Call Center Quality Assurance

4 Uses of AI in the Contact Center

Artificial intelligence (AI) has multiple uses across the modern contact center. While some people mistakenly believe that robots are going to replace live support agents, the truth is that AI in the contact center actually helps customer service agents perform their job better. Here’s how.

  1. Data capturing during customer interaction.

There’s a lot of data to be captured during every interaction. AI can be used to capture this data and then feed it into the contact center’s analytics system. With features like sentiment analysis, AI can also be used to spot certain emotions, like anger or dissatisfaction, which can then escalate the issue to be handled by a live agent.

  1. Management of customer data.

Capturing all of that customer data is just the first step to actually understanding it. Data has to be analyzed and leveraged in order to actually improve the customer experience. AI can help with this by capturing and cross-referencing data, then sharing it across different channels and platforms. This way, the customer won’t have to repeat their details every single time they contact customer service, and they won’t receive offers or messages that don’t truly relate to them.

  1. Smart replacement of IVR processes.

A contact center’s IVR system will have a set of pre-defined rules to follow. These rules are generally simplistic – for example, the IVR system may transfer a sales call to the sales department. AI can take this several steps further by using natural language processing and machine learning to understand customer statements instead of just giving them a set menu of choices.

  1. Directing customers to different areas of the website.

Many customer requests can be handled simply by pointing the customer to a specific area of the website. For example, a customer may be able to get information about their account or recent payment by viewing their account information, eliminating the need to speak with a live agent. Customers can also engage in self-service by finding the answer to their FAQ. A virtual assistant can direct the customer to the right section of the website, freeing up live agents to handle more pressing issues.

There’s always going to be a need for live, human interaction. In order for contact center agents to deliver the best, most personalized support possible, AI tackles easier-to-handle customer queries, speeding up the process on both ends.

4 Essential Components of Your Workforce Optimization Software

Delivering a positive customer experience is no small feat ­– there are a lot of moving parts that have to work together, with workforce optimization (WFO) being a major component. When considering which WFO suite to go with, keep the following four must-haves in mind.

  1. Integration with Existing Systems

The WFO system you use should be compatible with the rest of your contact center. Cloud WFO solutions are typically the easiest to integrate ­– they can be custom-fitted to your contact center, prepped and tested before going live, and even run along with your current WFO solution as you make the switch so there’s no downtime.

  1. Creation of Reliable and Adaptive Schedules

With the right WFO solution, scheduling becomes much easier. Your WFO software should generate schedules with enough agents to cover daily shifts, accounting for agent requests like certain days or times off, flex shifts, or work-from-home shifts. At the same time, your software should review shift data to accommodate for high and low patterns, which will affect things like breaks and training sessions. Your WFO solution should also be flexible enough to adapt when something unforeseen occurs that requires a quick change in the workforce.

  1. Real-Time Schedule Adherence

In order for management to know if an agent’s daily activity is in line with contact center objectives, you’ll need to see reports about schedule adherence. Your WFO solution should monitor and record real-time adherence, tracking log in and log out times, plus lunch breaks and other types of breaks. For contact centers that have out-of-the-box needs, like after-hours coverage, your WFO solution should let you create custom guidelines.

  1. Accurate and Robust Reporting

WFO (and just about everything else at your contact center) revolves around reports ­– otherwise, it’s very difficult to know what’s going on in your business. Even the best managers can’t be everywhere all the time, which is why they rely on reporting. The data that’s gathered will help you figure out where changes need to be made and what type of training needs to occur moving forward. Comprehensive reports will help you make the right workforce decisions.

The philosophy of WFO ­– shifting the workforce for the sake of optimal productivity ­– has been around for a long time, but actually embracing this philosophy by seeking out the tools to achieve it is still new for many contact centers.

4 Best Practices for Analyzing Quality Assurance Data

If you’ve ever called a business, you’ve heard the recording that goes, “This call may be monitored and recorded for quality assurance.” The best contact centers don’t just record customer calls, though – they also monitor them and then make changes to improve the customer experience. Here are four best practices to help you put call analysis to work.

  1. Clarify your goals and processes.

It’s impossible to know what parts of call analysis to pay attention to if you don’t have a game plan. Which KPIs will you measure? How many calls will be evaluated and within what timeframe? Who will review the data? What will be the process for recommending and implementing changes? There’s going to be an overwhelming amount of data coming in and you need to know what to do with it.

  1. Get everyone involved.

Determining goals and processes should involve the whole team, from management and executives to the agents themselves. Everyone should feel comfortable making suggestions and providing feedback. Also, make it clear that quality assurance is an evolving and collaborative process. Quality monitoring can sometimes be stressful to agents because they feel like their every move and decision is being judged. By emphasizing constant improvement instead of perfection, agents will be more relaxed and effective on their calls.

  1. Pay attention to repetitive comments.

Trends should get a lot of attention, whether they’re positive (compliments) or negative (complaints). While single incidents on either end of the spectrum are not unimportant, tending to the more common issues will have the biggest impact. Start with the major influencers before working on one-off issues.

  1. Allow for subjective ratings.

Scores shouldn’t always be taken at face value; allow some wiggle room for context. For example, you may have an agent who has lower-than-normal Average Handling Time scores. However, their customer satisfaction scores may be higher than average. It’s possible that they’re taking extra time to better understand and soothe the customer, which could be more valuable to the contact center than if they sped through the call just to solve it quickly. Before trying to fix a problem, make sure that there’s an actual problem to fix.

By monitoring phone calls and then using the information you discover to your advantage, you can streamline the agent-customer communication experience, improve contact center performance and greatly increase customer satisfaction ratings.

How Speech Analytics Affect the Outcome of Calls

By the time a customer has contacted a live agent, they’ve probably tried to troubleshoot the problem on their own with self-service tools. When they’ve reached the point of wanting to speak with someone, they’re already part of the way through their customer journey. Real-time speech analytics take into account customer history so they can pick up where they left off instead of starting from the beginning.

Real-time speech analytics help agents determine the right thing to say to a customer in the moment in a variety of situations. On top of making sure the customer is directed to the correct agent or department, this technology also gives agents the current, relevant information they need to solve the customer’s problem. Examples of up-to-the-minute information agents will receive include:

  • Issues that are trending on social media.
  • Topics customers are currently calling about the most.
  • Recent updates to products or services.

Real-time speech analytics technology, combined with information being fed to agents in the moment, means that the customer support offered will be tailored to the individual.

Management can program speech analytics to choose agent scripts based on specific speech cues. Software is able to identify words and phrases that are present as well as those that are absent. The software also takes into account sentiment; the point in a call when a word or phrase is said; and the absence of a word or phrase when it should have been said. On top of improving the course of a call while an agent is on the phone, speech analytics can also pinpoint larger gaps in training and find areas for improvement.

The best speech analytics technology will understand the context of a conversation in order to appropriately guide the agent. Customer calls are analyzed in real-time and conversational indicators make it possible for agents to proactively handle a call in a way that’s highly beneficial to the customer.

Advanced speech analytics software helps contact centers in a number of ways. It increases first call resolution and improves the customer experience. It monitors agents for regulation compliance and adherence to company policies. Agents can also use real-time speech analytics to recognize and take advantage of sales opportunities.

3 Ways to Use Analytics for Quality Assurance

Trying to optimize your contact center’s quality assurance (QA) program? Consider these three tips.

  1. Your QA should be focused on high-value calls.

Randomly selecting calls for QA gives you an average of QA rates, but no actual insight into particularly good or particularly bad calls. The best and worst calls are important because this is where agents either improve or ruin customer loyalty.

Every interaction between a contact center and a customer costs money. When you QA a low-value call, you’re just adding onto the cost without getting anything insightful in return.

Instead of randomly choosing calls to score, first assess calls based on the data that’s auto-tagged to them via speech analytics. Call transfers, calls that are put on hold, high-value orders and repeat calls are the most important types of calls to score. You can still randomly select calls, but select them from within a certain category.

  1. Use speech analytics to monitor 100% of calls for quality assurance data.

You can setup speech analytics categories for many of the QA agent conformance checklist items. These include:

  • Proper greeting
  • Call recording disclaimer
  • Verification of personal information
  • Proper closing

You’ll be able to get a percentage score of calls where this information was and was not used. By looking at the scorecard, you’ll be able to glean important insight. For example, you may see that a particular agent isn’t closing the call properly most of the time.

This is a way to off-load a big portion of QA monitoring. Instead of doing this manually, you can have your system do it automatically. Then, you can shorten the amount of QA questions you need to answer, such as if the agent provided the customer with the right answer and if the agent had enough product knowledge.

  1. Adjust the amount of QA evaluations you do for agents based on performance.

Agents who regularly perform well don’t need to be monitored for QA as much as other agents. For example, if an agent has a QA score above 90%, they can be monitored less frequently the following month. Furthermore, agents who have low QA scores should be monitored more frequently the following month. While it may take extra time to setup additional monitoring or to reduce the frequency of monitoring, it’s a good way to give struggling agents the extra help and attention they need.

How to Improve Your Contact Center’s Quality Assurance Program

Quality assurance (QA) programs are designed to make sure customers receive consistent, high-quality service every time they get in touch with a contact center or an agent gets in touch with them. Data collected through QA programs provide the contact center with information needed to plan training and incentive programs for individuals and teams.

Monitor Agent Conformance Questions Using Speech Analytics 

Speech analytics can be used to monitor conformance-related agent tasks for each call, including:

• Greeting

• Call Recording Disclaimer

• Verification of Name

• Verification of Account Number

• Verification of Mailing Address

• Verification of Contract Terms

• Empathy

• Closing

While speech analytics cannot be 100% accurate due to factors like accents and poor call quality, they can still be used to show the estimated percentage of calls where each phrase was or was not said. Speech analytics can be used to monitor the conformance questions on the QA scorecard, allowing analysts to manually record and focus on more important agent skills, such as accuracy, tone and product knowledge.

Shift How Contact Center Agents View QA Evaluations 

It’s important for agents to see that QA evaluations aren’t punitive, but that they help companies provide customers with better service. Explain how calls are selected and ask for input about QA scorecard questions. When agents understand the benefits of a QA program, they’re less likely to get defensive and more likely to work with the company toward a common goal.

Training New Contact Center Agents

When hiring new agents, set them up with a Boot Camp QA Scorecard for the first month or two of their employment. Instead of standard “yes” and “no” questions, make each question subjective by giving options like “excellent” and “needs improvement.” There can also be a section for agents to make additional notes. This will help them understand the aspects of calls that matter most and give them a way to track their progress.

Flagging Calls: The Good and the Bad 

Agents should be allowed to flag two types of calls: ones that are especially challenging and ones that they think they performed highly on. When a call is difficult, agents can flag it to ask for coaching. These calls should not be scored. When a call goes particularly well, agents can flag it to be reviewed and added to a best practices library. Excellent calls can be used for training purposes and can also help agents compete in QA contests. Great calls should be celebrated and agents should get feedback about what went especially right during the call.

Weekly vs. Monthly QA Evaluations 

Often, QA evaluations are due at the end of each month, which means the last few days of the month are prone to an overload of monitoring and feedback. Instead, have a couple of QA evaluations due weekly. Agents will get feedback consistently throughout the month, which will help them avoid repeating mistakes over and over until their next evaluation.

Quality Assurance Self-Evaluations 

Allow agents to self-evaluate a percentage of their calls each month. If you’re going to go over a call with an agent, let them review it ahead of time. They’ll be better prepared do discuss what went well and which aspects of the call could be improved. You can also provide them with the QA evaluation prior to the meeting. If they received a low score, giving them the evaluation beforehand will offset any shock they may experience during the meeting, clearing the way for you to have a productive coaching session.

How to Deliver Proactive Customer Service

When contact center agents are proactive about the customer experience, customer loyalty increases while support calls decrease. By identifying and solving small problems before they become larger, contact centers receive fewer inquiries overall. Additionally, proactive customer service is a way to control word-of-mouth. Customers routinely discuss service experiences with friends, family and social media networks. By being proactive, brands are better able to frame those conversations and present their company in a favorable light.

Offer self-help options.

Every brand’s knowledge base should be regularly updated. Customers will be able to find the answers they’re looking for, reducing the number of support tickets agents receive. The knowledge base should include product information, usage advice, common customer issues, best usage practices, troubleshooting tips, terms and conditions, and shipping info.

Provide live chat support.

Live chat is ideal for shoppers in all stages, including when they’re casually browsing and also when they’re ready to make a purchase. Customers like to use live chat for certain types of enquiries, like getting their order status, finding out about current promotions or asking about shipping. Live chat adds a human touch to the customer experience and gets questions answered quickly. Having a live agent walk a customer through part of the sales process via web chat, without waiting on a phone queue, is extremely convenient.

Monitor customer conversations.

Customers talk about brands both in private and online. When it comes to social media mentions, software and online tools make it easy to monitor brand mentions. When customers say something that requires customer service action, agents can act quickly. If a person is complaining, it’s best to apologize and offer a solution. If a customer recommends a brand or product, thank them and show gratitude for their business. Every mention, whether positive or negative, is a customer service opportunity.

Monitor customer actions.

In order to be proactive about the customer experience, it’s best to monitor and track how customers use the brand’s website. This allows for audience segmentation based on actions. Then, relevant messages can be sent to segmented groups. For example, one segment can include people who have created an account on the website but who haven’t made their first purchase yet. Those people can receive an e-mail asking them if they need help finishing their purchase. Another segment can include regular shoppers who are sent a special “thank you” coupon every few weeks.

Deliver post-sales support.

Companies often forget about providing post-sales support. Shipping an item doesn’t conclude the customer-brand relationship; there are customer service opportunities following a sale, too. Plus, it doesn’t take a lot of effort to provide post-sale support. After a transaction, contact customers to find out if everything is going smoothly. For example, sending an e-mail to find out if the customer is satisfied gives the customer the opportunity to explain a problem before becoming frustrated and posting online.

Pay attention to support trends.

Support tickets should be monitored in order to spot trends and recurring issues. When a problem repeatedly comes up, like an issue with a particular payment method, fix the root of the problem. This will help manage minor issues before they get out of hand, which will in turn increase customer satisfaction and limit the number of incoming support tickets.

Announce company mistakes ASAP.

Customers should find out about mistakes from the company, not on their own. By discovering problems themselves before the company is able to get ahead of them, issues can escalate quickly and publicly. Brands should always be transparent with customers. For example, when an order is shipped later than promised, e-mail the customer to let them know. Problems can snowball when they’re hidden or ignored.

Positive customer experiences can mean the difference between having a good company and a great company. By being proactive about customer service, the customer will have the best experience possible.

Contact Center Quality Assurance

As customers become more educated and informed, it’s becoming increasingly important to monitor and measure quality assurance in relation to the customer experience. Listening to the way the customer is greeted and parted with; finding out how the agent responds to the customer’s needs; and preventing duplicate records are three ways to oversee quality assurance.

1. Greeting the Customer and Closing the Conversation

The way the agent answers a customer call, e-mail, text message or social media message will set the precedent for the remainder of the customer experience. Standard, friendly greetings, such as, “Thank you for calling Company X!” or “Hello, this is John Smith, thank you for calling Company X! How can I help you today?” are simple phrases for greeting customers and starting off on the right foot. The closing has to be done right, too. The way the agent says goodbye plays a big role in how the customer will remember the experience. Asking particular closing questions and adhering to a friendly sign-off are important factors of the closing.

2. Measuring Customer Need vs. Agent Response

Often, inbound customer inquiries are for buying, upgrading, troubleshooting or canceling a product or service. It’s important to monitor how agents handle these inquiries in order to get insight into the quality of the conversation. Do the agents know the answers to the questions being asked? If not, do they know how to find the answers? Are they delivering the answers appropriately, such as by using the correct tones and offering clear explanations? By monitoring these quality issues, you can determine which ones need to be addressed.

3. Preventing Duplicate Records

When duplicate records are prevented, day-to-day contact center activity improves, and forecasting and analytics become more reliable. There are two main ways to prevent duplicate records with quality assurance software: during batch record import and when entering new leads.

Batch Record Import: When several new records are imported to the system at one time, duplicates will be identified and flagged before the duplicate is recorded. Administrators can decide which record to update. Eventually, once administrators are comfortable with the software, the system can match and merge duplicates automatically.

New Lead Duplicates: Quality assurance software can identify a duplicate when a contact representative is creating a new record. Advanced software is able to do this even when errors are made, such as misspelled words or data that was entered in the wrong fields.

Quality Assurance vs. Quality Control

There’s a difference between quality assurance and quality control. Quality assurance refers to a set of processes that ensure a quality result. Quality control refers to evaluating and testing the end result to check that quality specifications were met. Quality assurance goes beyond ensuring that the company’s processes and procedures are being adhered to. While you won’t ever be able to avoid quality control entirely, when quality assurance is successful, there’s often less need for quality control.