Budget Thuis enhances customer experience with (Gen) AI
Budget Thuis is a leading energy and telecom services provider in the Netherlands, committed to providing reliable services at affordable prices. Serving more than a million customers, Budget Thuis holds a prominent position in the industry. Recognizing the substantial potential of Generative AI to enhance internal processes and the customer experience, Budget Thuis faced challenges in realizing its value independently. Therefore, they asked SparkOptimus’ support in identifying and prioritizing company-specific use cases, rapidly testing and proving the value by collaboratively building, testing, and improving MVPs, and designing an organizational setup for long-term, sustainable (Gen) AI innovation.
"Although we usually prefer to do as much as possible in-house, Spark's practical support enabled them to blend into our teams easily. They made Generative AI easy to understand and focused on practical, effective results."
We kicked off the project with a one-week value scan, collectively identifying 50 Generative AI value cases. From this longlist, we prioritized three use cases to start testing in across-functional pilot team, and further selected 12 use cases to implement with off-the-shelf Gen AI tools.
We then developed the implementation plans and set up the pilot teams to kickstart the development of the pilots. In just 8 weeks, we built and tested the MVP.
Following the successful test-and-learn phase, we developed detailed scaling plans for each pilot project and established a new organizational setup for innovation to enable Budget Thuis to independently replicate this test-and-learn approach for future Generative AI applications.
Take the insights with you
During this project, we built successful MVPs for several Gen AI use cases:
Call log automation pilot
We collaboratively developed an MVP tool that accurately transcribes and summarizes customer service & tele sales calls according to a specified format, reducing time spent by agents by as much as 15%. As a byproduct of the Gen AI-based summarization, the quality and consistency of call summaries also improved. Automated call logging enables agents to spend more time on value-added activities, like providing a better customer experience.
Chatbot improvement pilot
We developed a Gen AI tool to identify chatbot improvement areas using large amounts of historic live chat data. The tool categorizes conversations and identifies reasons for customers to escalate to a service agent. Furthermore, the tool also generates typical reasons why customers escalate for each category. This type of analysis, which is virtually impossible to effectively execute without Gen AI, allows the team to rapidly improve the chatbot, reducing escalations, and thus time spend for customer service.
Integrating off-the-shelf tooling
In addition to the pilots, we introduced off-the-shelf Generative AI tooling across key departments. To ensure successful adoption into daily operations, we appointed dedicated ‘champions’, tasked with driving Generative AI within their department. Working closely with these champions, we refined potential use cases, identified suitable tooling, documented best practices, and conducted training sessions for comprehensive adoption and alignment on the way forward.
Organizing for continuous innovation
To meet the challenge of driving continuous innovation in the business, we designed a tailored test-and-learn approach. This involved setting up dedicated cross-functional innovation teams, and implementing a quarterly meeting rhythm for ideation, prioritization, and evaluation of ideas. Altogether, this positions Budget Thuis to autonomously generate, prioritize, test, and expand innovative ideas, unlocking untapped (Generative AI) value!