Human beings are flawed – we’re prone to subjective evaluations and in the world of customer service this is a problem. It’s really a problem when traditional monitoring methods within our contact centres are reliant on these judgements, costing time and money. Luckily, speech analytics has emerged to unearth the opportunities previously lost in uncharted conversations – but it’s crucial to use the right approach for your business requirements.
Whether your contact centre focuses on sales, debt collection, or providing customer service, monitoring and analysing calls is vital when it comes to assessing trends and reacting effectively. Traditional quality management methods are labour intensive, requiring supervisors and quality analysts to listen to calls, which is expensive. These costs prohibit most centres from monitoring more than around two percent of the calls they handle. Meanwhile, subjective evaluation criteria lead to disagreements about evaluation results among both evaluators and agents and make objective measurement difficult.
To really carve out those nuggets of opportunity, we need to go beyond personal judgements on whether an agent remained ‘focused’, or ‘attentive to the customer’. Automating the quality management process using speech analytics is one way forward-facing businesses are able to truly tighten up their workforce performance and customer experience delivery.
Making the switch
By switching to speech analytics, businesses can:
- Categorise calls according to every topic discussed within each conversation
- Automatically discover emerging trends and unexpected events within conversations
- Search for words or phrases spoken during calls
Gone are the days when managers, supervisors and quality analysts are frantically trying to identify calls which are relevant to evaluate for quality purposes. Once conversations are categorised by topic, those topics can be directly correlated to KPIs using speech analytics, exposing the root cause of performance issues and the drivers of success.
Great – but how does it work?
Well, there are three types of technology available but all are not equal:
- Phonetics – Phonemes are the smallest unit of speech and individually, are meaningless. Phonetics is good for searching for rare occurrences within large volumes of audio, but it isn’t as reliable for normal quality monitoring needs, and users always need to tell the application what to look for.
- Speech to text – Transcribing speech into text is more accurate than phonetics. As the output is text, text analytics can then be used to identify emerging trends and unexpected events that users may not have known to look for.
- Speech to phrase – Entire phrases are detected within the audio itself without first being converted to text or phonemes. This is the most reliable technology for detecting phrases. Any topics that are important to the business to measure and monitor on an ongoing basis, such as agent skills, must be defined and detected as phrases, not merely keywords.
Despite the obvious difference in technologies, and clear disadvantages, the most common use cases within interaction analytics are Phonetics and Speech to Text. Speech to Phrase capabilities are crucial and the only vendor offering Speech to Phrase capabilities is Genesys.
Despite the obvious difference in technologies, and clear disadvantages, most speech analytics vendors offer either Phonetics or Speech to Text; or a combination of the two. Speech to Phrase capabilities are crucial for automating your quality management process and the only vendor offering Speech to Phrase capabilities is Genesys.
The second largest Internet Service Provider in Australia deployed Genesys Speech Analytics to conquer high average handle times and understand the intricate nuances and purposes of conversations taking place. Like most organisations using traditional quality management methods, their quality team was only manually evaluating less than one percent of their calls. With speech analytics automatically evaluating 100 percent of their calls, their quality team was able to transition to selective reviewing of what are automatically flagged as important calls to evaluate further, and recommending actions such as targeted agent training. This resulted in a seven percent reduction in average handle time across all their support call centres ―a savings of over two million pounds. “Every extra second we spend on the phone costs us money, so speech analytics starts exponentially saving us money,” said the Chief Information Officer of the Internet Service Provider.
To learn how you can achieve similar improvements by leveraging speech analytics to revolutionise your organisation’s quality management process, please read this White Paper.