Customer: Large Online Retail

Florium is the largest online plant and gardening retailer in Ukraine. The company offers a wide assortment of more than 5,000 plant varieties, selling planting materials across Ukraine and abroad. According to the latest publicly available figures from open sources (including Forbes rankings), Florium reported revenue of approximately USD 5 million in 2022. The leadership places strong emphasis on high-quality customer service, timely communication, and maintaining loyal relationships with returning customers to support stable and controllable demand.

Challenge: Gather Unbiased Customer Feedback Across All Customers

Florium required a scalable and unbiased way to measure customer satisfaction (CSAT) and Net Promoter Score (NPS) across its entire customer base, particularly after order delivery. Manual outbound calls were resource-intensive, as Florium operates a massive customer database. As a result, obtaining statistically representative customer insights at scale was challenging. An automated solution capable of engaging all customers consistently would help the company effectively improve CSAT and track NPS over time, capturing structured feedback from all customers, while maintaining a high standard of customer care in every interaction.

Solution: Post-Delivery Customer Satisfaction (SCI) Survey

We launched an AI answering agent for post-delivery customer experience calls, focused on gathering structured, unbiased feedback about plant blooming. The assistant:

• greets the customer and states the purpose of the call clearly,
• explains the 1 - 5 rating scale and asks for a score,
• responds differently depending on the rating or questions/complaints,
• records the feedback in a structured format for the marketing team.

Conversation Example: 

Transcript:

- Hello.

- Good afternoon! You are greeted by the Care Department of the Florium online store. It is very important to us that every order you receive brings you only positive emotions. Please rate, on a scale from 1 to 5, where 5 is the highest score, how satisfied you are with the blooming of the plants from your recent order.

- Five.

- We are extremely pleased to know that you are satisfied. Thank you sincerely for the high rating. May your plants bring you joy every day! By the way, we have prepared discounts on our entire assortment, so feel free to visit our website and choose the best for yourself. We will be happy to see you again! Wishing you a wonderful day — take care!

Results

To evaluate the effectiveness of the conversational AI communication, EVE.Calls applied its proprietary evaluation methodology. During the pilot phase, EVE.Calls reviewed and analyzed all call recordings involving the AI Voice Agent. Following customer approval, the AI Voice Agent was deployed in live operations. From approximately 1,000 outbound calls processed across multiple successful campaigns, a representative 10% sample was selected for in-depth qualitative analysis.

Using this methodology, the selected calls were assessed across various dimensions, including logical consistency of the dialogue, relevance of responses, response latency, accuracy of personal data recognition provided by the client, and tone of communication. Based on this analysis, the AI Voice Agent demonstrated over 97% accuracy in managing conversations.

In the remaining cases (3%), minor non-critical issues were observed, such as brief response delays or occasional conversational imperfections. These did not result in incorrect information, policy violations, data exposure, or any negative impact on clients. All such issues are being addressed through ongoing system improvements. 

There are multiple advantages of the AI agent:

• High level of engagement: 61% of customers stayed engaged and provided a satisfaction score on a 1–5 scale, where 5 indicates the highest level of satisfaction.
• Significant reduction in call center load: 13 hours of human labor saved across 1,000 calls, with an average call duration of approximately 45 seconds.
• High customer satisfaction: the average score given by customers was 4.4 out of 5.