OTC Leader: Automating Market Analysis with AI Knowledge Graph, Saving 600 Hours Monthly

Overview

As a leader in China's OTC market, the company needs to constantly monitor the market dynamics of thousands of chain pharmacies nationwide to formulate precise sales strategies. However, faced with a massive amount of Key Account (KA) chain store data and scattered sales reports, the traditional manual analysis model has become severely outdated, becoming a bottleneck that restricts business agility. To address this pain point, the company chose to collaborate with Tezign to introduce an enterprise-level 'AI Think Tank' based on knowledge graph technology, aiming to reconstruct the market analysis process and elevate data processing capabilities to an intelligent level.

Challenges

Before digital transformation, market analysis work relied heavily on manual completion. Analysts needed to manually collect, organize, and analyze sales reports from different provinces and different KA chain stores. This work was not only time-consuming—completing an analysis report for a single province often took 3-5 hours—but also highly repetitive, leading to inefficiencies and a high likelihood of human error. The team spent over 60 hours each month just analyzing 10 major KA chain stores.

More critically, there was the issue of 'data silos'. Key information such as sales data, competitive dynamics, and market trends was scattered across cloud storage, internal ERP systems, and third-party market data platforms, lacking a unified entry point and standards. This resulted in inconsistent output standards among different analysts, and data assets could not be effectively consolidated. Valuable expert energy was consumed by a large amount of low-value data handling work, making it difficult to focus on high-value strategic thinking and in-depth insights across regions.

Solution

To break this deadlock, the company deployed a customized 'AI Knowledge Graph Agent'. This solution goes beyond a simple automation tool, reconstructing the logic of analysis and decision-making from the ground up through an innovative model of 'knowledge extraction + Q&A':

1. Unified Knowledge Base (ContentLake): The system first ingests scattered unstructured and semi-structured data such as PDF reports, PPT proposals, and Excel spreadsheets into the company's ContentLake. This provides AI with a comprehensive, real-time updated knowledge base.

2. Intelligent Extraction and Reasoning: Utilizing large language models (LLM) and knowledge graph technology, the Agent can 'read' and 'understand' these complex documents. Users do not need to perform cumbersome query operations; they can simply ask questions in natural language (e.g., 'What is the sales budget for skincare products in 2023?' or 'Compare the performance of Brand A and Brand B in Guangdong'), and the AI can accurately extract key information from a vast array of documents and perform logical reasoning and data synthesis.

3. Automated Report Generation: The system supports one-click online report generation. It can automatically convert the extracted data into intuitive charts, trend summaries, and strategic highlights, formatted according to the company's standard templates. What used to take hours of manual tabulation can now be completed in about 1 minute.

Results

The launch of the AI Think Tank has brought a qualitative leap in efficiency. The time taken for a single analysis task has been reduced from 3-5 hours to under 1 minute, achieving hundreds of times improvement in efficiency. The system saves over 600 hours of analysis time each month, greatly freeing up expert energy, allowing them to invest more time in formulating and optimizing market strategies.

Currently, the system has successfully covered over 10 core KA chain stores, achieving standardized management of cross-regional data and multidimensional comparative analysis. It has transformed market analysis from a passive, labor-intensive task into an active, intelligent core capability, making the company's knowledge management and business decision-making unprecedentedly simple and precise.