Automation in customer service has seen rapid development in recent years. This is true for IVR solutions/voice portals as well as for the comparatively young, automated chat and messenger dialogs. Therefore, it is all the more important for companies to fully exploit the potential of solutions already in use or to further optimize them. In the following article, CreaLog Key Account Manager Stefan Riesel reveals what is needed and which tools can be used for efficient spring cleaning, especially in telephone customer service:
What exactly does the customer want, and can they quickly reach the desired goal via automated dialog? In customer service, are you already recording real customer conversations and chat histories for subsequent, AI-supported analysis? If not, you should probably modernize this area right now. After all, these analyses are the only way to effectively determine which different terms and phrases customers use for identical issues. You can also see whether and how dialogs with existing customers differ from those with new customers. In addition, you can identify whether the reasons for calling have changed significantly over the years. If that is the case, the dialogs will of course have to be adapted.
If customers end the call without having achieved anything, it may also be due to the dialogs. It is not only important what is said (content), but also how it is expressed (form). Pick up the phone yourself and ask yourself the following questions: Are wording and tonality still up to date? Does the speaker's voice still fit the corporate image and is it motivating? Is the sound quality still optimal from today's perspective? Often, the company's own employees have recorded the texts, but unfortunately it often sounds exactly like that. Especially in this situation, it is worthwhile to commission professional speakers to produce newly phrased announcements. Compared to the positive effect on the callers, the costs incurred are low. If your announcements are dynamic or change frequently, for example due to seasonal promotions or information that needs to be updated frequently, it may make sense to switch from static pre-produced texts to dynamic TTS (Text-to-Speech or speech synthesis). The quality here is now much better than it was a few years ago, as you can easily discover on your smart speaker. Read more at: https://www.crealog.com/en/products-solutions/one-number-voice-portal/
Friendly and clearly understandable announcements (not for you, for the callers!) in top sound quality can already reduce the number of dropouts. In addition, the fewer choices you offer in the dialog, the less confusion for callers and the lower the dropout rate. Basically, it's worth determining how many callers actually go through the paths you provide. Are there frequent detours or do callers even move in circles? If so, there is a risk of abandonment or more calls are forwarded to the service center than originally planned, which can significantly worsen the defined KPIs. Do your dialogs already fully master free text or voice input, which gets customers to their destination much faster, especially for more complex requests? If not, now is the time to bring the voice portal up to the current state of the art (gold standard). Then no longer anyone does have to listen to endless choices or "wade" through multi-level menus. Just let callers freely answer the question "How can I help you?"!
For self-service applications, it pays to pre-qualify or even legitimize the customer. Your CRM can then support dialogs with customers and provide the VoiceBot with important customer and contract data for optimal, automated processing of the request. Secure caller authentication also speeds up service center processes. Valuable or particularly important customers and escalated tickets can then be given priority. Equally important: If callers are already identified by the VoiceBot, the data on the customer and the customer history can be provided to the agents even before the call is transferred. A prerequisite for this is the coupling of telephony with the company's CRM systems. Have these IT systems been changed, modernized or perhaps even outsourced to the cloud recently? If so, it should be immediately checked whether all systems still harmonize smoothly with each other and whether the data is actually exchanged flawlessly.
The essential prerequisites for a powerful conversational AI solution include innovative technologies in the area of speech recognition and speech processing, as well as fast access to different data sources via APIs. Only this way can an AI-supported customer dialog ensure that the appropriate answers are found for almost all inquiries. Which then also particularly satisfies the customer/end user: Around the clock, she/he can, for example, use a VoiceBot or ChatBot to start inquiries, place orders, check delivery times, and use many other services. Today, modern speech recognition (NLU: Natural Language Understanding) enables not only the recognition of individual words, but also of entire sentence contexts. This means that even complex questions posed by the caller can be correctly interpreted.
So how many intents still require a human employee to intervene and (re)process them? Which voice dialogs are structured in a way that the call still has to be transferred to a service employee for final resolution? Is it worth automating further processes? Our tip: Take the time to go through all the use cases with the CreaLog experts and discuss with them which of these can now be handled reliably and with case completion by a VoiceBot. Read more at: https://www.crealog.com/en/professional-services/
Many companies that started their self-service strategy with a Voice Portal now also use ChatBots and Messenger services for automated customer communication. These contact channels can provide customers with the same information and use the same data. This has been made possible by modern, AI-supported omnichannel solutions that use a uniform database for voice and text services, so that the customer experiences the same quality of service seamlessly on all communication channels and around the clock. Here, too, it should be checked whether this was considered during implementation and whether voice and chat bots can actually access the identical database - or whether content has to be maintained multiple times.
Have our tips given you ideas for your spring cleaning?
For even more support, simply get in contact with us.