Humans Vs Technology QA in Call Centers – 17 Great Points to Consider For QA
- Contact Center
- BY Eddie
- No Comments
Over the past couple of months or so, we have looked at a few articles comparing human analytics and voice technology QA. In this article, we are going to take another look at how each of them competes when they are in more complex environments. Before we get into the article in-depth, I would like
Over the past couple of months or so, we have looked at a few articles comparing human analytics and voice technology QA. In this article, we are going to take another look at how each of them competes when they are in more complex environments.
Before we get into the article in-depth, I would like you to ask yourself one question: How many calls have you seen that are easy to score? My assumption is going to be very few.
Having answered that question, let’s take another look into the ways that Voice Technology works. (You can find one of our other articles about this here.)
Voice Technology for Quality Assurance
As we have previously said, voice technology has evolved immensely over the years, and we would be silly to think otherwise. Technology in calls started as a basic ability to have a recorded voice asking questions, and the caller requiring to press keys to answer. This technology is still active, as I am sure you are aware.
However, it has now evolved into something even more. Speech analytics has become an almost integral part of call centers to analyze key performance indicators or KPI’s. The way that voice technology does this is very simple in theory:
- Distinguishing words. – Voice technology has become very good at identifying certain, pre-programmed words. This identification is the complete basis of the system.
- Marking those words. – If the AI hears a name, for example, it will signify that as a point onto a scorecard. If it is not recognized, then it will mark it as a point lost. Of course, names are just a simple example here. A more complicated case could be phone numbers. If the voice technology hears numbers, there are a few different things that they could be: Date of birth, zip code, or anything else that contains numbers.
Of course, those words can be anything that you like, and they often are. It seems like a very alluring system to incorporate into your call center, doesn’t it? It is almost a “Set and Forget” system that only requires an initial set-up.
However, as with everything, there are downfalls to it.
Voice Technology Downfalls
Several downfalls come with voice technology, but they often get overlooked. The reason for the overlooking is because of the benefits that speech analytics give you. I will not go into the benefits in great detail in this article, as you can read about them here.
In this article, we are going to focus on the downfalls of voice technology. We have touched on these downfalls in other articles that we have written. However, we feel like it is time that you understood them entirely so that you can make informed decisions:
The Expense of Voice Technology
This point is the first thing that you may think of when you are looking at the downfalls of anything. Buying a car, for example. Who doesn’t want the newest and best Mercedes Benz? But that car comes at a high price to your personal income. But, let’s say, for argument’s sake, that you have the spare money to buy it. Brilliant.
Two years down the road, and you now do not have the newest or best car that money can buy. This fact is going to leave you wanting again. The same theory applies for your call centers voice technology.
As I said in one of our previous articles, voice technology is not cheap to develop. Therefore, it stands to reason that it is not going to be low-cost to acquire. This cost is not usually just an initial cost either. Depending on call volumes and recording/monitoring times, you can also be charged monthly “upkeep” costs and a cost per minute. That cost can become an enormous expense, for technology that will be outdated in less time than a car.
To put that into context, take a look at how mobile phone cameras have developed over the last five years. As a quick comparison, the best phone of 2015 was the Galaxy s6 edge. It had a considerable recording rate of 120 frames per second on its rear camera, and a front camera of five megapixels. Today, the best phone is arguably the Note 10 Plus. This phone has an incredible 960 frames per second record rate on its three rear cameras and a ten-megapixel front camera.
You can see by that one paragraph; technology is improving faster than anyone has ever imagined. Buying Voice technology for your company could quickly leave you wanting better equipment in less than five years.
The Complexity of Voice Technology
Next on the list is the complexity of voice technology. We will use the same analogy as above, mobile phone cameras. Even they get more and more complicated as time goes by. How many times have you searched for “How to use the camera on my phone” or “what is iso on my camera”? Probably more times than you are willing to admit.
Mobile phones, however, do not lose your companies money if you do not know how to take great photos on it. Voice technology and the software that goes with it is infinitely more complicated. To top it off, it will lose a lot of your companies money if you get it wrong or take months to work it out. Even when you have worked it out, you will need staff that can deal with it.
The staff that you employ for this task will likely require training and monitoring that they are doing it correctly. This training takes even more money and time out of your business.
Uncertainties of Voice Analytics
Can you be one hundred percent sure that you have everything set up correctly, and that the analytics are doing their jobs properly? The answer to that is probably no. You will have teething problems, and they could go on for months, or even years. This fact is especially true when you are migrating from one software to another.
Take this example: You have just got used to your software when you find out that compliance regulations have changed. You have been thinking about new, up to date analytics for a while, and this is the push. You have had a salesman visit you showcasing their latest equipment. They tell you how it complies with all of the new regulations, and it does more than your old equipment. Great, I’ll take it.
All is going well until you realize that it is not monitoring certain aspects that you previously observed. Now you have met the uncertainty. How many of these aspects is it missing? How much is it going to cost me to put it right? You are now paying out more money for more stress.
Voice Technology Overview
It would be wrong for us to say that voice technology has no upsides at all. Of course, it does. You can read about them in the links that we have earlier on in the article. For basic calls and scoring, and generalized work, it may suit you. But how many times have you found yourself talking to someone on the phone, and you can’t understand them for some reason? Maybe the line is terrible, or they have an accent that you can’t quite understand?
If you can’t understand those people, then how can you rely on technology to follow it and score your agents? Especially to do it correctly.
Human Analysis for QA
Now is the point that you may think: “Well, what do I do then?” You have probably tried to monitor calls yourself or even appointed a small team to do it for you. In this section, we will quickly talk about the main benefits and downfalls of human analysis.
The workload of In-house QA (Downfall)
We have covered the pros and cons in more detail here, but let’s take a quick look at the workload of it again:
It is quite self-evident that if you wish to control your own QA team, you are going to be increasing your workload as a business. One company alone, with 100 agents can handle well over one thousand calls per week, maybe even two thousand or more. But for this example, we will take it as a round number of one thousand.
If the average call is ten minutes long, it is likely to take your in-house QA team longer than that to analyze the same call. Even if that number is only by a couple of minutes, so, let’s say that it takes twelve minutes to analyze and score a call. In reality, it is longer, but for this, we will go with twelve for ease.
You now need twenty percent more analysts than agents to monitor and score those calls. That is just not plausible.
The Accuracy of Human Analysis (Benefit)
Of course, people are much more able to complete an in-depth analysis of calls. They can listen to a call more than once to determine factors that get marked, and they can also complete more complex tasks. They are the main benefits. It would be great to have that 6/5 ratio of analysists to agents, but that is, as I have just said, not a realistic expectation.
So, what can you do? That is where Call Criteria step in to help.
What Can Call Criteria Do To Help?
When you have a company solely dedicated to a single task, you are going to get the best results. Call Criteria is THE company dedicated to Quality Assurance monitoring in call centers. We have developed and perfected the QA process that we can incorporate into any situation that you may find yourself in.
Complex situations often require trained and experienced people to deal with them correctly. We have a dedicated team and software that they are entirely competent in using. We stay up to date on all of the rules and regulations in the Call Center industry. Therefore, we update our software and staff accordingly. I have written about how we monitor calls and provide scores in some other articles that you can find here:
- Scorecards. – Our system allows for custom and complex scorecards that fit your requirements. Not only do we allow it, but we also compile and provide them. All of our agents are then trained to use them and how they should be marked. Read about our scorecards here.
- Monitoring Agents. – Monitoring individual agents can be difficult, especially if you are only able to analyze a small percentage of their calls. You may only catch one or two good or bad calls. When you only have the data for such a low rate of calls, you are likely to get incorrect results. That will lead to staff with lower morale, or you are not seeing the poorly performing agents. You can see how we can improve your monitoring in this article.
- Tracking an Agents Performance. – To go along with the monitoring agents section above, you do not just want a snippet of their performance. The best way that you can monitor your agent’s and their performance is over some time. You also want to be able to compare their performance for specific periods to others. For more information on this, see our tracking article.
- Coaching Queues. – When you have your scorecards right, and you are tracking the performance of your staff, how do you help the team that is struggling? In our software, we even provide a coaching queue to help you see who needs coaching, when, and with what. Check out how our coaching queue could help you here.
If you want more information about how we can help you with our dedicated human analytics, then please contact us for a demonstration.