The multinational cleaning equipment company increased consumer self-service and decreased contact center traffic with a chat-first strategy.
Founded in Germany in 1935, Kärcher is a family-owned enterprise and the world’s leading provider of cleaning technology. With pressure washers, vacuums, steam cleaners, pumps, vehicle washes, cleaning agents and more, it offers a range of innovative solutions within both the B2B and B2C space. The company's ambition is to make the lives of its customers easier, more efficient and more pleasant. Meeting customer needs of availability, advice and quick response time with a chatbot solution is one part of this strategy.
Kärcher began working with boost.ai in January 2022 to build and develop a virtual agent for its Finnish, Swedish, and Norwegian markets. The bot, known as Aina in Finland and Airene in the other two markets, serves as a dynamic channel for website visitors to easily access information on the company's products, such as how-to instructions, cleaning tips, and general maintenance advice.
Using proprietary Natural Language Understanding (NLU) algorithms, the boost.ai conversational AI platform enables anybody with a non-technical background to quickly train sophisticated Machine Learning models in any language.
This ease-of-use allowed the AI trainer teams across Kärcher’s three Nordic offices to coordinate in collaboration with each other and boost.ai to build and deploy the virtual agents.
“I really love our chatbot Aina. She is so clever and can answer anything. Even the Finnish language doesn’t confuse her.”
- Tania Anttolainen, Customer Service Manager, Kärcher Finland
Each virtual agent is based on the same model that can answer questions on more than 500 topics. This is accomplished by utilizing the boost.ai platform's filtering feature, which enables a language-neutral strategy. Any of the bots can be asked a question by a consumer, and conversational AI will automatically route to the right intent (or topic) and provide a response in the relevant language.
The project's main objective for Kärcher was to reduce the workload on its contact center. The business intended to employ the virtual agent to first minimize traffic to human agents by at least 20%, with the goal of eventually raising this amount.
Kärcher used a chat-first approach with each of its virtual agents to accomplish this goal, making them the main means of communication with the business via its website. All customer support traffic is first directed to a virtual agent to try to resolve any issues, with contact center details provided if a request falls outside of the bot’s scope, or is expressly requested by a customer.
With each virtual agent conducting more than 4,000 conversations between January and July 2022, this strategy produced outstanding results. Kärcher identified that 21% of all customer service traffic, including phone and email traffic, was going through virtual agents during a service peak in July 2022. This, together with an average 87% resolution rate across all three markets, showed how much the virtual agents were helping to empower customer self-service while also reducing traffic to the main contact center.
Now that Kärcher has achieved its desired 20% reduction in contact center traffic, the company’s AI trainers are looking to expand the capabilities of the virtual agent beyond these initial markets. A Danish version of the bot is scheduled to go live in November 2022, with boost.ai’s Human Chat - a feature-rich live chat alternative - to be integrated in a staggered rollout across all markets over the coming months.