Verint Best Practices – Automated Quality Management (AQM)

Transcript

Matthew: Hello and welcome everyone, my name is Matthew Storm with Group Elite. Thank you for joining us today, we have a great team on the line here with us. So say hi Casey. 

Casey: Hi everyone.

Matthew: Hello Amanda Haney. 

Amanda: Hi everyone. 

Matthew: Hi, Hi. So without further ado, let's just get started. So what I'd like to ask first is just a really simple question, and I’m going to use the polling feature if it's all working for me great today, and that is I’ve put up a poll here and over on the right you should be able to see it, and we'd just like to know about your AQM experience. I’ve already had feedback people asking me if it was what was going to be educational or about AQM or whatever. So just go ahead and score the question. Do you do you guys see it there Casey Amanda, do you have it pop up yet?

Casey: It has not not yet, no I don't see it. 

Amanda: I don't see it. 

Matthew: Okay here we go, open the poll how about now. So tell me do you have Verint AQM and you use it, You have a QM for Verint and you don't use it, we use AQM from another provider, maybe you're using it from another vendor. We're considering AQM, or wait what what's AQM I’m just here to learn. So just if you would I’m gonna leave this open here for 30 more seconds. So just open up the polling over there on the right and just mark your answer tell us about that, while we're waiting for these 20 seconds, welcome Eric, hi welcome to the show. So looks like we've got a lot of we have it or they're just learning, got a few people that use another solution, some that have AQM and they're not using it, some people that are doing that. Alright. So we've got 15 more seconds. So if you're just joining go to the polling question really quick, do you have it, we don't have it, we have somebody else's, we're considering it and I don't even know what AQM stands for. Alright, okay great, well thank you very much for the poll that really tells us a lot. So we've got a lot of people here that they're using it, a lot of people that are considering it. So just to get us started here what I’m going to do is I’m just going to tell us a little bit about AQM itself. But first Group Elite is a very strategic partner and we're here to help Verint customers and Verint partners around the world get the most out of their Verint solutions. We've been doing this for 10 years, and we have great results and we've done this for hundreds of clients. Our services are for implementations and new projects and tailor-made projects and consulting engagements and, so without you know getting into a lot more detail, let me just tell you a quick update about what AQM are, automated quality management. Sometimes people say analytics, based quality management, and really what that's doing is that today in most organizations, if you take all of the interactions that you're having today, you're probably touching one to three percent of them through QM. Maybe even less than one percent. So in that blue box there that you see you know through your traditional QM methods you're looking at those issues and you're giving those as coaching sessions or collecting data. And it's giving you a partial view of the organization, but it's the other 97% of calls that don't get touched is where AQM comes in and actually uses the power of speech, can take the transcription of the call and then actually score those calls and give you an update about a hundred percent of the interactions. So what that means is that the transcription is used, so that you can take the rules and say did they do this did they not do this how many times did they do this and then either fully automate evaluations for 100% of the calls, or partially automate some of the questions that are in an evaluation form. So you open up a call you have an evaluation going and then some of those questions are already scored for you by again the power of the speech transcription of the call. So overall, that means that you can auto score some or all of the questions. A hundred percent of the calls if you choose to. It it creates a lot of engagement, Casey is going to be talking about you know some best practices about that because the employees really need to know more about what's going on there, maybe it allows you to shift some of your QM resources to coaching, or maybe speech analytics or some other tasks, or you can incorporate some automation and you can do it at your own pace. You don't have to do 100% of the calls 100% of the time but you might want to do some and in some organizations or in some departments. So what we're going to be talking about today are some best practices of how you take some of those questions you have in QM today, transition them using AQM so that you could take the power of that transcript, and it will score every call. Now you have all the calls they're all scored they how do that. How do you communicate that to your team, how do you go to your employees and you tell them I know not just one random call, I know the scores on all of your call well how do you turn that into coaching and make that effective and not overwhelming, those are some of the things that Casey's going to be talking about today and some of those examples. So that's a little bit about what AQM is, if you have more questions along the way, for Casey, or about what AQM is just hit the Q&A box or hit the chat box and we'll be able to take care of you. So what I’m going to do next is I’m going to stop sharing, I’m going to turn this over to Casey. Casey I’m going to make you the presenter, so that you can share your desktop. So before you do that as you're loading up your screens. Casey tell us a little bit about yourself how long you've been in and AQM, and and get us started.

Casey: Yeah, so thanks Matt and thank you for the introduction, I’m Casey McCloskey, I have been a senior business trainer and consultant with Group Elite almost three years. So July will be three years. I have about five years or more of just speech analytics dedicated and all around encompassed of seven to eight years of quality management. So really been diving deep and of course the last three years just really grinding in there with training and helping businesses along with their journey. So thank you for the introduction. I will share my screen in just a minute when we start really diving into a little bit more about AQM, also known as automated quality management. One thing that I do want to just kind of start with is what AQM is not. Right, so AQM is really like Matt had pointed out, there to automate a hundred percent of your interactions, whether it's a audio interaction or a taste text based interaction, but what it's not meant for is to replace your quality team. So that's a really big misconception sometimes when AQM is discussed is hey this will replace quality but really what it's designed for is to truly just AQM should help businesses get more insight on how their agents are currently handling those calls, it gives the opportunity to allow that quality team or the supervisors to really be tasked for other things set like Matt had mentioned speech analytics, or more coaching sessions, and it really just gives more flexibility. Also keep in mind through this process of AQM is quality is about subjectivity, where AQM is about objectivity. Right, so as a human, humans grade that subjectivity, where AQM doesn't have the possibility to do that. So if you have questions please make sure you drop them down in the little Q&A Matt will interrupt me and I’ll try and keep an eye on the webinar as well to see if any questions do pop in there. So let me share my screen with you. 

Matthew: Alright we see it.

Casey: Okay perfect, Matt just let me know if we get any questions then. So as Matt mentioned you have your two different types of forms, but let's talk about your fully automated forms right. So this is a published version of a fully automated form in Verint, and just to kind of walk through your fully automated forms should always well they have to be in the form of yes and no questions. You don't have a limit to how many questions that you can actually have in there, and again this is going to be based on a hundred percent subjectivity. I’m sorry objectivity when scoring. It automates a hundred percent of those interactions and it's submitted on the back end without your quality team actually having to go in and hit that submit button. So just to kind of talk through this. So this is actually showing you on the left hand side that it's based off of the transcription that is done on the transcription servers kind of like your speech analytics if you're familiarized with speech. When the question is answered it will populate greed. You also have the ability as an end user to overwrite these answers. So if you know just as an example that that employee did not follow that authentication process, you can actually as long as you have the privilege to do so is go in here and overwrite this, and change this answer to no. So that is what you're fully automated form looks like. The second one is your partially automated form. So like Matt discussed earlier when you have a partially automated form, it starts just like your normal evaluation when you're building it in form designer. It can be your manual for your manual questions can be the custom scored list types questions, but you're partially automated have to be in the form of yes and no. So this one is based on yes and no manual questions up at the top, but as we scroll down you start to see that there's different types of automation here. So we have the ability up here to manually monitor, and then down here those are questions that they did on a partially automated form. So it works the same way just a little bit different from your fully automated, where you still have to have those yes and no questions. So a couple of things to keep in mind as you're starting your AQM journey or your automated quality management journey, discuss internally, discuss if you want to use an existing form because you do have the ability to take an existing evaluation form and export it and then re-import it as a automated quality monitoring management form, and then define your rules that way. When you are going through your internal sessions, understand what types of questions you want to try and automate. During the calibration sessions, start thinking about your text based words. So because AQM is based off of that transcription server, that transcription that speech analytics uses as well, it's really using those text-based words to identify what are the callers saying, what are your agents saying. So when you start thinking about these internal conversations get an idea of the current questions you use and how can we change them into a text based form. Also partnering with your agent. So if you have in your business your quality team an AQM team, a speech teamm partnering with the agents to really understand their call types, how are they handling them, what types of questions and procedures do they have to follow when you're listening or communicating with your agent, and then work directly with your quality team and leadership on what their expectations are of AQM and what they would like to see out of that process. Once you meet internally, you've started to set your goal right, you've started to really hone in on an idea of what AQM is, how do you start the process, you partner with your agents, you partner with your quality team, understand what leadership's expectations are and once you start that foundation then you move on to your next journey of AQM. Matt do we have any questions so far.

Matthew: Yeah, actually we do. Does it handle acronyms well came in. 

Casey: So when we talk about acronyms, it's a lot like speech analytics right. So it depends, for example like dot-com where you would just put in the period com. it's not really going to search for that because your transcription is going to search for the word dot. So there's ways to kind of finesse that a little bit just like you would with speech analytics, and really getting in there with those acronyms. Any other questions?

Matthew: No not for now, I think we've got a few that I'd like to take at the end. 

Amanda: So I have a question.

Casey: Sure. 

Amanda: So Casey. you went through what AQM what AQM is not, and how it works, what about best practices, like if you have you have four minutes left what are those best practices that we should know to be successful with AQM.

Casey: Oh yeah absolutely. So partially automated forms right. So I’m a big believer in starting small and then building big right. So when you start with a partially automated form you're doing a couple of things for the business. A: you're getting comfortable with the system, you're getting comfortable with AQM to eventually broaden your horizon right. So that's the first step with a partially automated form. Your second step is really it's a time saver. So yeah we have the ability to have a fully automated form, but if you are in the position to do a partially automated form, and you have a long one like the one on my screen, where you have 15, 20 questions, these evaluations can take upwards of what 30 to 45 minutes sometimes to fill out. Partially automating some of these elite takes that time away. So then it opens up to going back to hey quality now is only spending 25 minutes on these evaluations, now we can allocate their time to something else. So partially automated forms start small before you go big right. So we all know with your fully automated forms, we can push your top what three questions out there the compliance statement, the grading, and the closing. Those are great right, but other type questions that you want to really like start small on, start small and then get comfortable. Your next one is KPIs. So if you're familiarized with form designer and you're the one that currently creates the forms, whether it's a manual evaluation or an AQM evaluation, creating a question as a KPI can be critical for AQ, because they really do go hand in hand together, and it's as simple as a click of a button. So when you're creating your form, again manual and or AQM, there is this little button in the question section that says automatically manage KPIs, and what that does is it pushes it to performance management, and allows it to be published for your scorecards. This does a couple of things it gives great visibility, you give your agents visibility to see on their scorecards where they're scoring on specific questions and how they're doing up against their peers, and just to kind of give you like a little bit of a visual here, you can see that some of these questions were submitted as KPIs. Coaching, it gives your quality team a better visibility on what types of questions that they should coach on right. So when we think about that currently right now as Matt mentioned you may be only doing one to three percent on an annual basis of manual evaluations, and sometimes you're coaching on the whole call versus the really impacted questions, and this really helps with that. Automating coaching sessions. So this can be completed under the organization alerts in Verint where you can actually automate those coaching sessions to be sent out to the agents as well as your quality team and supervisors depending on who actually does the coaching sessions for the agents. Again your KPIs go hand in hand. So AQM form data is actually only available in the solution for 60 days, but the really neat thing about having it as a KPI is it's an easy way to quantify your data, you can pull the KPI information in daily weekly monthly and even annual basis. So even though you may not have the reports for AQM specifically, six months to a year down the road by, putting it as a KPI you can quantify that data and pull that data from those AQM questions. Another one Amanda, I’m going through a long list right, because we got a lot of them. So tone, we all want to be able to measure tone on calls we all want to be able to say man they were super positive they were very empathetic, but it's hard to do sometimes without that human touch but then that's where we get into the subjectivity right, that's why this isn't to replace your your human, your quality team your supervisors, but think about how we can change the perspective of a tone-based question a little bit. Make a textual right. So think about empathy, we all want to know did our agents have empathy on the call were they concerned for the caller, did they meet the caller's needs right. So when you go through and you start thinking about how can we change a tone-based question from a manual perspective to an AQM perspective, it's all about the verbiage. So think of keywords as I’m sorry you're going through that, we're happy to help, I'd be happy to help, let's take a look and see what we can do about that. And you got to think of it from a little bit of a speech analytics perspective right. So if you are familiarized with speech analytics and even if you're not familiarized with speech analytics, think about the words that are being said on the calls, whether it's your agent or the caller, think about how receptive each individual person is being when they're on these calls. So just think about that. Another thing you know with this being close to speech analytics it does use the same transcription as the speech analytics categories used to be created.

Matthew: And Casey to that point, we're this is a great topic, and I’ve got a lot of questions coming in, and I know you have a couple more best practices to share. So let me share with everybody if you were looking for 20 minutes today we're going to go long. So you need to drop right now I’ll send you a link with the recording and you can catch up and I apologize for extending, if you can hang with us for a little bit longer stick around I think we're going to probably go another maybe 10 or 15 minutes today. So again if you're attending and you only had 20 don't worry about it we'll send you the link and thank you for joining. But I’m gonna also at the end we need to talk about that whole the transcription server and speech analytics I’m getting some questions about that, but I want to stick with the best practices for now, and get to that. One of the questions that came in while you were doing I think that's with this is can visibility rules be applied to a partially automated form.

Casey: It can, yes.

Matthew: Okay, alright. So go ahead and keep going, I think you were after wrapping up tone, go on to the next one and we'll get to these questions at the end. So go ahead.

Casey: Absolutely. So the types of forms you use right. So currently a lot of businesses have outbound forms, outbound call forms and inbound call forms right, and a lot of them have the exact same type questions. This is a really good way to combine those forms together because you actually have the ability when you're creating the AQM forms to select filters for each specific question. So, for an example, an outbound call you really want to measure the compliance statement right, this call may be monitored and recorded, but it's only for outbound calls. You can actually set that one specific question as the direction to only look for that on outbound calls, and then when it's answered because you defined it with that rule, it's going to if it's an inbound call that comes through it's just going to score that as an n/a and not count towards the end user. Another big point is touch points, touch points with your team internally is crucial for the success of AQM. I always recommend either doing it on a monthly or quarterly basis. A really good way to actually do this and you may be familiarized with it is calibration sessions right, we've all talked about calibration sessions and using the flags within Verint, you can actually do the same process but use an AQM session flag right, build an inbox for it and then really hone in on that and what that does is it does a couple of things. So it allows for when you meet internally for the users and your leadership to just go in pull those calls that were flagged into that inbox and listen to them. It's also a really good way to get your quality team and supervisors more involved. And that's what we want rightm we want all aspects of the business to be as involved as possible in this process, because education is key. 

Matthew: So Casey,not to cut you off there sorry. So you when on this screen where you talked about outbound and inbound, a question came in about can a question be applied only to a sub organization or a branch? or is it is it strictly at call flow and direction. 

Casey: So it can, and you actually do that on the filtering right here you would select your employees, and then select the organization that you would want it to apply to, and let's just say we would do contact centers, or a sub organization of tel telco mind and apply that specifically to that one organization.

Matthew: Okay great thank you.

Casey: Yeah no absolutely. So just going back a little bit to the internal conversations, it really I think I was talking about involving your supervisors and quality team. Let's say as a quality team member or a supervisor, I’m listening to a call but I’m noticing there's no automation to this specific type of call right. If you actually allow your users to flag these types of calls, they can then use private annotations to notate why they believe that, hey listen to this call and you may be able to automate some of the questions because we don't have a current form based on this call type, and really that's what your calibration sessions are going to be used for. So it's just another really good way to engage your end users in this process. 

Matthew: So you know I you know somebody asked that question I think the way they ask the question is actually quite brilliant they said. So just. So we're understanding you're building the speech analytics category saving the verbage then it's going to get picked up through that and then somehow you apply that to the form in the form designer yes. 

Casey: No. So it's two separate applications right. So speech analytics and AQM are entirely separate the only. The only thing that they have in common behind the scenes is the transcription, the transcription is what makes them to work similar and the text-based keywords. So, for example when you're creating a speech analytics category you use operators such as cancel near in capital letters account, whereas in AQM you don't use those types of operators you don't use the near you don't use the and. The system actually does it for you. So just taking a look up here, we actually in a category a speech analytics category you don't put several different syntaxes together right. You want to look for one specific thing at a time. Where here you can see that we're looking for the birthday, the birth and then an asterisk date, and that asterisk is actually your near operator that is used in speech analytics, it's called the wild card. So it's a little bit different. So no speech categories do not export or import to AQM. Two totally different applications and modules within Verint. So they do have to be built separately.

Matthew: Okay, is it a best practice to separate forms for different languages or to combine them into one form. 

Casey: So that's a good question right. So you can do the same form for multiple languages. If you have a dedicated spanish speaking form it can be good to separate it out sometimes right, especially when you start integrating in those KPIs, and really get an understanding of how to export and quantify that data via reports. So if you want a more visual visible aspect of it, I would say to keep it separate, but you do have a really good opportunity to use the same form based on those language types as long as your system is supporting Spanish or you know German, English, whatever the case may be.

Matthew: Okay, gotcha thank you. Do you have one more best practice to share? And then I i have a couple things to follow up.

Casey: Yeah absolutely. So I just got give me two minutes and I’ll let you wrap this up. So continuous accuracy tuning. So if you're in 15.2 there is an opportunity to allow your team members to submit suggestions based on transcription to your speech analytics and phonetic boosting team, which just allows as an end user to say hey this should be transcribing as compliance but it's coming up as complaint. So allowing your end users to kind of submit those suggestions can be really beneficial for the transcription. And then the last one which we've already touched on is education is key. Education in this whole process, your agents need to have the education your quality team needs to have the education, as well as your end users of the solution, because you're going from five to ten evaluations a month to a potential 50 in a day depending on how many calls or text based interactions that are coming in. So educating everyone is crucial to the success of AQM

Matthew: Sure. So let me walk through one quick thing before we end which is a little example of an example that we had with a client where we were working with the utility company just their QM wasn't addressing all their compliance needs that they needed, the randomly audited evaluations rarely hit on those regulatory issues, they had you know check boxes for everything that was extremely overzealous. We helped them use AQM to answer just those AQM questions, and have those easily located, really ready to coach, and the the end result for this was actually that they were able to design a form that cut the whole process in half before their QM team. They were able to allocate two of their QM team, they actually moved over into speech analytics and it was able you know they were able to do that and kind of shuffle their resources a little bit. The instances were ten thousand dollars a piece that they were trying to avoid. So it had a huge value to them if they didn't do that. This was just through six regulatory questions that were answered on 100 of the calls and and then that way that these remaining customer experience questions could be done subjectively as you said Casey. So I think this is an example of what we've seen, some other great resources before we run out of time and take a couple more questions is that on Group Elites web page as well as verint.com, there are a lot of great blog posts about cherry picking, about you know how to coach people from home, about multiple channels like speech and text and things like that, there's a new ebook that's coming out about building a speech analytics team and then maybe how you could incorporate AQM. Also check out the Verint marketplace, the Verint marketplace is a great resource where they've pre-built some things. So some compliance some order some very specific terms that are already built out, you could just download them and they're there for. So if you wanted to build out the 50 different ways that people could say I want to you know cancel my account, or the 80 different things that you could say when you're doing something that that's really good. So there's a lot of great resources out there. The other thing too, at the end today you're going to get an email from Eric and Eric's going to ask you a couple questions, and if you have questions maybe we didn't get to them today about your existing forms or you know do I want to go full or do I want to partially automate evaluations, or maybe you don't understand the impact it's like what does this mean for compliance, what does it mean to my QM team, what does it mean to customer experience if I do 100 of the calls. Feel free to answer you know, Eric can chat with you maybe set up some time or we could talk about if if AQM is right or what's your readiness level for AQM. We'd be able to answer some of those questions. So I’m going to jump back over here to Q&A because we've got quite a few more here. So there's a few other questions here, KPI's support of AQM powered coaching that's really brilliant. So that was just the comment for you Casey. We had another question about yes we'll send you the link if you missed this or you want to see the full recording again, we talked about that, we talked about these sessions. There was some questions, and Eric and Amanda do you see people that are using AQM and then they use speech analytics, or I already have speech analytics and then I’m using AQM as part of that. Maybe Eric, Amanda, and Casey, can you tell us a little bit about the the differences in the journey kind of between the two. 

Casey: So Eric if you want to all take that. So we've seen both right. So there is a lot of journey with just having a QM, solely because they want to be able to automate right. Automation is the new age of things, we want to move into that new technology of automating and being able to allocate our resources and other ways. So we've seen both, there are also a lot of customers that start with the speech analytics journey right. You start with speech analytics because you also have the ability to do what's called analytics driven quality, where you can build your speech analytics quality, I’m sorry your speech analytics categories, and then implement it into your shared inboxes to really help filter out some of those call types like retention right. Retention is going to be a big one in a lot of calls and call centers where we create a speech analytics category drop it into our retention shared inbo,x and then our quality team can go into that shared inbox and start measuring out those calls and evaluating them. So we've really seen it in both perspectives. So there's not the wrong way or the right way it's just understanding at the end of the day, how to truly build, start from the ground and build it, and each way you do it whether it's speech first or AQM, understanding the operators to use is going to be your key to success.

Matthew: Wow what what a great transition, because my next question is can AQM measure if statements, talk about operators. So the example they said if objection from the customer and a rebuttal was given. 

Casey: So you can finesse it right in two-fold. You would want to create two different type questions of the customer says this the agent should say it like this but not an actual if then statement. So there's ways to finesse it to make it look that way, but from an AQM built for specifically it doesn't do that if then statement.

Matthew: Gotcha. So the other question was about. So you don't have to have speech analytics to do AQM you can just have AQM. So that that came up, and then there was a question about retention on AQM forms of just being 60 days. Can you tell us about why is that that way what can be done and how it can be used maybe with KPIs and ao forth.

Casey: Yeah. So right now it is not adjustable, you know of course Verint it is a Verint product and at this point in time it is 60 days for attention to be able to pull those AQM forms. So think of it from a one to three percent perspective that you talked about in the very beginning, right now your system is seeing what 10 maximum maybe more calls per month per agent. So 50 calls a year and now we're going into so much larger data. You're talking every single audio or text based interaction has the potential to be evaluated with those fully automated forms. So it's much larger. So when you actually take that data and put it in as a KPI, KIPs can be ran on an annual level. So if you're a business who needs to have those reports pooled every single month, but you know that what's done today is not going to be available in say May 15 for an AQM form. Having that as a KPI will actually give you the advantage to say okay well I know that Casey scored at seven percent when our goal was 50 percent for the greeting for the month of February and that just gives us insight and you know providing it to leadership at the end of the year if you pull it on an annual basis. Again monthly basis, daily basis, weekly basis as well.

Matthew: Gotcha. Great well thank you for that question. So I guess it is you know making it a KPI makes that huge huge difference. So well Casey that is all the questions we have and for all of you I mean wow we've had nearly 100 people today that have stuck with us to the end. 

Casey: Thank you.

Matthew: Thank you, and again if you missed something or if you had colleagues that you know had to drop off we're going to be sending out the link to the recording. So Casey wow thank you very much, thank you for going 20 minutes long today, that was good 20 minutes.

 Casey: Yes thank you everyone, I appreciate your time.

Matthew: Thank you Amanda for your time and we hope to see you on a future webinar. So look for your next invite and if you have some other questions look for that email from Eric and he's going to be able to help you out or maybe arrange a quick call where we can discuss some of those other topics. So thank you everyone, have a wonderful rest of your afternoon or morning.