Get More Value From your Data with an IVA
Companies that use an Intelligent Virtual Assistant (IVA) to collect and validate data can generate an immediate ROI.
Garbage In, Garbage Out
Bad data is worthless. People can be sloppy, and when we enter data manually, there’s a good chance human error will come into play and muddy up your results.
Contact center agents are crafty, and they will find shortcuts to make their jobs easier. This is often at the cost of accuracy, especially if they are skipping key steps to collect and verify data. Those steps take time, and if agents have incentives to drive their handling times down, well, those seemingly “extra” steps may be bypassed in some way. If it’s not a required field, it may just be skipped. If it’s required and they can’t skip it, they may not choose the most accurate option.
And then there are typos. Names, email and mailing addresses and order numbers, can all have characters missing or transposed. This leads to duplicate records or interactions that can’t be associated with the customer.
Business intelligence relies on data integrity. If the data is riddled with errors, your decisions may be misguided. Using an Intelligent Virtual Agent (IVA) or other similar technology can eliminate human error and complete data entry tasks accurately, every time.
Upgrade the IVR to an IVA
Most contact centers have some form of Interactive Voice Response (IVR) to greet their callers. It may be as simple as “Press 1 for sales or press 2 for service” or it may have complex routing, company directories, and features to initiate a callback rather than wait in queue. But the IVR is a tool of the past. Today, we have new resources available to digitally transform our customer experience.
The Intelligent Virtual Agent (IVA) replaces your IVR with a virtual assistant. The IVA can do what the IVR did, but it does it better. It can interact personally with customers, greeting them by name, gathering information from them, comparing it to other data, making decisions on next steps. As a digtial assistant, the IVA summarizes the CX history for the agent in a screenpop, whisper, or other audio-visual cue at handoff.
EXAMPLE 1: Using a MotionCX IVA for Caller Verification
In our first example, the MotionCX IVA uses a caller’s ANI (i.e., caller ID digits) to query a customer’s data warehouse for matching account and order information. For proper verification, the caller must verify 2 pieces of information, per the business rules outlined by this client. A successful ANI match counts as 1, and the IVA lets the caller know they need to provide another piece of information: their zip code, street address, or a recent order number. The caller just needs to provide the digits, the IVA will compare it to all possible matches.
With are seeing consistent success rates of ~30% of all calls verified by an IVR in the retail market. Given that some callers are not eligible for verification (e.g., new customers), the IVA is consistently able to verify over 60% of eligible callers.
EXAMPLE 2: Using a MotionCX IVA for 24×7 Data Entry / Updates
After verifying a caller’s identity, the IVA checks the contact record to see what information we have on file, cross-referencing systems and filling voids, linking call records to email and chat histories and painting the picture in a 360-degree customer profile.
The IVA is filling in gaps in names, email addresses, and phone numbers for 30% of the verified contacts, on average. This equates to about 10% of the overall volume. It’s working 24×7, so after hours, while the contact center is offline and we are all sleeping, the IVA is answering calls, verifying customers, updating records, and cleaning data.
Good business stories should have a dollar sign in them. In this story, the money comes from time saved. Greeting a verified caller takes approximately 20 seconds. Going through full verification, which includes data entry and repeating the spelling of names and email addresses can take from 40 to 90 seconds.
On average, the time saved on a verified call is around 45 seconds. Let’s say your volume is around 1 million calls per year, and your fully burdened labor cost is around $0.50 per minute. You can save $375,000 / year by avoiding the costs for 750,000 minutes of labor.
And that’s just on the quantifiable side of the coin. There are soft savings and trailing benefits. Improvements in CX reduce churn and drive repeat business, increasing revenue. Agents who have a better toolset and easier job are more likely to stay, helping to offset attrition. And for those who do attrit and are replaced, the speed to proficiency is improved through IVA assistance.
Other IVA Use Cases
The examples above are effective and easy to implement with the MotionCX platform. Taking things a step further, we can also leverage generative AI to summarize customer order histories and past interactions.
MotionCX can calculate the total lifetime value of a caller in milliseconds by measuring the customer’s purchases and sales histories against the time spent on service and support inquiries. Customers with a high lifetime value and low support burden can be prioritized or routed to special teams for white glove service. Customers with a higher support burden and lower lifetime value may interact more with the IVA before being introduced to a live agent for support.
Evaluating order history data, we can see if orders are still open, or had a recent status change, and use these events as prompts to help clarify customer intent. Once we know which order(s) and items are related to the inquiry, those data points can be pushed to the CX profile and other systems, taking another task off the agents’ plate and boosting data integrity through automation.
Can the IVA Replace all my Agents?
The IVA will quickly become our new best friend, but it has its limits. Think of it as a virtual SME or a new digital BFF, but not as full replacement of human talent. If key data points are not available, the IVA may struggle to verify a caller’s identity. If the systems the IVA pulls data from do not allow the IVA to initiate actions or events, then self-service may be limited. The capabilities of generative AI are impressive, but are not infallible, and engines still may need to be trained on brand-centric traits. Above all else, your agents still have the ability to empathize with your customers and personalize the CX.
The IVA of tomorrow is a bot empowered with machine-learning and natural language processing, capable of using generative AI and neural text to speech to resolve most customer interactions without the need to escalate to an agent. But that’s not the IVA of today.
Real-World Next Steps
Let’s face it, there’s a lot of talk and legitimate interest in AI and its capabilities. But to focus on present day tools and solutions that can be delivered in the near-term, MotionCX can help you harness the power of the IVA to gather and summarize information for your agents, saving them time and effort. It’s a real-world use case for generative AI that MotionCX can fulfill for you immediately.
Reach out to us, we’d love to talk to you about ways our IVA can assist your agents and improve your customer experience, data collection, and business analytics.