“There is a limit to what people alone can do. That was when we received the AICALL proposal.”
These are the words of Mr. Koizumi, Managing Director and registered pharmaceutical salesperson at NR Co., Ltd. Serving customers ranging in age from their 20s to their 90s with pharmaceuticals, skincare products, and general merchandise, the company previously optimized its cancellation process by integrating BOTCHAN Keeper with BOTCHAN AI (a chatbot), reducing the number of complaints by approximately 40%.(See the previous article, “Consolidating the Entire Cancellation Process with BOTCHAN: Delivering a Consistent Customer Experience Through Data Integration.”)
Subsequently, as new customer acquisition for products targeting older consumers increased, the call center faced a new set of challenges. In this article, we explore the reasons behind the implementation, the innovative scenario design tailored to pharmaceutical e-commerce, and the changes that followed.
Meet the Team: From Frontline Operations to Management, Always a “Player”
Mr. Koizumi: “My responsibilities as a key player in CRM, SCM, and fulfillment remain unchanged from before, but I have also become involved in management. I had already been overseeing the company’s financial performance in my previous role, so it does not feel like a major change. What has changed is that I now have more opportunities to view the company as a whole from a longer-term perspective.”
Ms. Tomomi Koizumi, Managing Director and Registered Sales Clerk, NR Co., Ltd.
Pre-Implementation Challenges and Background: Growth in New Customer Acquisition Exceeded Call-Handling Capacity
Mr. Koizumi: "While we were able to provide adequate support to customers who contacted us through Keeper, our chatbot designed to prevent cancellations, and other chatbots, growth in new customer acquisition for products targeting older demographics meant that more of these customers ended up calling us than anticipated. We had expanded our capacity by relocating and enlarging our call center operations, but when staffing levels fell due to absences, influenza outbreaks, typhoons, or other factors, we sometimes struggled to handle customers who contacted us by phone. Inquiries were particularly concentrated on Monday mornings, and there were times when we could not handle them all. We received the AICALL proposal just as we were realizing that staffing alone would not be enough, so we saw it as a good opportunity and decided to participate."
Morikawa, head of the BOTCHAN AICALL business at wevnal Co., Ltd., reflects on this challenge as follows.
Morikawa: “With limited resources, I believed there was universal value in bringing the experience of being unable to get through by phone as close to zero as possible. At the same time, because AICALL was a new initiative for NR, we were particularly careful to ensure that it would operate without incident and handle end users appropriately and respectfully. The challenges were clear: staff shortages, unanswered calls, and potentially rising costs. We were truly grateful that you worked with us toward the same goal to resolve them.”
Discovering the service and deciding to adopt it: “Even if it cannot replace humans entirely, it will undoubtedly be useful if there is a clear division of roles.”
Mr. Koizumi:“To be honest, I had two conflicting feelings: hope that it would be great if this could really work, and concern about whether it could perform at the same level as a human. Since I was not yet familiar with AICALL itself, I did not know the extent of what it could handle. However, I decided to implement it because I believed that even if it could not replace humans entirely, it would certainly prove useful once we could clearly define how responsibilities should be divided.”
BOTCHAN’s proven track record across its products and the trust built over time were also factors in reducing resistance to adoption. There were also cost concerns about an operational model that relied solely on increasing headcount, so the company evaluated the solution from both perspectives.
Morikawa explains what level of results should be expected in the initial stages of implementation.
Morikawa: “Handling cancellation requests is work that someone has to take on, whether it is AI or a human. In the future, we hope to improve the quality of our interactions so that customers may reconsider canceling. However, during the initial implementation phase, we proceeded on the basis that securing sufficient resources and maintaining a strong response rate should be the top priorities.”
Implementation Process and Runtime Decision Criteria: Defining Requirements with a Focus on the Boundaries Unique to Pharmaceuticals
When designing the scenario, we finalized the requirements in consultation with the customer support team, using the cancellation criteria data managed in the BOTCHAN Keeper cancellation prevention chatbot as a foundation.
Mr. Koizumi: "All the conditions governing whether a cancellation could be accepted were stored in Keeper, so we used that information as a reference while coordinating with the person in charge. Our company places great importance on standards and rules, and they carefully configured the system by confirming every detail with us, asking questions such as, 'How should we handle this?' They even clearly identified which customers should not be handled by AI, so we felt confident moving forward with the implementation."
Morikawa's words also convey just how meticulously the requirements were defined.
Morikawa:“Calls requesting cancellation inevitably require identity verification, and the contract details must be accessed in real time. As we refined the process, the requirements continued to grow, including delivery lead times, the brand policies each company prioritizes, and the extent of the authority granted to update the status after a call. We believed that people and AI could work together to provide a hybrid customer service experience, but from the customer’s perspective, some aspects presented nothing but risk. Our greatest challenge was minimizing the risk of critical errors, such as incorrectly changing a contract status.”
Given the nature of handling pharmaceuticals and quasi-drugs, scenario design for routine interactions proceeded smoothly. However, both parties devoted particular attention to determining where to draw the line on risks that could lead to health hazards and should therefore fall outside the scope of AI support.
Results and value achieved after implementation: Maintaining response rates during staff absences and peak periods
Since implementation, call center staff have been able to determine how much of the workload AICALL should handle and adjust that share flexibly based on staffing availability. AICALL’s share is not determined automatically; instead, call center staff review daily call volumes and staffing levels and adjust it as needed. On weekdays, AICALL handles a set percentage of incoming calls, while its share is increased further on days following weekends and public holidays, when inquiries are particularly concentrated. This approach allows for fine-grained adjustments by time of day and day of the week. As a result, even as incoming call volumes have risen due to growth in new customer acquisition, the call center has maintained an answer rate above 90% without significantly expanding its workforce.
Mr. Koizumi: “The ability to flexibly adjust the balance—for example, by increasing the proportion of calls handled by AICALL when staff are unavailable—has made a tangible difference. Even during the recent natural disaster, we were able to continue assisting our largest customer segment without completely suspending phone support. The fact that we did not have to shut it down entirely was truly significant.”
Morikawa:
"The allocation is not determined solely by our company. To a large extent, call center staff make these decisions based on daily call volumes and staffing levels. We believe this approach is only possible through close coordination with the frontline teams."
With that in mind, Morikawa reflects on the situation at the time as follows.
Morikawa: "Even if we assume that this is what should happen in theory, we cannot know what will actually happen. Given that even some large companies have not incorporated this into their business continuity plans, we were grateful to have had the necessary preparations in place."
Koizumi also discusses the insights revealed by the conversation data.
Mr. Koizumi:“Reviewing the conversation data revealed that some terms we had taken for granted were not understood as well as we had expected, and that some customers did not listen through to the end of the explanation. Human operators had been adept at navigating these situations, which had obscured them, so gaining insight into customers’ natural tendencies was a valuable discovery. Unlike one-way communication such as automated voice messages, the AI considers and responds to customers’ questions. I believe its strength lies in enabling flexible interactions rather than following a fixed script.”
Koizumi also senses a change in the mindset of frontline operators.
Mr. Koizumi:
"We do not have many opportunities to speak directly with the operators, but I feel that when call volumes surge, knowing they can rely on AI CALL rather than having to manage entirely on their own gives them some peace of mind and helps ease the pressure."
Morikawa says that AICALL also offers unique advantages when it comes to keeping records of customer interactions.
Morikawa: “I sometimes sit in on reports presented by call center BPO providers, but because the data is manually compiled and summarized after the interactions, the level of detail in the records inevitably varies. With AICALL, evaluation criteria can be applied retrospectively, and 100% of the interaction content can be retained as logs. I see real value in the foundation it provides by digitizing the traditionally analog touchpoint of telephone calls.”
Future Outlook and Key Initiatives: Delivering an Authentic Customer Service Experience Through Effective Collaboration Between People and AI
Mr. Koizumi: "What we really want is to have meaningful conversations with our customers, but given the volume of calls, the call center has KPIs to meet, making it difficult to give each conversation the time it deserves. While we hope AICALL will eventually be able to interact just like a human, even if it does not reach that point, we want to create an experience closer to the person-to-person customer service we originally envisioned, whether through a division of roles, collaboration, or other approaches. Because AICALL enables consistent testing, we would also like to use it to identify effective phrasing for customer retention conversations—which can otherwise vary by representative—and share those findings with our operators. The team proactively reports on where the issues lie, why they are occurring, and how they can be addressed. As we operate with a small, highly capable team, we are truly grateful that they identify and address details we might otherwise overlook."
Morikawa describes what lies beyond this “division of roles” as follows.
Morikawa: “Rather than viewing this as a binary choice between leaving everything to AI and entrusting tasks exclusively to people, we aim to find the right balance along that spectrum to suit each company’s circumstances. We see our work with NR as one of the first milestones in that journey.”
A Message to Companies Facing Similar Challenges
Mr. Koizumi: "Mail-order companies can experience rapid growth when new customer acquisition increases, and it is difficult to adjust staffing levels immediately in response to such significant fluctuations. With a system like AICALL in place, which can respond flexibly once configured, I believe the issue of unanswered calls can be mitigated to some extent, even if not completely eliminated. I also feel that AICALL is more cost-effective than adding more locations as a risk-mitigation measure in the event of a disaster or similar emergency. I would recommend implementing it as a first step."
"BOTCHAN AICALL" is an AI call service for companies seeking to maintain stable response rates regardless of call center staffing levels. If your company is struggling with high call volumes, whether in the pharmaceutical or mail-order industry or any other sector, please contact us to learn more.
