Smart Service Desk with an ROI
July 08, 2015
Routine technical issues can slow down a support center like password reset calls or system outages. Find out how to streamline incoming service management inquiries and quickly build a ROI strategy using automation.
Hi folks, my name is Mike Heberling, I'm with HEAT Software. I'm the Director of Business Development and I focus a lot on our service desk, our call center tools. And what I want to talk to you today about is the ability to sort of smarten up your service desk and establishing some ROIs. As we all know, sometimes creating a return on investment strategy is both difficult and painful in terms of reporting, and I'm hoping to show you some ideas today that will help you through that process.
In terms of service desks, we look at many types at HEAT Software. We are supporting service desks that have HR support. That might include payroll, benefits, those types of functions. We also have, within those HR payroll service desks, we have the need for both Self Service Voice Applications as well as Live Agent connectivity. And you'll see displayed in this diagram, some of the types of things that we can provide via self service. We can update phones, we can provide personal time off updates, benefit updates, annual enrollment functionality, all of those can be done via self service, and then we can also deliver traditional skills-based agent routing to those individuals as well, that are calling in.
Besides HR, we have significant need for customer service types of service desks. And the key to customer service desks and the type of services that HEAT Software offers, is combining both the telephone and the back-end CRM system together to provide a consolidated customer history. Again, in those instances there's a great opportunity to deliver self service via voice as well as an improved agent experience.
One of the things that I'll be focusing on today is our Enterprise Service Management for IT. And again, with those, there's a need for skills-based routing to agents that are ready and available to take telephone calls and there's the opportunity that we have to reduce the amount of calls that the agent takes and to streamline the services for the employees through IVR Voice Self Service. So we'll be focusing primarily on Enterprise Service Management for IT today. The sweet spot of value that HEAT Software brings in that the ownership of HEAT Software has worked on for the last 11 years, is combining both telephony applications as well as the back-end database, the CRM or ticketing portion of the service management application, again, whether that's HR, Enterprise Service Management for IT, or Enterprise Service Management for Customer Service. And that red area in the middle is that service automation that HEAT can provide. Because we own the source code for both the telephony application as well as the back-end CRM, we can provide advanced functionality from an automation perspective without customers having to spend a lot of money on middle-ware and applications that need to be developed to bring those traditional solutions together.
Some of the examples...I'll have a couple of examples in here as we go through these slides today, but one of the examples is a company in Tampa, Florida, PMSI, that did some work with us, they're a pharmaceutical company. And they were working with Avaya and some lesser-known CRM kinds of tools to provide services to their clients who were calling in regarding workmen's compensation and other things like that. So, over a period of, you know, three months, we measured the functionality differences as they adopted the combined solutions of HEAT Software as well as the telephony products. So you'll notice, pre-HEAT installations, they were looking at an average time in queue of up over 40 seconds, and then as the months went on, they were able to reduce down to you know, the 28...sub-25 and then in the 15 second range. So, as time went on and as we were able to refine the information that was contained within the HEAT database and then direct that in through the IVR, we were able to actually to reduce time in queue on a consistent basis. Another statistic that they monitored was their percentage of calls answered in under two minutes and you'll notice that pre-HEAT, pre-installation of the HEAT application, then compared to post-HEAT plus voice implementation, the percentage of calls answered within two minutes greatly increased.
Another statistic that many organizations look at in service management is calls abandoned, and you'll notice that they were up pre-HEAT installation with voice, they were up above nine percent, and then that drops significantly. And the reason it was such a dramatic drop is because there was intelligent routing that was capable within the system. Now, we are able to look within the HEAT database with the pre-defined connectors and identify the information that was important to the callers, and route that to the right agent so that the callers were being answered more timely and they were being answered more intelligently and so there was less of a need for them to abandon at some point in the IVR because of routing. And lastly, PMSI was looking at calls per day over the four minute time, which is, really the opposite of course than the two minutes so, reducing the amount of calls. So, pre-HEAT was up over 30 and then below that over 4 minutes it begins to drop significantly over a period of time.
Now, the way are able to demonstrate this type of functionality was through a return on investment strategy that we worked out with PMSI. And so what they realized was that they had significant phone-in queries coming in and, over different types, and that what they needed to do was intelligently route them to the agents based on what they were. So, what we have here is the definition of IT service desks stuck in cycles today. And the Help Desk Institute, through their Practices and Salaries Survey, on an annual basis, produces figures that show the average fully-burdened cost of an agent based on these types of incidents that are created. So if somebody walks up to a desk, $29.30, if a telephone comes in, the average fully-burdened cost of handling that telephone cost is $27.60. Now, if we compare that to the very bottom, which is self service at $13.50, it becomes very easy to understand, if we can take a fully-burdened cost of $27.60 and change that into workflow representing a $13.50 completion, then we're saving the organization both time and money. We save time from the agents perspective, we save time from the employees perspective, and we're saving the cost associated with actually handling that incident.
The Types of Incidents comes from Gartner on the right-hand side of the slide, and so they've recognized that for IT service desks, that 20-35% of calls traditionally coming into the help desk can be associated with password reset, or 15-25% can be related to major outages or outages that have come in or requests for repair regarding large-scale outages. So what we're gonna look at today is the concept of password resets and how we can help manage that and then provide an ROI.
Now, the calculation for the ROI, rather than going through and determining what the cost of all your call center agents are, what the cost of your lines are, what the cost of your square footage is, it's very easy to come up with an ROI calculation based on the information provided by HDI on an annual basis. So, if we know that, based on their surveys that an average phone cost is $27.60, and we've actually worked with some organizations through Hopson and Company and done and have completed some calculations associated with that, and those are actually very close. We've seen anywhere from $21 to $35 as the variable within that. So, if we said that an average phone incident cost was $27.60, and if we can take that incident and move that through automation, some type of self service in this instance, we're gonna talk about voice self service, and reduce that cost to $13.50 we have a $14.10 savings per incident.
Now, what we want to do is spend a little bit of time in talking about the value that we bring besides lowering the cost per incident, and we're gonna get a little bit more into the incident costs in just a second. But there's the cycle that has appeared in the most recent cycle that we've seen, started in 2008 with the difficulties that we had with the economy in the United States. We saw a lot of organizations starting to reduce their staffing levels at service desks, whether that be HR, Customer Service, or IT, to make up for the lost volume of revenues that were seen from the reduction in, you know, from the... depression or recession that we went through at that time. So, we've seen low staff levels, reduced supports and budgets, but now as we've entered into 2009, 2010, 11, and 12, and certainly according to the Wall Street Journal and some other papers like that and studies, we've seen increased activity not only from Wall Street, meaning stock prices have not only gone up as we've seen since 2010, but actual volumes of market activity is going back up. And so what we're facing, from a service desk perspective, is we're facing the reality of increased real volume of activity with reduced staff levels that are hung-over from 2008 and 2009 and a reluctance on the part of CIOs, CEOs, and CFOs to re-spend on staffing because we don't know if this recession is bottomed out or if we're just hitting a lulls and we may hit that again.
So one of the ways that we have seen the ability to break this negative cycle of reductions and improvements and re-staffing and de-staffing, is to use automation. And some of the automation that we've seen at the IT service desk from customers that we've done business with that have been very successful, is to use our IVR, the HEAT Voice IVR, and make that available to callers to what we call call-deflection. And call-deflection means not only just answering the call with an IVR, but actually performing all of the activities that an agent would normally do associated with that. So, password reset is a good example. Call comes in, "I need to reset my password," well an agent's going to check the security of the caller, create a ticket for it, go into the systems and reset the password and then close the ticket out. And if that can all be done through self service, through in this instance voice self service, we can lower the cost from $27.60 to $13.50, right? So, that's the opportunity that we have for breaking this negative cycle of staffing, de-staffing, and the changing volumes associated with markets.
One of the organizations that we have done business with in the past is Radiant, they work with point of sale systems for organizations like Burger King and Jamba Juice and they maintain their point of sale systems. So, previous to the deployment of HEAT, they were working with Avaya phone system, great phone system, and then Vantive CRM solution to manage the tickets that were happening. Now, both of those are good solutions standing alone but what HEAT provides in combining those, without having to build a lot of middle-ware, the most expensive part of this type of solution, providing that value in the sweet spot that we talked about earlier, they were able to see their Average Daily Call volumes drop from 1,106 to 619, their Average Speed to Answer drop from 1 minute and 54 seconds down to 38 seconds and so forth. Their Call to Call ratio increased significantly, which, of course, drives value at the service desk, and in fact, Radiant was able to turn what was for them a cost center into a profit center. From what was before, everything costing them money for every call that came in, they were actually able, through some intelligent routing and through some service automation, HEAT service automation, they were able to turn that into a profitable center for themselves.
As we look at the ROI perspective, some of the things that we sometimes miss is that we focus on the Level 1 Support. And, you'll notice Level 1 Support, according to the Help Desk Institute in the Eastern part of the United States, Level 1 folks are costing organizations $40,000 a year. In the Central part of the United States, $39,000 a year...these are, again, fully-burdened costs...and in Western United States, $48,000 a year. Many times what we focus on is, "How can I save a few seconds at the help desk?" Certainly Radiant was able to see that value and change it from a cost center to a profit center, but sometimes maybe the focus is wrong regarding internal IT support.
If we're focusing on strictly saving some seconds at the service desks through screen pops or through password reset or some of that service automation, certainly we can potentially pay for the automation within a short period of time and reduce our staffing, Level 1 support staffing. But that is not what we recommend on an ongoing basis because that does not help us in that negative cycle that we get into regarding de-staffing, staffing, de-staffing, staffing, and all the associated costs with that regarding training, hiring, all of those long-term costs that we have associated with keeping people up to date with technologies and keeping them effectively working with our clients and/or employees.
So I believe that sometimes what we're missing is, instead of focusing on savings seconds at the service desk, what we really should be focusing on is preventing downtime on the manufacturing floor, making sure that executives are up working. Primarily making sure that sales are not being lost and that we're actually hitting the revenue lines that we're expecting as we come out of this recession that we went through a few years ago, making sure that shipping dates are being met. So if we can take service automation and take rote and redundant tasks that are being handled by people, and move them into what I call "people helping people." And so in other words, a password reset is really sort of a mechanical effort. We need to find out who the caller is, we need to provide some security around through authentication and validation, we need to reset the password, and we need to create a ticket to it or for it, to auto-start the event. If we can change somebody from spending seven or ten minutes and with chitchat maybe some, maybe fifteen minutes on that call, take them from that activity to actually helping the CFO get the payroll system back up or making sure that the manufacturing line, responding to the manufacturing line being out, or making sure that the shipping labels and so forth are out there, and if the, you know, the shipping printer is working so that shipping is going out the door. If we can change it from rote and redundant activities to people helping people, then we've got a win on our hands and we're saving more than $40,000 in the East and $39,000 in the Central part of the United States and $48,000 dollars in the West. We're actually adding millions of dollars to the bottom line for organizations. It's just really a change of focus, and really a change of focus on where the ROI needs to come from. But I realize that hard dollars need to be shown and so we're gonna show you that as well.
Here's an example of what I've been talking about. So in this instance we have IT Support for Bill, the Finance Guy, who's calling in, and Jason, the IT Guy, and he calls in and leaves a message for Jason. Jason doesn't get the message right away, it's in the voicemail. He calls Bill back 40 minutes later, he looks Bill up in the tool, HEAT, whatever the tool is if you're not using our voice products with it, and looks him up, we create a new ticket, we dispatch the service text and we click Complete the Call. So, Jason's salary is sitting in there at $35,000 a year, he's created a ticket, talk time was seven minutes, cost was four dollars for associated to that ticket, for just that portion of it, again, not a fully-burdened cost.
On the other hand, we have the clock ticking over here on Bill. Bill's an executive, he's our CFO. He's waited around at 47 minutes at a cost of $57 and that's really clicking it. So instead of focusing on the $4, maybe we can save working with Jason, even though that's important, and we want to show our return on investment there, maybe some of the focus should be, "If I could have freed, excuse me, if I could have freed Jason's time and he wasn't busy doing something else that was maybe a little less important, there was more of a rote or redundant nature, maybe I can have gotten Bill back up and working a little bit faster."
So if we can identify that Bill or Jason was the person we actually needed to take the call. Bill, the same Finance Guy calls in, he's got a laptop that won't boot and it's going to cause a slow-down in processing of the payroll. So, the call comes in, the system looks up to see who is that's supporting Bill or who is it that's supporting payroll, we do some skills-based routing from the telephone system, and then we look up where Jason is and we deliver the call to Jason wherever he is and he immediately accepts the call, takes the call, and so, instead of a cost of $4 because of the chitchat and calling him back and all the other things, it's now a cost of $2.32 for that particular telephone call, not the incident though, but for that particular telephone call we've gained a 42% efficiency increase.
But that's really not what we're looking for; what we're really looking for is to make sure that Bill doesn't miss payroll and all the employees get mad, right? And so, we can save on 4,000 calls a month, an annual savings of $80,000, which is a couple of agents' times, right? If we look at $43,000 for the East or $48,000 for our first level support person, then we kinda have to save that time, there's no question about it. Over a period of calls, we can save one or two people. But the idea isn't to let those one or two people go, the idea is to re-focus their efforts to make sure that higher revenues are dropping to the bottom line.
So let's look at a password reset example. And I will warn you that the math doesn't work out straight away, right off the air, and so don't take 875 and multiply it by 14.10 and expect 43,050. I will explain how this all works in just a second.
So, let's say an organization has, an example organization has 3,500 incidents a month, and if we went back to the Gartner rules, we would say that 20% to 35% of all of their calls would be associated with a password reset. And obviously 25% of 3,500 is more than 875, but what we have found in real-life implementations, is that most organizations have something besides Active Directory or single sign-on type of applications. Health care institutes are a good example, so hospitals have all kinds of applications out there. So, while they may have a 25% password reset concern, maybe only 10% or 12% or 15% of those can be actually be handled via single sign-on or Active Directory. And so what we ship out of the box is an Active Directory password reset tool, right? So not all that specific application in oncology may not be associated to the single sign-on, so I cannot reset that password reset. So a 25% reset need might actually, for active directory, might represent, you know, a 12% or 15% effective password reset adoption rate.
Now, you can use the IVR to reset some of those other passwords but that would just require additional professional services and the point that I've been making throughout this presentation is that we have that sweet spot. We've already built all of those applications so I don't want to show value or ROI on something that we're not out-of-the-box ready to deliver.
So, let's go through the numbers. An average cost per incident was $27.60 according to the Help Desk Institute. A self service incident, fully-burdened, would be $13.50. If you subtract those two, you would see an incident savings of $14.10 or, in other words, every time I can take an incident and stop an agent from actually taking that telephone call and move that into IVR voice self service or web self service or something like that, the one thing I would caution you on is looking to email as a savior in terms of driving self service. So, web self service, voice self service, we can do call-deflection with, that actually is deflecting the call, creating the ticket, resolving the issue, all of those kinds of things. When someone sends an email in, if you look, if you study the HDI numbers carefully you'll notice that for every email received then it generates two and a half times more calls. So, I got an email, yes, it created an incident for me but I may need to do some follow-up on that. I may need to do some other things associated with that so I don't necessarily get a strict ROI from email like I do from voice self service or web self service. Email is great, and it's a great way to receive an incident, but it doesn't really help you finalize and close out that incident or do what we call, call-deflection. Okay, sorry for that little interlude there.
But, so 875 would represent around 13% of the 3,500 incidents that are being received and if you multiply that times a cost per incident of $14.10, that would give you an annual organizational or opportunity cost savings of $43,000.50. So, in theory you could say, "I could deploy HEAT and HEAT Voice self service and in the first year, if I were in the Central or Eastern part of the United States, I could pay for that in nine months to one year." It would have a nine month to one year payback, and that's true. But, I'm not sure again that that's what should be the focus. The focus should be that person that I now have available, after I've gone through my de-staffing programs back in 2008 and 2009, this critical resource that I still have available to me, I am now going to use to have them work with people, talking to people and delivering solutions that are more complex and harder to deal with than a password reset or an outage or something like that, okay?
So let's walk through a password reset example. As we worked with mid-size organizations and small organizations, they've readily adopted voice password reset because it made such a tremendous difference to their business immediately, it freed up people at the help desk and so forth. But as we worked with large, enterprise organizations, we found security concerns to always be a real stickler and so, while we had the ability to identify, authenticate, and validate, what they really wanted to do was do that in a multi-factor way. And so what we've come up with, and there's a guy named Errol who works for K Force down in Tampa, Florida, and he's just a genius of a guy, he came up with a great solution to provide multi-factor authentication using PasswordReset. So, I'll talk to you about it right now.
So a call comes in and hits the voice Self Service, as you see there, and via the ANI or the Caller ID, or maybe some information that they've entered into, like their employee ID or some bit of information, we're able to identify the caller. So, you see the phone on the left, making a call in, hitting the Voice IVR Self Service, and on the right hand side we see the HEAT screen that contains the data that's associated with the information of that employee.
So, second thing that...so that would be identification, the second thing we want to do is to authenticate who that caller is. So traditionally, organizations that we've worked with have done something date of birth day, date of birth month, and sometimes they'll use the last four digits of their SSN. So, between those three we have been fairly successful at authenticating who the employee is there. We can sometimes identify that employee through that method. We work with Ronstadt, which is a temporary staffing company, and I think they had like 25 million records out there and they would put in date of birth day, date of birth month, and last four SSN and still have, I don't know, 20 or 30 people that they would match up with out of the 25 million. So, it's important to have some form of identification first and then go through to your authentication.
Now, the genius that Errol brought up from K Force recently, was the validation. And so, what we want is 2 Factor for validation, now 2 Factor means, or multi-factor would mean, piece of information we might both know, and then there's only one piece of information that you, the caller, would know, right? So, as the caller calls in and we've identified who they are, we've authenticated who they are, we now read back to them a PIN. And the PIN needs to be unique, right? And it needs to be randomly, in a sense, randomly generated. And what Errol came up with is just genius, was, it's the incident number that's associated with the ticket that was created. So as the call is coming in, we generate a new incident, it's not strictly random but it would be pretty difficult to figure out what the next ticket is that's gonna be coming back for the password reset. So we then generate the ticket, we capture that ticket as a property and we read that ticket number back to the caller and that becomes the PIN. We never tell the callers that that's the PIN, we just tell them that we've generated a new PIN for them. So we then say to them, "Hang up and we're gonna call you back," and that's what we do.
So they hang up, we look inside the HEAT database, and we call the user that hits called back and we actually look inside the database itself and make sure that we're calling a number that's in the database, I can't say that enough. So, we can actually get, we can, we could have actually asked them to input the phone number we would like them to call them back at, I believe that's not quite as secure, so I would rather have their cell phone number and/or their work phone number or home number in the HEAT database and then we would call them back at that number. And when we call them back, what we do is we ask them to input the PIN that was just created, which is a very close to random number which is associated with the incident that they created. So they then input the PIN that's back into them, we receive the PIN, we generate the password reset and then complete the call at that time. And we then take that HEAT ticket or incident that was open, we set it in a closed condition and we, via the API, reset the password and we close out the ticket.
Now, we follow the same rules for the password reset that you might find in Active Directory. We have complex, case sensitive, you can have different types of digits. Now, this actual password is randomly generated, so there's a generation tool that's picking up and generating a new password for the caller. Or if an organization decides that for the month of July they want the word ACME to be the reset password for everybody, you can do that as well. I don't remember any organization that's doing that, but that is a capacity of the tool as well. But we generate a password and read that back to the caller and then they call in. We can also require a change of password on first login as well.
So, I want to thank you for joining that presentation today. The items that I wanted to rehearse to you very quickly are Providing Service Automation where we do call-deflection where we're providing the ability to take what would normally have...the complete solution that would be driven by an agent, and dropping that down into self service through creating the ticket, resolving the problem, doing an announcement, and all of that can provide you a very structured ROI that your CFO can believe in, and it can pay the salaries of one or two people at the desks just with password reset, and we can go into outages and all those other things for IT support.
But beyond that, I believe it gives you a greater opportunity than saving $43,000 at the help desk; I believe the value is millions of dollars that can drop to the bottom line in terms of revenue and in terms of improved services and in terms of people actually helping people. One of the hits that I get all the time as we start to talk to organizations is, "I don't really want my people talking to an IVR, I don't want that automation portion in there." And you know what, I completely agree with that. That's really not what this is about. What we're trying to do is take rote and redundant activities, move those off automation because those can be handled, literally better by automation than they can by people. Because if somebody's gonna call me and ask to password reset and I get into a conversation with them, I could have, what was gonna be a two-minute call, I could turn that into 15 or 20 minutes. The IVR's not gonna do that. The IVR is gonna get to business, it's gonna reset the password, and it'll allow that person to get back to work. It's gonna take the information about the outage and it's gonna post it and read the person back the workaround, whatever the form of automation is that you want to provide.
But we can show the CFOs that real savings from the automation standpoint, but then provide on the back-end, you know, like, the real savings to the business, which is getting the payroll out on time, getting the sales folks back up and working, providing those types of services which can be more impactful and mean millions of dollars to the bottom line versus a few cents of savings at the help desk.
We have monthly presentations up on our...that are going on in terms of demos that we're posting up on the HEAT page on the site that you, on the URL that you see here. We encourage you to go look at it and encourage you to study some ROI activities that we have going there as well, and would be happy to provide you some information in terms of white papers on ROI and other types of information that could be meaningful.
We appreciate your taking the time today, thank you very much.