Voice Bots & NLP
The answer to FCR woes?

What makes Voice Bots so good?

With customer identification, product/service identification, problem identification, acknowledgement and consent all being delivered by the customer verbally, and captured by the bot, by comparison makes the traditional IVR, and all too familiar Agent startup script seem a little clunky.

The Voice Bot removes the heavy and time-consuming human processes and enormously strengthens the integrity of data capture surrounding the reasons behind the enquiry or problem.

For the Contact Centre, each time this happens successfully, it means one call less, and one potential problem less (humans can make a hash of it sometimes!).  For the Customer, their enquiry is still perceived to be resolved on their first attempt (so FCR = YES), and this leads to good CX and a likelihood of retention or repeat purchase.

How does FCR come into play?

So if executed correctly, voice bots should reduce the repeat call rate for common enquiries, but is this where your FCR problems really sit? Probably not.

One would hope that the support team would be more familiar with common call types, so these would mostly be resolved first time anyway. So by removing these from the Contact Centre, whilst undoubtedly easing the workload, this will invariably concentrate more complex calls into the Contact Centre.

So as before, support staff will still very much be dependent on systems, process and 3^rd parties to close cases, so those same escalation, feedback and follow up loops still remain, with the customer often stuck in the middle.

A Drop in FCR % Expected??

With a higher proportion of the calls being handled by staff being more complex in nature, there is a strong likelihood the FCR % in the Contact Centre will actually go down (so be careful when setting FCR % targets to your Digital Transformation Team!).

Remember, the essence of the Contact Centre is to help customers that need information or who have problems. Equipped with easy to access and accurate information about the reason for call, ideally companies will be able to pre-empt future traffic by providing the necessary information proactively, so customers don’t need to call in the first place.

If the bot can help intercept customers and solve some of their problems by a process that can be administered without having to deal with a human, then this will become a game changer, but for the rest, the Contact Centre will always be at the mercy and dependability of the organisation’s products and services, technical infra, and their SOPs as before.

It is also worth being mindful that if bots fail to resolve specific enquiries or problems, they need to be able to offload the transaction to a human being immediately, to be able to still classify as a FCR (same transaction).

If poorly designed, and the process sends customers down the wrong path, or hits a wall, forcing them to hang-up, or worse still, disconnecting them, this strategy can backfire spectacularly. This will leave customers feeling frustrated and holding a negative perception; that the organization is deliberately trying to get rid of them by using robots, rather than using them to actually improve their experience.

Voice bots versus Humans in Problem solving & CX

Customers usually know that their problems are not unique, and don’t like to have to explain or justify themselves and rely on other people. They just want to get their problems resolved as quickly and painlessly as possible.

And the robot voice can weirdly be quite calming and reassuring, and so long as it is functional and helping customers, it will probably deliver a better CX than humans can achieve.

So beyond basic transactions, to what extent can a robot solve a complex human problem? The opportunities for voice bot automation in problem solving are almost endless, and limited only by the imagination of CX Digital Transformers and the technical capability of existing systems.

For now, although often referred to as such, this technology is not strictly Artificial Intelligence, as the robot is simply following a process that has been pre-prepared by a human being.

So how do we tackle the Non-FCR cases that fall through the Voice bot net?

When it comes to the isolation of Non-FCR hotspots, you’re still going to need visibility mechanisms to identify the customer profiles affected, which products and services are involved, who is contributing to the repeat traffic, and the reasons why, in order to make any significant inroads.

Ongoing tracking of your current enquiry types or reason codes across all human support channels from the data captured in the voice bot, as well as from the CRM, and close monitoring of the success of your existing processes being handled by the bot, will help identify customer pain points and specific opportunities for further automation, and FCR improvement areas.

Comprehensive FCR reporting across all channels needs to be a top priority in your Business Intelligence toolkit, as it holds the strongest opportunity to drive significant change, to reduce wastage and improve CX. This will give you the best possible chance to ensure the mechanisms on the back end are well oiled and improving in parallel.

In a similar way to legacvy FCR tracking, compared to the post call survey, the method yielding the highest integrity is to track for repeat customer activity systematically. Although now, with vital customer data capture taking place ahead of the call,  this will serve to strengthen the integrity of the FCR reporting, compared with CRM Reason Codes and Ticket tracking which have historically been heavily dependent on manual data capture.

In Summary

Voice bots will surely transform the landscape of Contact Centres as we know them, and provide a refreshing new interface for customers to take advantage of, navigating almost effortlessly simply by using their voice, to state what they want to enquire about, without the hassle of keying in data or selecting from a limited menu.

For the more complex calls being handled by the Contact Centre staff, this won’t necessarily improve FCR %, or make your Non-FCR reporting and investigations any easier, but it will certainly improve the accuracy and dependability of your Non-FCR data.

And if you really want to measure your Customer CX and get a true reading of your FCR using NLP, since it is all automated anyway, why not put in a simple, but hugely powerful question into the voice bot up-front… “Is this the first time you are contacting us with this enquiry?”