Verint Speech Analytics

It's not as appropriate. I'll go with the not first. That's easier if you're looking for call volume for specific volume questions and that's not speech. Speech is not going to provide you, like I said before, that absolute number of this is how many calls came in for this exact topic, why natural inherent limitations of speech, background noise and audio issues and strong accents like myself, so things like that. When we have had some departments come to us and some individuals have requests for insights from speech and we've said, and we discussed it for any request we receive, we have what we call a scope meeting to ensure that that request not only aligns with the goals and the value, it's going to bring value to the business, but to ensure that that is the best source of information for who's out. And sometimes they just didn't realize you can get it somewhere else. So if you're looking for actual call volume, actual numbers of the counts of what, that's not going to be the best source. It is the best source. I'd say when you really want to know what your customers, what your members are thinking, what are they saying? When you really want to know the voice of the customer, that is crucial because that interaction and listening to them is different than what the system can identify the tone and inflection and can identify the terms that they're saying that express joy or disappointment or dissatisfaction or negative emotion, whatever that may be. And yes, the system does an amazing job at identifying all that. For those calls, for example, where there is an accent, where there is background noise, where there are children and dogs in the background, the window's open because calling us from the car front, the freeway, none of those calls are really captured accurately. And so it's an entirely different insight. And sometimes you may have numbers, you may have data, but what impacts data, what brings the story together? I can give you numbers and say, Hey, we've had an influx of calls and the majority of them are talking about, they're expressing negative emotion regarding the Blue Shield website. Let's just say his hypothetical. I'm not saying that's what they're doing. So let's say they call and they have, so the metrics will tell us we're receiving these many calls, they hit the website category that we created, they're expressing negative sentiment, but is it going to tell you exactly what's wrong with the website? Nope. That requires human listening. So yes, the metrics provide great insight and I believe it provides a lot of direction. But when you provide to say to someone in senior leadership, those metrics, and then you also attach, oh by the way, this is verbatim a quote from our member regarding this issue. And that member says some things that are not very pleasant or they're truly expressing their genuine frustration and discontent that has more of an impact because it's not just a number, it's a human, it's the voice of the customer. So I think you need that whole picture to be able to present the scenario to be able to offer a solution. You can't offer the solution if you only have half of the problem or half of the issue. And I love that Verint Speech gives us the ability to do all that because then we partner with the analytics part with the desktop analytics and we can build triggers to know the actual counts and then we can bring that together with the voice of the customer and then present in an entire view so that informed decisions can be made.

That's still to be determined. Again, I'm still learning the overall picture here, so that would probably be something I can probably gauge later this year. Once I understand more what Verint can offer and what we have on the table to kind of pull those two together.

Well, my only experience with it is in a call center environment, so I guess the biggest piece is more so on the automated, since we've kind of been dealing with the automated scorecard, the functionality of that is absolutely great. But your AI does have some limitations. I mean it's just the nature of the beast in my opinion. So if I had to say anything right now, it's that we can't quite get everything we need out of that ai, but maybe there's more to come with that too. So yeah, my sentiment was an example of that, if I needed to give one so you better understand and we're just kind of getting with the voice of the customer and the irritants and things of that nature, we want to be able to try to really gauge that sentiment. And sometimes that's harder to do just through select words or to really get the feel of what's going on.

It is perfect for all of our interactions. It might have a redaction edit on transcriptions if we need to share data that contains HIPAA

We get good data for process improvements and for getting ideas for business cases.

Verint Speech Analytics is well suited for call reviews, honing in on customer sentiment, reviewing trends.

It easy easily suited to finding terms that are said in day to day language. Sometimes when you are looking for a unique scenario, it is difficult to find if the program name is not already in the Verint Speech Analytics dictionary.

Speech is helpful to localize areas of improvement. For example, we used to track FCR via telephony data (frequency of contact). Speech not only gives us a more accurate picture (if they call for the same issue) but allows us to dive into the top drivers, outliers within our product lines, etc which is much more actionable.

If your ticket categorization is not quite where it should be yet, or you find yourself having difficulty getting real time data quickly, then Verint Speech Analytics can be a great help to you.

Taking call data and translating it to text data to enable analysis. Text data is used downstream to determine the call driver to help guide business decisions and ensure adequate resource allocation.

Verint Speech Analytics is wonderful to get a more accurate view of how agents are performing across the company in specific areas. It also is a wonderful tool to help pinpoint areas of concern that need more attention and possible coaching opportunities.

Its well suited in :
- First call resolution
- Agent performance
- service quality measurement
- high topics
- enhance the process inside the company .

Speech Analytics works if a report is required for certain words used in a call. Contact Center administration is required to send a stop payments report to compliance every month and relies on Speech Analytics to find those calls. Speech Analytics is also used sometimes to indicate trends for issues for example mobile banking app/website is down or debit card issues.

PCI compliance to determine if credit card information is being taken

It is well suited for finding out near real time issues that are driving call volume to any particular area. Also good for training and development It is not as appropriate for use where instant feedback is necessary

It’s well suited for any call center to understand customer behaviors and call intents.

finding dead air on calls and displaying a graph for agent call reviews works well...some specific words that are wanted or not wanted in a call conversation are easily findable. In my opinion, other more generic words are often mixed in with other words inaccurately.

After training and using the system for 3 months, our team was able to find at least 6 items to bring to leadership for improvements in 1) agent performance, client experience, contractual improvements, and general insight into all recordings.

Simply put, if you're not looking for anything speech related here, then forget it.

Verint Speech [Analytics] is great for getting you to a group of calls for a particular topic quickly. You can see where the keywords occur so you can "listen" to a long call in short period of time by getting right to where you need to be in that call. It's so easy to use and manage, no IT involvement needed once it's up and running! However, it's not an exact word-for-word transcription, so it is not good for review calls by looking at the text alone as sometimes it doesn't make sense.