience and saved millions.

“We're seeing increased cross-channel traction with performance analytics that links Web, contact center and retail channels together, which provides visibility into how these channels interact to produce an end-to-end customer experience,” says Hildebrandt. “Companies have been surprised by the amount of unnecessary cross-channel traffic that has been identified. If this were reduced, it would not only improve the customer experience but also save money.”

Performance management, predictive analytics, and performance analytics
Aspect sees performance analytics as a subset of performance management. Performance management drives alignment and visibility across all levels of an organization. The demand for which is growing as its customers see dramatic results, such as improved productivity and enhanced customer interactions.

“Analytics is descriptive; for example, it might focus on the service level for the previous half hour or an agent's real-time adherence to schedule, “explains Allyson Boudousquie, Director of Business Process Marketing, Aspect. “Performance management is prescriptive and enables a structured process through which a company manages and improves its overall performance against certain key performance indicators. Examples include improving customer satisfaction and customer retention for the year by aligning agents’ operational activities to support the customer satisfaction improvement objective.”

A companion to performance analytics is predictive analytics. Predictive analytics permits firms to anticipate customer behavior by mining the same voice and online interactions as performance analytics, but is also then pulls in other data, such as demographics, transactional and market research information, to analyze this complete dataset.

Colin Shearer, SPSS senior vice president of strategic analytics explains that employing predictive analytics can bolster performance analytics by enabling managers to obtain accurate behavior prediction, and use that to drive customer interactions. This data is then used to build and enhance customers’ profiles that can be used, when customers identify themselves, to guide them directly to IVR self-service, live agents or specialized care via CRM systems. Predictive analytics can, by intimately knowing customers, makes each call more successful, which improves performance, increases customer satisfaction and revenues.

“These days everyone is being stressed to get more out of existing assets,” explains Shearer. “One way to generate more value is to apply predictive and performance analytics together to make sure customers have better experience, remain more loyal and to do things like see if there are previously untapped sales opportunities.”

Performance analytics issues and solutions
While there is little question that performance analytics is effective, in many instances it is being hobbled in its adoption by high costs that arise from having too many features that add to purchase price and to labor costs by having to hire dedicated analysts. All told these factors can in some cases add 50 percent or more of the total costs; speech analytics solutions typically price in the hundreds of thousands of dollars. Implementation times are in the 12 to 18 month range: twice to three times the tolerance of today’s senior management.

“Even if there is a solid payoff at the ends of it the entry price is so high that only a handful of customers are going to be looking at it and even many of them are saying it is too cost-prohibitive to tackle," reports Kraus. “But by taking the bloat out of these products and providing simple to understand reporting more companies will buy, use, and benefit from these solutions.”

Envision Telephony has done that with Envision InteractionIQ™ that simplifies speech data processing, search, and reporting from within audio recordings. An integrated component of the Envision Centricity WFO platform, it contains the core speech processing, capture, reporting and analysis capabilities needed to fully incorporate audio data into WFO analytics.

InteractionIQ does so at a fraction of the cost of standard speech solutions, by doing away with some of what the firm reports, via comments from its customers, are largely underutilized features. They include emotion detection because emotions such as anger do not necessarily translate into lost business, reports Envision’s marketing director Jim Shulkin. Another tool, full-time transcription, often provides too much data for management to parse through when they need to quickly get important information such as cancellations. In this batch are real-time alerts that do grab urgent utterances that require responses, with examples including ending service. Yet serious problems with products or threats to agents or firms where real-time action is truly needed are limited.

Envision InteractionIQ features a processing filter that enables self-managing the total cost of speech analytics by focusing processing power on the most pertinent and relevant interaction. It also has a subscriptions tool to set-up ‘saved searches’ to have only filtered recordings including specific voice data delivered to an inbox to expedite review. It has an ad-hoc search that performs on-demand queries for new or specific terms as/when needed. There are ‘Smart Tags’ that permit unlimited terms to be marked in all processed interactions for immediate reference and “click through” ability to recordings during review.

“Up to now we have not seen widespread adoption of speech analytics even though it can offer tremendously valuable insights to agent, center and organizational performance and trends,” says Frost’s Dawson.” InteractionIQ drastically reduces the total cost of ownership for speech analytics while still delivering the market’s core and most practical needs for speech processing technology.”

Verint’s new Impact 360 Speech Analytics Essentials solution features its proprietary Complete Semantic Index, which automatically identifies significant changes in customer behavior as expressed within recorded customer interactions. Such changes are proactively surfaced by the software’s Automated Trend Analysis, which identifies increases or decreases in terms and phrases used during customer/agent conversations.

With the Complete Semantic Index, users do not need to know in advance what terms to search. Intuitive search engine-like, guided search capabilities—including contextual suggestions and search visualization functionality—help users find relevant calls quickly to determine the underlying causes of rising call volumes, costs and customer dissatisfaction.

“Most small and medium sized centers don’t have the resources and time required to build search categories, nor the business analyst staff to interpret them,” says Ziv. “The Impact 360 Speech Analytics Essentials really brings this type of technology into mainstream contact center operations without costly setup overhead, lengthy consulting engagements, and separate analysts.”

To get the most out of performance analytics solutions firms need to apply them across all contact center functions: billing, collections, and sales as well as customer service, and in all channels, thereby giving them a ‘single version of the truth’. Merced Systems’ performance analytics products integrate data from all of these areas, giving managers a single view into every customer interaction.

“Firms need to understand where they are to know where they need to be,” says Mark Gally, director of marketing at Merced Systems. “Performance analytics is today a must-have system in the contact center just like ACDs and workforce management, providing visibility, performance improvement, and delivering a culture of accountability. It is that window of where companies are today, that allows firms to know how best to tweak their operations in this difficult economy. It guides them along their overall process. To do that though they need a complete view of their customers via their analytics solutions.”

The following companies participated in the preparation of this article:

Aspect
www.aspect.com

Enkata
www.enkata.com

Merced Systems
www.mercedsystems.com

NICE
www.nice.com

SPSS (News - Alert)
www.spss.com

TelStrat
www.telstrat.com

UTOPY
www.utopy.com

Verint (News - Alert)
www.verint.com

CIS Magazine Table of Contents