Companies worldwide are seeing an increase in the volume of interactions and changes in customer behaviour. Both changes force companies to make choices. What choice do companies make? During the Verint event 'Data-driven Customer Contact' (8 June, with approximately 100 participants), Martin Melis provided insight into the results of recent research in this area. In addition, both Nationale-Nederlanden (NN) and VodafoneZiggo shared their experiences about using data to improve processes.

 

In the latest survey (which also involved Dutch respondents), Verint investigated what decision-makers (N=2.742) did with their insights from last year's survey. About the results wrote Ziptone before; in his presentation, Melis elaborated on current issues. The decision makers surveyed see three key challenges: delivering and sustaining high-quality customer experience during disruptions (such as in the supply chain, think of the chip industry), delivering a complete digital experience and responding to the talent shortage. Setting up hybrid customer contact (people and bots) is also high on the agenda.

The Netherlands: especially busy with hybrid working in customer contact

While 'The Great Resignation' is playing out worldwide, the Netherlands is mainly busy managing home workers, according to the research. Many companies are still looking for a new balance; for the time being, WFM is a hot item in addition to Quality and Performance due to the labor market shortage and the after-effects of corona. The Netherlands scores well on digitization – we are a front runner compared to many other countries. This leading position raises the question of what the next step will be for Dutch companies.

There was also room for practical stories. Both NN and VodafoneZiggo gave a look behind the scenes in remarkably candid presentations. NN showed how the bank-insurer optimized two relatively confusing back-office processes with Operations Manager; VodafoneZiggo talked about the implementation of Speech Analytics from Verint.

Nationale-Nederlanden optimizes back office processes

Verint supported two business units of Nationale-Nederlanden with the implementation of Operations Manager: Income Insurance (130 FTE in a mixed front and back office, 200.000 administrative requests per year) and Non-Life Insurance (300 FTE for front and back office and 400.000 administrative requests per year). years), both primarily aimed at intermediaries. The business units both work with small specialized teams that have a large number of skills, but which are not always clearly defined. This makes responding to peaks and troughs in both volumes and in the SLAs to be achieved a vulnerable process.

Both business units have therefore again mapped out what the work supply actually entailed. They have put skills in a matrix and involved employees in the desired change: a more efficient organization with better predictability. The way in which the work reaches the employees is partly automated. Team leaders monitor productivity and can adjust the way in which the robot distributes the work. Processes are further categorized in priority based on customer expectations and impact: a move is different from a benefit.

Working systematically is a process of change

The business units have also introduced work schedules. Berber Harkema (business optimization manager at Schade Intermediair) indicates: “Each department gives its own interpretation to this and the experience of employees also differs. It reduces the pressure, and within teams that distribute the work themselves, there is a higher involvement in the process. The result: an almost completely predictable process, more insight into the duration of tasks and therefore more peace of mind for the teams. Collaboration with other departments has also improved. This allows IT to plan changes better, because data shows where the peaks are.” Yuri Min (Manager Control, Change & Data, Nationale-Nederlanden) adds: “Work inventories are decreasing, so there is less repeat traffic, the SLAs are achieved, and the NPS is rising.”

Implementation tips

Nationale-Nederlanden also has tips for the Verint public. According to Min, communication is the most important thing. Rationalization of back office processes is a major change. Therefore, ensure support from employees (why are we doing this?) and support from senior management. In the implementation of Operations Manager at Nationale-Nederlanden, various forms and moments were used: from demos, soapbox sessions, newsletters, super-users to periodic rounds along the fields and an escalation line. Min also sees the added value of an intermediary in the Netherlands to speed up the collaboration with Verint UK. There are indeed cultural differences between England and the Netherlands. An intermediary can offer a solution in this regard.

Speech Analytics at VodafoneZiggo

Roger Becker is lead insight at VodafoneZiggo, analyzing data is its core business. The mobile operator/cable company processes almost 7 million customer contacts with consumers every year. Beckers explains: what the customer says, what the employee hears and what the agent registers; a lot of information is lost in the 'classic' logging of interactions. In other words: the average conversation lasts 660 seconds and leads to an average log of 73 words. From the purpose viewed from VodafoneZiggo – 'Pleasure and progress with every connection' – every customer contact should also lead to insights and learnings to make it more pleasant and better for customers. Speech Analytics fits in well with this. Within customer contact, the use of data should lead to more insight into the context, which in turn forms the basis for process improvements and smart automation - also for example for scaling up coaching of agents.

VodafoneZiggo uses the Avaya platform and Avaya's recording and QM solutions that are already in use are based on Verint technology, so the choice to build on Verint for interaction analytics is understandable. The analytics project was set up in collaboration with Avaya and NTT. Goal: working towards a platform where all data comes together and can be analysed.

Technology is just the beginning

VodafoneZiggo's analytics model consists of three phases: converting speech to text, creating semantic indexes and categories; analyzing and implementing possible improvements with an eye for impact and ROI. VodafoneZiggo therefore started 'from scratch' and in the initial phase even 'good morning' seems to be an important word. In the analysis process, speech and text are treated separately and sentiment analysis is also added. The process of categorizing relevant concepts (including interrelationships and coherence) is labour-intensive and time-consuming simply because products and processes are constantly changing.

Beckers is not the first to point out that purchasing technology is not enough to get started with data: you need a team of experts to make analyzes and translate them into relevant actions. For VodafoneZiggo's B2C customer service, half an FTE is involved in managing the analytics environment. In addition, there is active collaboration with the broader analytics community within VodafoneZiggo for the more challenging investigations. To work more efficiently in the construction of new categories, experts and coaches are also involved who can combine testing categories and coaching their agents during listening back.

Results at VodafoneZiggo

What has interaction analytics brought? Beckers sees the biggest advantage that VodafoneZiggo gains insight into current processes very quickly. Calls are analyzed and ready in the system after XNUMX minutes. This enables VodafoneZiggo to make adjustments on an hourly basis. Knowledge and insight from the conversations can be added to the knowledge base, but you can also turn it around: if you want to enrich the knowledge base with new topics, you can specifically look up relevant conversations about it.

Wifi at home and Viaplay

Beckers also has tips for colleagues. Make sure senior management understands what interaction analytics can do. The system is not designed to spit out 'Top 10' call driver lists, it's about in-depth insight into what to do next. Beckers gives as an example the boom in questions about the Wi-Fi connection, when customers started working from home en masse. In addition to a higher volume, calls also took longer and it took time to gain insight into the various root causes. That insight led, for example, to the decision to set up a campaign for WiFi amplifiers, because many consumers opted for new workplaces at home.

Speech analytics also played an important role in the rollout of ViaPlay. At the time of rollout, the moment at which information could be shared with agents was similar to that of public announcements. As a result, near-live insight into the extent to which all materials are clear for both customer and agent was crucial. The ability to use data to investigate whether customers called specifically for ViaPlay – and thus started the conversation with it – or mentioned it somewhere in passing, also provided a great deal of insight into how VodafoneZiggo could make the rollout as successful as possible.

Basics in order

The fact that the foundation is now in place means that Beckers can think about realizing his wish list: transferring all transcripts to the data warehouse (a PoC is in the making), further optimizing automated quality monitoring and automated recording in the CRM system spoken summaries. (Ziptone/Erik Bouwer)

Customer Experience, Featured Verint