Congratulations to our customer Sky

Pienso’s long-standing customer, Sky, has retained its position as the provider with the fewest complaints per customer across pay-TV, broadband, landline and mobile services in the United Kingdom. The UK’s communications regulator, Ofcom, is there to receive complaints from customers who are frustrated with their telecom services. And to better understand the reasons for dissatisfaction among residential customers, Ofcom publishes quarterly reports on the number and nature of complaints related to each provider. 

For the fifth consecutive quarter, Sky has received the fewest complaints among providers, having received just 5 complaints per 100,000 customers, well below the industry average of 12. This is a testament to Sky’s commitment to customer satisfaction and their corresponding technology and resource investments. Since 2019, Sky has used Pienso in their contact centers to help monitor why customers are calling and inform proactive service improvement.

Sky’s Customer Insights team originally invested in Pienso because they wanted a more holistic understanding of their customers’ issues so they could address them more comprehensively. They had tried different off-the-shelf AI approaches before, but none had satisfactorily unearthed the insights that Sky knew hid within their mountains of unstructured text data — call transcripts from recorded customer conversations. It would be a difficult task. But it turned out to be an excellent, early success story for Large Language Models (LLMs).

In 2019, LLMs were still a nascent subfield of Natural Language Processing (NLP) research. And even as the space began to expand thanks to Google’s introduction of BERT, leveraging this technology still required advanced technical personnel, i.e. data scientists and Machine Learning engineers.

Cue, Pienso.

Pienso’s no-code/low-code LLM platform allowed Sky to analyze their customer interactions with more precision, control, and accuracy than ever before. Using Pienso, Sky’s nontechnical business users could fine-tune an LLM so that it speaks the language of their data — without needing any help from Sky’s over-extended data science team. And by putting AI/ML in the hands of the people who understand the data the best, Sky was able to start tapping the wealth of knowledge in their data.

Sky now analyzes each customer interaction to identify the topic of the customer’s issue, the emotions presented by the customer during the interaction, and the overall sentiment expressed during the interaction (scored from -1 to +1). These new metrics go beyond the out-of-the-box CSAT analytics or NPS surveys that Sky relied on before, thanks to the power of Large Language Models that have been fine-tuned on actual Sky data.

Taken together, these metrics have enabled Sky to develop a more comprehensive picture of specific customer journeys, identifying points of friction or areas for improvement. And these insights have driven the level of customer satisfaction demonstrated in Ofcom’s report.