|
There are many tools available today that perform analytics. The goal of these tools is to help organizations gather intelligence from customer transactions. They do not, however, have the capability to extract meaningful and strategic insights from customer interactions and combine them with insights gathered from transactional systems, which is the key to improving customer analytics and reporting.
Today there is a powerful new domain evolving for extracting insights from customer interactions: interaction analytics. Interaction analytics are driven by qualifying principles. In the past, up to and including today, the various contact center systems that have been implemented have focused solely on quantifiable customer metrics; for example, average handling time, service levels (what percentage of calls were answered within a certain time frame), etc. Interaction analytics bring an understanding of the customer experience to the next level by merging these data with qualifying metrics. For example, why did a customer become irate and how did the agent handle objections? How and why is an agent successful at identifying and closing sales opportunities? How well did he or she fare on customer feedback surveys?
The kind of input that can be derived from interaction analytics provides a strategic benefit at all levels. The contact center can better understand the performance of its agents, supervisors and the contact center overall; the organization’s marketing department can gain critical competitive insights and input regarding customer “wish lists”; the business development department can better identify and leverage new business opportunities; and more.
The key to extracting qualifying insights from customer interactions is to consolidate both quantifying and qualifying information that is extracted by as many contact center systems as possible; i.e., recording, quality monitoring, speech analytics, application screen content analysis, customer feedback, CRM, CTI and workforce management. Furthermore, the greater the number of different analytics methodologies that are applied to customer calls (for example, keyword spotting with ad-hoc query capabilities, emotion detection, talk-over analysis, speaker recognition), the easier it is to understand not only what was said, but in what way, by whom and, most important, what the implications are.
Interaction Analytics
The goal of improved customer analytics is to gain a deep understanding of customer behavior and be able to address a broad spectrum of key strategic issues both on the contact center and enterprise levels. These issues could be improving agent performance, streamlining coaching packages, increasing customer satisfaction, decreasing customer defection, upsell/cross-sell, ensuring compliance or preventing identity fraud.
Within the analytics realm, speech analytics (speech recognition and verification) have been used primarily by contact centers to automate and improve self-service applications and to authenticate customers automatically in an attempt to reduce transaction time, not to gain insights from the customer interaction.
But when deployed in one unified platform, speech analytics can allow contact center supervisors and decision makers to cross-reference data to gain insights into what is truly going on with their customers, their agents and in the contact center in general.
Ideally, the interactions platform, which captures and analyzes data from customer interactions via voice channels (traditional and IP telephony), CTI and agent computer screen activity, should apply multi-dimensional analytics, which include a variety of methodologies and technologies such as those mentioned above.
The Importance Of Multi-dimensional Interaction Analytics
The multi-dimensional approach to interaction analytics entails compiling results from monitoring keywords and phrases, detecting the customer’s emotion level, gathering input from the agent’s screen activity and application events, as well as gathering input from various business systems.
This kind of approach can quickly alert managers to customers at risk of defection. For example, multi-dimensional analytics can combine input from the CRM system, where the customer’s average monthly buying patterns have significantly shifted towards fewer and fewer purchases, with key phrases being uttered during calls to the contact center — phrases such as “not satisfied” or “didn’t work” — along with poor customer feedback results for the handling agent. All of these can alert management to a customer about to defect, allowing the company to take steps to remedy the situation before the customer decides to go to the competition.
When combined with emotion detection, the benefits of interaction analytics become even more powerful. Emotion detection is a recent exciting development in speech. This new capability detects a speaker’s emotional “event” — or heightened emotion, which is triggered by happiness, frustration, anger and so on. There are several features that help point to a speaker’s emotional state, including voice pitch levels, speaking rate and stress distribution, among others. First, a baseline of emotion is determined during the first seconds of a call, when the speaker is least likely to be excited or frustrated. Next, the software engines pick up on any deviation from that baseline and conclude that the speaker is in a heightened emotional state.
In the contact center, emotion detection is critical in identifying true customer satisfaction/dissatisfaction and pre-empting potential defections. If a customer expresses dissatisfaction, the call can be flagged and routed to a member of the management staff. The issue will then be reviewed, and the caller will receive a callback instantaneously. This can result in unprecedented responsiveness and customer loyalty.
The traditional approach to spotting keywords and phrases also needs to be revised to improve customer analytics. The benefits of keyword spotting in monitoring the content of calls have become well known. However, this benefit is not being fully exploited when implemented as a discrete engine, separate from the recording/QM platform, as is normally done. This approach is limited in that it enables the creation of reports that can provide only statistical compilations.
Ad-hoc search capabilities can transform saved voice interactions into a searchable database by creating an index file for every recorded customer interaction. Once calls are indexed, ad-hoc queries can be performed to retrieve calls that contain keywords or phrases that were spoken at any point during the call, without having to pre-define in advance. In the contact center,
emotion detection is critical in identifying true customer satisfaction / dissatisfaction and pre-empting potential defections.
Another way of improving the accuracy of interaction analytics is the separate capture of the customer and agent sides of the interaction, and performing independent audio processing for each. This provides critical qualification of relevant calls. For example, through a unified approach, supervisors can verify whether the words “buy again,” as picked up by the word spotting software, signified an upsell opportunity (Customer: “I will want to buy again”) or a routine sales pitch (Agent: “Would you like to buy again?”).
Reporting
Improved reporting entails centralizing the results of multi-dimensional interaction analytics into unified dashboards. This enables managers to review all the relevant and critical performance parameters of the agent, the team, the supervisor and the contact center. He or she can spot trends regarding customer satisfaction, sales targets, handling time of customers and more. The manager can then correlate exceptions and take corrective action to improve performance at all levels.
Unified dashboards support KPI-based (key performance indicators) management. Driven by organizational and individual performance objectives, this is an extremely effective management tool. The agent, for example, can receive the kind of feedback that is required for increasing motivation, feeling connected to the organization and understanding what underlies the success or deficiencies in performance.
Supervisors can answer questions such as: How are agents/teams performing relative to pre-defined objectives? Who is excelling? Who needs help? Which agents are receiving the best feedback from customers? How does the performance of my team compare with other teams? Is there a correlation between my agents’ adherence to scheduling and the quality of their interactions with customers?
Furthermore, business users can answer questions related to customer and market dynamics, such as: Are my customers happy with the service they are receiving from our customer service representatives? Which products are getting the most calls? Why are customers calling in about these products more than others? Is my marketing campaign effective? How is my telesales force doing?
A Case In Point
To illustrate the above principles, let’s take, for example, the contact center of a bank’s credit services division. This contact center is looking to apply advanced customer interaction analytics and receive comprehensive reports to help them increase dollars collected, ensure compliance with bankruptcy and privacy laws, and ensure high levels of customer service and satisfaction.
This requires improving each step in the collections model: being more effective in contacting the account holder; improving collection skills in securing the customer’s promise to pay; getting immediate payment on the phone to increase kept promise rates; and, ultimately, increasing dollars collected.
The bank can achieve these goals by implementing a unified platform that consolidates information, applies multi-dimensional interaction analytics and generates cross-application reports. Furthermore, the implementation process should entail certain preliminary steps critical for successful deployment: business processes benchmarking and reviews, requirements definition, interaction analytics definition and tuning and further engineering where required.
This bank can best leverage the benefits of this approach by creating a customer analysis group or business interactions analysis group in addition to the traditional quality monitoring group, which performs random monitoring and focuses on standard agent skills.
This second group is compromised of a function that is new to the contact center, business interactions analysts. These analysts would perform precision monitoring, enabled through multi-dimensional interaction analytics, to focus on specific business issues and improve agent skills that have a direct impact on dollars collected per head.
Once these processes are implemented, with the proper human and technology infrastructures in place, the bank is ready to go. By integrating business data generated by their CRM system, the bank can identify, for example, calls/customers who had terminated the service. The bank can then correlate these data to actual calls recorded with these customers. The calls are identified by a mix of high emotion detection and by picking up on words and phrases such as “I have other credit cards” or “I will close my account.”
This information can then be aggregated in a report that is generated and pushed to the user’s desktop to analyze all the relevant parameters; e.g., agent performance and adherence to scripts and compliance regulations, effectiveness of marketing campaigns, workflows, coaching/training, processes and more.
Another benefit of this approach is preempting customers at risk of defection. By combining keyword spotting with emotion detection, customer dissatisfaction can be assumed when high emotion is detected in conjunction with words or phrases such as, “I already told you,” “lawyer,” “supervisor,” “manager,” “I don’t understand” and “How many times do I need to say?”.
Consolidating customer insights generated from as many contact center data systems as possible and applying a broad variety of interactions analytics enables contact center and enterprise decision makers to improve customer analytics and reporting. This enables them to address key business issues such as driving and protecting revenue, ensuring compliance and corporate governance, nurturing customer loyalty, increasing agent productivity, improving agent negotiation skills, improving upselling and cross-selling and compiling critical business intelligence.By Illan Kor
NICE Systems
Ultimately, interaction analytics can improve operational efficiency and strategic effectiveness to help companies understand what is actually going on during customer interactions and, most important, why. CIS
A Luxury Item No More
By Steve Sanden, Cincom Systems, Inc. |
If you have ever moved in with someone — whether out of undying love or a more practical need to pay the rent — you know how hard it can be to merge two sets of belongings. Suddenly, you find yourselves with two sets of furniture, two sets of silverware and two sets of TVs and remotes. How will they fit into one abode?
Most companies have felt the same way about upgrading their legacy CRM systems. What is the best way to merge your old data with any new data? And how do you collect data and report on the old and the new?
In the past, only large companies could afford to purchase an expensive new proprietary database that would convert their old data to fit new technologies. But now, smaller companies can take advantage of new streamlined systems that provide data collection and reporting capabilities and enable organizations to easily tap into their legacy customer data. Small businesses also gain the ability to anticipate customer trends and make strategic decisions. In short, what was once a luxury item for a few is now mainstream for the masses.
A Mountain Of Data
Most companies have no problem collecting data. IT systems that gather an impressive amount of customer data, call data and even advanced metrics have been around for years. What many smaller companies lack is the ability to put those data into readable and practical formats for better business decision-making.
Many legacy systems also fail to collect the right kind of data. For example, if a small company wants to discover where its customer calls are abandoned, an older CRM system may not be sophisticated enough to decipher whether a call was abandoned at the private branch exchange (
PBX (News - Alert)
), in the customer queue or during mid-conversation.
Why does this matter? If calls are being lost in the queue, for example, the company may want to hire more CRM staff. It may also want to change its sales or support approach, since it appears to be turning off some customers.
Beyond the mass of simple statistics, the real value in gathering data is anticipating customer trends. Many smaller companies or departments don’t know which metrics to study. They simply don’t know where to begin climbing the mountain of data to reach some sort of business enlightenment. They are not alone.
New Insights From Old Data
After spending a great deal of time and money installing expensive, older CRM systems, many veteran CRM managers remain extremely frustrated. Most feel they need a Ph.D. just to run a report. Even when they can run a report, it often doesn’t give them the information they need.
Many older CRM solutions also ignore the wealth of customer data sitting in a company’s legacy systems. Most solutions that have bridged the gap between legacy systems and new data are also cost-prohibitive for a small or medium-sized company.
Fortunately, new solutions build a better bridge between legacy data and new, real-time data. What sets them apart from old solutions is their primary goal — to help companies get to know their customers better by easily tying old data to new information. This new, merged information is also readily available in easy-to-navigate, customizable reports.
Marrying these two data streams instantly creates new efficiencies. The new technology solutions allow CRM managers to see events and transactions across the organization for every customer. The “marriage” ties each customer’s current interactions to previous transactions. Reporting on this conjoined data allows anyone to immediately understand what’s going on.
These data need not sit in a report on a manager’s desk. The real-time capabilities allow a CRM manager to create business rules that change the customer service or sales script based upon a customer’s previous transactions. A CRM agent can instantly view previous transactions and preferences for that customer.
Each transaction can be individualized and specialized. The customers feel their unique issues are being understood, and they don’t feel as if they are wasting time repeating information. Agents spend less time gathering and securing redundant data. In addition, the company’s image as a responsive organization is immediately boosted.
The Cost Of Understanding
The technology that makes these solutions possible is now affordable for small and medium-sized companies. Many vendors offer hosted solutions, which means small companies no longer have to invest heavily in a massive IT infrastructure to compete with larger companies.
By using a hosted solution, companies use their existing IT infrastructure while leveraging the technology and experience of a vendor. The vendor houses much of the necessary processing power and data storage at a central location. Companies can access and report on their data from any PC using an easy-to-navigate, Web-based interface. Customer data remain secure, protected by the vendor’s enhanced knowledge of the latest security and identity-theft-prevention measures.
Moving to a hosted solution also means that companies can open their data to access across their enterprise. Sales, marketing, customer support and purchasing can all gain new access to existing and newly generated data. Companies can establish administrative rules for which areas (or even which users) have access to certain types of data. By expanding and customizing data access, companies allow every employee to respond much more quickly to customer requests, as they can now instantly know the history and customer preferences behind each one.
A hosted solution dramatically reduces the necessary business cost of knowing your customer. More important, a good vendor provides tools to thoroughly and efficiently mine your new and legacy customer data so you can make better decisions about every aspect of your business. Even in the best marriage of data, the ability to hone knowledge about those data to make strategic choices provides the true return on investment.
Eliminating Barriers To Customer Retention And Growth
When you share a dwelling with someone, you may learn something shocking about that person. You may discover he or she still loves Duran Duran or is fond of teeth flossing at the table. In today’s highly competitive global marketplace, those revelations — good and bad — make a company much more responsive to its customers’ changing needs.
Small companies need to think and act like the big guys. A hosted CRM solution allows them to do just that. They can marry new data with legacy customer information to gain new insights into their customers. They can report more effectively on customer trends. They have resources to better understand their customer data and make strategic decisions. Putting serious analytics and reporting into the hands of any company makes bold decision-making for growth possible. CIS
Steve Sanden has worked for seven years in software development utilizing technologies ranging from PHP/Perl to Java/JavaEE. He is currently a software engineer at Cincom Systems (www.cincom.com), working as part of the Synchrony Development Team.
|
|
|
|
SUBSCRIPTIONS
TMCnet LOGIN
WEBINARS
