AI sparks widespread discussion, and we’ve just begun to discover its potential in category management. While reading a book on AI, I explored the concept of containment and the limits of AI’s reach. However, amidst these topics, we must stay cautious and mindful of the potential risks associated with AI’s rapid advancement. I believe rational human thinking always plays a vital role, especially in providing necessary oversight in data analysis.
AI and Data Quality: Garbage In, Garbage Out
In AI, like in many other data analysis tools, garbage in equals garbage out. A wealth of data streams from numerous sources makes it hard to keep up with it all, which is why BI tools are necessary today. Data users must carefully evaluate their data sources. The crucial questions are: can we trust the data we receive? Where does it come from? Was it artificially synthesized? Artificially synthesized data is emerging, but data should always be questioned before starting anything, even from big data houses. They aren’t perfect, and sometimes, their methodologies intrigue us. So, trust your instincts; if something doesn’t look right, it probably isn’t.
Current AI Applications in Category Management
I want to highlight a few current AI applications in category management that have positively impacted the field.
- Data Connection, Processing, and Analysis: The main benefit of AI is its ability to save time. It enables users to sift through large datasets and quickly present data comparisons. I use the term comparisons because AI analyzes data but does not fully explain underlying reasons. For example, we may find that product XYZ has increased by 55%, possibly due to improved distribution as it also rises. While correlations can partially explain cause and effect, human consumption introduces numerous other variables. This scenario illustrates where companies might err, assuming that out-of-the-box AI provides absolute truths. It is also why consumer insights and research are essential for understanding category management analysis.
- Web Scrapes, Data Capture, and Compilation: Here’s a time-saving example. When implementing a BI tool, my team had to build a cross-reference library to harmonize data from sources like IRI Unify, NPD, POS, SAP, and EDI. The main issue involved ensuring proper data coding so the data could align for a unified output. Coding product details line by line, including category, subcategory, and description, consumed hours. Now, a tool called Harmonya simplifies this process. Harmonya scrapes product descriptions, content, and even reviews down to the UPC level and creates a consumer-centric attribution. This tool harmonizes data, providing a unified view both internally and externally. This aspect proves significant because many retailers and manufacturers code categories differently. AI rapidly resolves the data alignment task, saving countless labor hours. Contact Andy Nielsen at andy.n@harmonya.com to learn more about Harmonya.
- Image Capture and Analysis: Another area that has advanced significantly over recent years is robotics, specifically its application in image capture and supply chain management. Various companies now provide these services, leading to reduced labor-hour costs. I remember having to capture individual item images to complete planograms and analyze them. Today, robotic tools can scan shelves, capture product images, and automatically identify out-of-stock items, item performance, and inventory status. For insightful articles on robotics and category management, check out Georges Mirza’s work [here] www.InnovatingCategoryManagement.com.
- The Role of Human Capital in Category Management: This leads us to the importance of human capital, which will never fade away. I cannot emphasize enough the value of hands-on experience. Correlating findings from spending time with buyers, walking stores, and talking with consumers while they shop is a must. Only through practical experience can individuals identify issues or concerns in data. Data machines effectively gather and process vast amounts of information with minimal errors, but they do not replace the thought process of a human, at least not yet. Just remember: garbage in, garbage out. Data accuracy remains paramount to the AI revolution. When examining the reasons behind data trends, statistics will help us understand correlations, standard deviations, predictions, and accuracy in data analysis. However, statistics won’t reveal what the human brain is thinking and how that relates to the data behaviors we observe.
Contact us if you belong to a retailer, manufacturer, or broker group seeking category management and market research services. There is no need for three vendors; we serve as your one-stop shop to build a bespoke solution.
