Call and contact centres need to continuously remain up-to-date with the cutting-edge in IVR systems. Old DTMF IVR simply will not cut it anymore. In this article, we talk about AI-powered IVR systems. If you want your call centre to remain effective, read our article below about why you need to make the switch.
The problems with legacy IVR systems
Traditional DTMF IVRs that are based on a menu tree structure becomes less and less effective the larger and more complex they are. Think of an IVR with eight different options. Most people will struggle to remember the first options by the time the eighth is played to them. There is a lot of wasted time; listening to the prompts, fumbling with your phone to find the keypad, being slowly funnelled into the correct submenu ready for the next onslaught, having to call back to restart the process. Most IVRs simply funnel callers, and the IVR system has no idea what the callers objective is. Getting these IVRs to complete customer service transactions is time-consuming and frustrating for both the callers and administrators.
Speech-driven traditional IVRs simplify the user experience by recognising specific keywords using automated speech recognition (ASR). The user can then be asked a broad question such as “How can I help you?” and assuming every possible answer or keyword has been pre-loaded, the caller can be either directed to the correct person or the IVR can complete the transaction faster. Furthermore, they don’t have to fumble around with their phone keypad or stress their brain remembering the various options. ASR systems can also process alpha-numeric caller responses which DTMF systems can’t. While the user experience is better very often the system has little intelligence and is just a funnelling system that uses speech instead of keypresses (DTMF).
Traditional IVR systems are limited therefore in what they can do by their complexity which is why most of them simply route calls to the correct department rather than providing solutions to problems.
What is an AI Powered IVR?
Interactive voice response (IVR) is still the correct term for a system which uses AI and understands natural language (spoken) user inputs on a phone call and provides solutions or assistance. The underlying technology is the same as is used by customer service chatbots albeit with ASR, so the caller can essentially speak to the chatbot. The main advantage that this type of system has is being able to recognise the callers intent or objective from the outset using natural language processing (NLP). NLP allows the caller or chatter to communicate as though they were talking to a person rather than answering a long series of tiresome funnelling questions.
An example conversation with an NLP chatbot might go as follows:
Bot: “Hello, how can I help you today?”
Caller: “I’d like to buy two air tickets from London to New York on June 21st in the morning, please.”
Bot: “Of course, you like to buy two seats from London to New York on the 21st of June, we have a flight at 10:30 AM, would you like to book this now?"
Caller: "Yes."
The bot immediately understands the callers objective using keyword analysis or natural language processing. This example transaction not easy using a DTMF system and is rarely attempted.
As well as understanding the callers' intent quickly, AI powered IVRs have some other advantages:
Main benefits of an AI IVR
- A Better User Experience
By understanding the callers intent quickly the bot can provide a slicker customer service experience. Callers will prefer to use the AI IVR because it can provide solutions 24 hours a day in a conversational manner. 80% of everyday transactions for most organisations can be handled by phone or chatbot, the reduces the need for staff and provides service users with a quicker resolutions time.
- The Same AI Bot Across Multiple Channels
The same bot can provide support through multiple customer channels such as an AI IVR, website chatbots, and social media. This means that changes and improvements will instantly be replicated across all channels cutting down administration time, errors, duplication and cost. This also means that the customer service experience is consistent across the “omni-channel’ contact centre.
- Easier Setup and Programming
Chatbots generally use simple programming languages are therefore reasonably easy to develop, in terms of adding questions and answers to. There is no need for a software developer or highly skilled technical person to remain beyond the initial set up of the system; it can be done by anyone via a simple admin interface. This reduces the long term operational cost of the system and means it is more likely to improve quickly without the constraints of skilled resources.
- Real-time Visibility
Customer interactions are transcribed real-time in both chatbot and phone bot scenarios. This is useful as the phone call conversations can be monitored and analysed by supervisors who can quickly respond if the caller is struggling or becoming frustrated. If necessary an agent can step in a take over from the bot seamlessly to resolve the problem.
- Easy integration
AI systems and chatbots generally come with a full set of web APIs which can perform tasks such as authentication, providing confidential information, retrieving information from any database, sending SMS & emails. While the programming requires skilled people to code, the integration methods are based on widely used programming languages such as PHP and javascript, as such development costs are low.
- Potential for Machine Learning
Once set up the bot can be integrated with a machine learning system such as TensorFlow which will typically require an extensive database of customer interaction information from which it can formulate intelligent responses automatically. This requires a great deal of skill to set up correctly but promises to almost eradicate administration and the need to add improvements manually and continuously.
Tree structure IVRs are used by almost all organisations today and most of them are used to steer callers to an under-resourced department. AI bot IVRs can be easily inserted into these organisations to provide instant impact, enabling the organisations to provide service digitally, 24 hours a day freeing them from legacy constraints.
