03/02/2021
Our CEO shares in his words the advantages of speech recognition and its impact to CCCS.
Modern companies offer numerous ways to communicate; Phone, email, chat-bots. But to be effective it is crucial to use these technologies to your advantage.
For companies with dedicated sales, collections, and support teams, call center software has been a very popular means to communicate for good reason.
What do you expect to see in the large call center? Hundreds of agents, huge area like several basketball courts and a very noisy place. But can you imagine it be a quite and cold room of just 5 square meters with only the noise from cooling fans and many small lamps are blinking? This is how a call center with the capacity similar to 1000+ agent call center capacity may look like with speech synthesis (text-to-speech) and speech analysis (speech-to-text) along with speech recognition function (AI).
Speech recognition software is one of the emerging technologies that a company can take advantage of; some of the big companies such as Google and Amazon offer Automatic Speech Recognition software that is now available not just for existing big businesses but also for startups that can be used to improve your call centers’ efficiency.
The functions of an Automatic Speech Recognition Software combined with tailored fit call flow gives you the ability to maximize your company’s ability to cater to your customers. It can be different types of automation:
• An AI based robot that utilizes all three services: text-to-speech, speech-to-text and speech recognition:
• A simpler form, but in ways equally efficient – a combination of text-to-speech, speech-to-text and a call flow with set of keywords that will help the robot to “understand” what a customer say and to answer accordingly.
Let’s talk about the second version of the Robot, Building the perfect application of the Robot is a continuous process but can be done in as easy as one, two, three.
After choosing the supplier of the service you will need to integrate this cloud-based service to your existing telephony system, or you can find a provider that offers a call-center system already integrated with either Amazon, Google, or Microsoft cloud services.
Next, identifying key words that will direct the call flow of the conversation is to be identified by your team in order to proceed to the next step of creating a call flow.
Creating a call flow for your business or campaign can be as simple as 3 steps or multiple steps that will only be limited by your requirements. Since you have full control on how the robot will handle keywords on the conversation, you can set-up a full conversation using pure robot but there is also an option for you to combined both robot and human operator.
One of the key operations needs on implementing a robot system on your business will be the Quality control team. Just like in a regular QA team responsibility, they are tasks to check call quality and result regularly for phone operators. Within this process they are checking for call handling, call behavior, accuracy, proper tagging of the call, and other aspects of the call. A similar process is still needed when checking for calls that are handled by the robot, may it be a partial or a full robot call.
Let us take a collection call as an example; a QA can rate a robot call under 3 main categories of mistakes.
1. CRITICAL MISTAKE – a call that ended with a successful answer from a customer (like an offer to accept or promise to pay from a customer, which we will consider as a “yes answer”) but is incorrectly understood by the robot as a successful. This result will greatly affect future interactions to a customer and might possibly result on a complaint from the customer, thus the setting of the robot needs to be adjusted in order to rectify this problem on future calls.
2. MODERATE MISTAKE – a call wasn’t tagged as a yes answer but there was a clear statement of the intent from the customer. This result is considered as a moderate mistake as it will not negatively affect the customer but still needs to be corrected as it skews the expectancies that we have in terms of the performance of the robot campaign.
3. OTHERS – are calls with minimal errors such as improper call recognition.
This makes the work of the QA team very crucial in the maintenance and development of the robot system in your company. The use of the technology can bring more positive results as long as the process is being improved continually with the help of the QA team.
In conclusion, the availability of this technology provides businesses; no matter the size and the nature, a chance to improve the quality of the services they offer to their clients and in turn improving effectiveness in terms of performance and cost in the process. You no longer need to invest millions and spend years of implementation.
CCCS has started to embrace this technology using a call-flow Robot from Google that has the ability to speak in English and as well as in Tagalog. We have already automated 20% of our outbound calls that reduced our costs by 15%, we are planning to increase the share of the robot calls to 80% of our calls.
