NESIC: Leveraging Process Mining and Cluster Analysis to Optimize Sales

NEC Networks & System Integration Corporation specializes in the design, construction, and installation of private branch exchange and networks for both private and public clients in Japan. In recent years, NESIC has expanded its portfolio to include SaaS resale, systems integration, and operational services. This expansion notably includes the resale of web conferencing tool Zoom in Japan.

But that’s just the beginning. Keep reading to uncover how NESIC navigates diverse markets with innovative strategies, optimizes sales processes, and utilizes advanced analytics to drive business growth and staffing efficiency.

Navigating Diverse Markets With Standardized Sales Strategies

With recent advancements across various industry sectors embracing SaaS solutions, NESIC has adjusted its sales organization to better cater to specific industries. However, managing the complexity of sales methods and workflows remains challenging, as the absence of standardized processes makes it difficult to provide quantitative recommendations.

The organization operates with subdivisions, where each employee is responsible for conducting business negotiations within specialized industries. Therefore, it’s essential to recommend the most suitable negotiation methods for each individual or allocate staff accordingly. Given these circumstances, there is an urgent need to visualize the negotiation process in the sales domain, reassess human resource allocation, and formulate measures for personnel evaluation to identify and address any performance gaps among employees.

In the past, each sales representative at NESIC was tasked with conducting business negotiations independently, despite encountering a diverse range of clients and pricing tiers.

However, this approach presented several challenges:

• Lack of standardization in the process of acquiring sales from customers.
• Inability to quantitatively evaluate personnel performance based on sales activities.
• Suboptimal staffing allocation according to sales representatives.

To address these issues, NESIC utilized the following models:

• Business transition process mining.
• Cluster model based on the volume and value of business negotiations.
• Reports integrating process mining and cluster models.

Streamlining Sales Processes With Process Mining and Cluster Analysis

Dataiku offers a unified platform for data preprocessing, feature engineering, machine learning, and model monitoring. Data preprocessing tasks are managed using Dataiku recipes, while Business Solutions are utilized for the learning model’s predictive output method and process mining.

The project, led by a team of 10 individuals including two recipe creators and a service planning advisor, closely collaborated with the sales department, acting as the end user, to address two critical issues:

1. Process Mining of Business Meetings Registered in SFA
To optimize business meetings, a process mining technique was employed, analyzing the classification of meetings recorded in the SFA system over time. This analysis revealed the 10 most common business process patterns, providing valuable insights. Initially, creating the first process mining model took about a month. However, simply presenting the results wasn’t enough to guide end users effectively. To address this, two important innovations were introduced:

• Presenting Model Process User Results
  Sales representatives were interviewed to identify individuals showcasing exemplary negotiation practices. By focusing on these role models and narrowing the scope, a visual representation of the negotiation process was developed as a reference for others. Notably, findings revealed that sales activities often involve a dynamic process, with significant movement and negotiation concerning amounts and delivery dates.
• Visualizing Process Mining Results by Sales Team
  Recognizing the diverse customer interactions across sales teams, the approach was customized accordingly. Analyzing process mining outcomes unique to each team allowed for the recommendation of negotiation strategies tailored to their respective industries. This process facilitated the identification of biases within the sales organization and ensured that each team adopted a process aligned with their unique requirements.

2. Cluster Analysis of Deals Registered With SFDC
Business negotiations encompass both high-value and low-value projects, leading to significant variations in the amount, range, and number of projects handled by different individuals. To address this variability, clustering techniques were employed.

• A BIN was created to map and visualize the revenue, number of cases, and number of customers handled by each sales representative. This provides insights into the sales representative’s workload, pricing range, and capacity for managing projects.
• Additionally, the distribution of cluster results by organization was examined. This helps in organizing and optimizing the workload of sales representatives effectively. Nearly 30 recipes and one model were developed within the design node after data cleansing in Snowflake.

Leveraging Visualizations, Performance Assessment, and Quantitative Metrics

Benefits for the sales department include:

1. Visualizing the business negotiation flow based on industry standards, providing a reference for the negotiation process led by the model representative.
2. Assessing the performance of individuals based on project pricing, volume, and customer base, utilizing clusters as a reference for organizing sales teams effectively.
3. Incorporating clusters and process results as quantitative evaluation metrics, alongside reference information, for comprehensive performance assessment.

By using sales activities in the education of new employees and mid-career employees, I was able to have an image of the process as a guideline.

Empowering Users to Enhance Sales, Personnel Evaluation, and Staffing Efficiency

Dataiku enhances NESIC’s sales activities, personnel evaluations, and staffing efficiency. By incorporating sales activities into the training of new and experienced employees, a clear process guideline is established. In personnel evaluations, quantifiable sales activity data serves as valuable reference information, providing insights into the employees‘ hard work. Moreover, personnel allocation can now be based on individual suitability for specific activities and process methods within organizations.

Dataiku delivers valuable capabilities accessible to users of all skill levels in data science or programming:

• Streamlined Process Mining: With Dataiku’s Business Solutions, users can quickly construct process mining flows, even without advanced technical skills, streamlining the process.
• Interpretable Cluster Models: Dataiku enables the creation of easily interpretable cluster models through tools like partial dependency graphs, important feature graphs, and correlation analysis, enhancing understanding and insights.
• Integration of Latest Analysis Methods: Users can seamlessly employ the latest analysis methods from a variety of business solutions within Dataiku, enhancing their analytical capabilities and keeping up with industry trends.

This solution makes our sales activities, personnel evaluations and staffing smarter and more efficient.