Image by

The Enigma of Speech Analytics

Written by Sina Taavon

Within this age of technology there have been many unprecedented breakthroughs using artificial intelligence and voice recognition such as Alexa, Google assistant, Watson, Siri, Bixby, and the list goes on and on. These technological advancements are gradually becoming the norm within households but why are businesses not adopting this new technology. If it is making households more efficient why would it not do the same for business activities? The clear answer is that it does make businesses more efficient and helps with work flow optimization, speech analytics is a twist on artificial intelligence and voice recognition to suit the needs of organisations. However speech analytics is a relatively new phenomena which has been introduced into technology. But what is this mystery treasure? Speech analytics is an engine primarily used in call centres to convert oral communications into a written format which can be edited understood and more importantly analysed quickly and efficiently.

The main issues with speech analytics which is halting the use within businesses is that it isn’t really understood. However the advantages of speech analytics is that it reduces operating costs, increases revenue, improves productivity and aids in providing quality, consistency and compliance. Firstly, this engine has to be implemented into their servers which is problem number one, the second problem is explaining something like this to your CEO/CFO could be challenging.

Like all technological advancements it takes time to be accepted in to the world and into business, but the astonishing fact is that the benefits of speech analytics are not up for debate, all agree that this engines advantages outweigh its issues one hundred to one. Yet only 26% of call centres use speech analytics. Which is bizarre if you think about it because it is physically impossible for every call recorded to be analysed by even a team within a company, when there is a software which can analyse numerous recorded calls simultaneously which picking out and recording patterns, phrases, and other data which can be used to improve customer’s satisfaction. The important part is that advantages of this technology does not end here. I’m going to explain what you actually can do and the real potential of the software with a real scenario, A travel agency could get all sorts of request in shape of email, social media posts, calls and more with all these new technologies you can categorise all this information into meaningful segment so your marketing team can take advantage of the data e.g. extracting specific keywords from calls, emails and social media posts. You can even run a sentiment analysis on each call, so you can have better understanding of how your agents perform.

The two most common problems businesses encounter:

  • The business wants to use this software but do not know how to implement it.
  • The business knows how to implement the software but don’t know how to use it.

These problems can be encountered at different levels within the company and for many uncertainty and even fear is what is holding businesses back. Because personal data and recording calls is a sensitive topic it is usually best to stick to what you know which is why, so few organisations use speech analytics. Problems like this could be easily fix by a cloud hosted solution like Hermes.

First let’s follow the current path of analytics. A call has been recorded, an analyser would have to sit through and listen to the conversation, picking up on tone of voice, questions asked, phrases etc. and pick out desired traits of the employee which produce desired traits of the customers. It is only natural that this process can miss aspects, human error is a well-known concern.

Now using speech analytics this process is now automatic. A call has been recorded, and just to keep things realistic there isn’t one call being recorded at one time there may be 50, 100, 150 calls being recorded all at the same time in one call centre. Where a human can only analyse one call at a time. The software can analyse every call that has just been recorded. It will systematically covert every word uttered into text which will be recorded. The program will then go through each word, phrase and sentence picking up positive and negative connotations and compiling a script which portrays consists of phrases that have been said which produce desired traits of the customer. The final product will produce essentially a script which employees should follow to improve the calls. Not only this but if a particularly negative call was recorded the system can highlight this to be reviewed by management, this can then be followed up to improve customers satisfaction even further.

Sina, DevOps Engineer Ciptex

Sina joined Ciptex as DevOps engineer and is passionate about programming, music production and creating artwork. He is already a champion of cloud technologies — and In 2016, won an ‘Apps for Good — Sustainable Communities’ award for his App, DonateIT, which helps Charities acquire more second-hand goods by directly connecting them with Donors. Having mastered Handball, winning a gold and two silver medals, he has now turned his attention to running. Sina is fluent in English and Persian.

“I’m really pleased to be working with Ciptex as a DevOps Engineer. Taking this step will be a real game-changer in helping me achieve my life goals and objectives. The degree-level apprenticeship program run by MMU is just great — to be learning while working and achieving a degree is a dream come true for me”.