Real Time Speech & Text Analytics – does it really stack up?
There is a lot of focus in the market place on real time analytics and how your contact centre should be leveraging real-time speech and text analytics. Having instant access to data and analytics in real-time for your agents, supervisors, operations team, etc seems to be on everyone’s wish list, but it is important to critically ask ourselves – do we really need it?. What so often gets overlooked is that having information in real-time is one thing, but if the organisation can’t make a decision and then implement tangible operational change in the moment, the value of real-time analytics is lost. In fact, often the true value exists in using this data to make decisions, not in real-time but post the event as a far richer analysis can be undertaken.
Like many Australians I love following my sports and it is interesting to draw a parallel. The majority of professional sports teams have all their players wearing GPS trackers so that all the data related to the distance on the field is tracked and available. This data can be available real-time to the coaches in the box, however that data is not where the true value is derived. More of use is how they use this data to improve the performance of individuals and the team post game or post practice session. Let’s take a specific example.
In sport nothing can derail a successful season quicker than a few key injuries. There is a cost to the team, and for the athlete, any time they are out is development time they are missing. For both coaches and athletes, it is crucial to minimise the risk of injuries as much as possible. With metrics from their GPS performance monitoring system, coaches are able to use benchmarks they’ve established to determine when athletes are at a higher risk of injury. They can then tailor rest or training based on the athlete’s specific needs, and even prevent injuries before they occur. Detailed and real-time data – but this is leveraged post the event and allows a much richer assessment of the available data across many facets to make a considered decision and course of action.
This is the same in the Contact Centre where companies often perform Speech and Text Analytics. Real time does offer some opportunities but the low hanging fruit is using this rich real-time data that is analysed post interaction to derive actionable insights to drive better agent and team performance.
In the contact centre using speech and text analytics in real-time to determine when a potential cross sell opportunity may exist and to flash this on the screen to an agent seems compelling at first. Or perhaps to prompt a supervisor if an agent’s conversation breaches compliance and allow them to barge in is also touted by some as a real-time use case. However far more powerful and valuable is to leverage this data to determine whether agents are performing to the level we need them to, compare and contrast performance between agents, and identify specific training that can be put in place at an agent level to make a difference in future conversations. It is about making considered decisions using the detailed real-time data available at a pace and time that makes sense for the organisation.
There is no doubt very specific use cases exist that require a real-time decision to be made in minutes – for example bomb threats or similar emergency scenarios. However, from our experience the vast majority of value for most organisations is derived by analysing a broader and richer data set to provide insights to a key decision post the event.
It is actually more about the processes you have in place and the ability of the organisation to action insights rather than having software do something for you in real-time. Would you rather get real-time information that is 40% accurate instantly or wait 24 hours for information that is 95% accurate with insights and actionable.
It all comes back to what Use Cases or Business Problems are you trying to solve? Contact Centres are one of the most monitored environments in business, so adding real-time analytics without solving a problem or adding value … what’s the point. Just because you can, doesn’t mean you should. Focus on the business problems and then how to solve them in terms of processes, real-time decisioning and actions. Then you can define whether you need data on your interactions in real time, hours, next day or a longer timeframe. Once this is defined, then look to how technology can play a supporting role in achieving your business objectives.
