Automate your product data management processes and save time and costs! AI speeds up classification, description generation and error elimination. Find out how to improve operational efficiency!
On this page 4 sections
Introduction
Implementing PIM (Product Information Management) systems is a process that requires precise preparation, cleaning and categorization of product data. At LemonMind, through the use of artificial intelligence, we support our clients in automating these complex processes, resulting in significant savings in time and resources.
Artificial intelligence is playing a key role in the transformation of product information management. Its ability to analyze and process large data sets makes it possible to automate many routine tasks, such as product classification and description generation. This makes these processes more efficient and less error-prone.
At LemonMind, as part of our LemonHub.AI solution, we integrate AI with PIM systems to streamline product data management processes. Our approach includes automatically mapping product attributes from various sources, reading data from product labels, and creating intelligent assistants equipped with up-to-date product knowledge. This comprehensive approach to PXM (Product Experience Management) using AI allows us to deliver consistent and compelling product information across all sales channels.
Below are specific examples of the application of our solutions:
To address these challenges, we implemented a solution based on artificial intelligence algorithms that automatically maps products to the right categories.
Assuming that the traditional method required 1,400 hours of work, at a standard rate of PLN 150 per hour, the operating costs would be:
- 1,400 hours × PLN 150 per hour = PLN 210,000.
By implementing a solution based on artificial intelligence algorithms that automatically maps products to the appropriate categories, the workload was reduced by 80%. This means that manually processed will be only:
- 1400 hours × 20% = 280 hours.
Calculating new operating costs:
- 280 hours × PLN 150/h = PLN 42,000.
Through automation, a direct savings of:
- PLN 210,000 (manual cost) minus PLN 42,000 (cost after AI) = PLN 168,000.
In addition, the built-in data analysis module made it possible to minimize the risk of errors, which estimated additional savings of 10% of the original manual cost, i.e.:
- PLN 210,000 × 10% = PLN 21,000.
The total savings achieved by implementing the solution thus amounted to:
- PLN 168,000 + PLN 21,000 = PLN 189,000 in 10 months.
By implementing LemonMind's AI-based solution, the client has not only significantly reduced operating costs and reduced the time needed to classify products, but also achieved a consistent and standardized data structure. This standardization facilitates integration with IT systems and improves the quality of reporting, which is crucial for sales analysis and strategic decision-making. Automating the process has enabled employees to focus on higher value-added tasks, eliminating monotonous, time-consuming manual tasks.
In conclusion, the use of artificial intelligence in automated product mapping has proven to be not only a technological improvement, but also a strategic tool to optimize costs, increase operational efficiency and minimize the risk of errors, which has consequently contributed to the further development of our client's business.
Assuming that we have 150,000 products to process, the total working time would be:
- 150,000 products × 5 minutes = 750,000 minutes, or about 12,500 hours.
At the standard rate of PLN 150 per hour, manual processing would generate costs of:
- 12,500 hours × PLN 150/h = PLN 1,875,000.
By implementing an AI-based solution that automatically recognizes and extracts key information from labels (such as ingredients or nutritional values) and integrates it with the PIM system, manual work was reduced thanks to a 95% efficiency rate. This means that only 5% of products require further manual intervention. In practice, manual processing will be:
- 150,000 products × 5% = 7,500 products.
Calculating the time required to manually handle these products:
- 7,500 products × 5 minutes = 37,500 minutes, or about 625 hours.
The cost of manual processing in the new situation would be:
- 625 hours × PLN 150/h = PLN 93,750.
The automation resulted in a direct savings of:
- PLN 1,875,000 (manual cost) minus PLN 93,750 (cost after AI) = PLN 1,781,250.
In addition, the use of the AI system made it possible to detect errors as early as the extraction stage, avoiding the costs associated with correcting them later. We estimate that this reduction in errors translates into additional savings of 10% of the original manual costs, or approx:
- PLN 1,875,000 × 10% = PLN 187,500.
The total savings achieved by implementing the solution thus amount to:
- PLN 1,781,250 + PLN 187,500 = PLN 1,968,750.
With this solution, the client has not only significantly accelerated the data entry process, but also reduced the number of errors to a minimum. Automating the process allowed employees to focus on more strategic tasks, while increasing operational efficiency and improving data quality in the PIM system.
Assuming that it took an average of 3 minutes to manually develop a product name and that the client's portfolio included 50,000 products, the total effort would be:
- 50,000 products × 3 minutes = 150,000 minutes, or about 2,500 hours.
At the standard rate of PLN 150 per hour, the cost of the manual process would be:
- 2,500 hours × PLN 150/h = PLN 375,000.
Implementing our artificial intelligence-based system, which automatically generates product names, taking into account key product features and applicable regulations, reduced the time required by 70%. This means that the new process requires only 30% of the original workload, or approx:
- 2,500 hours × 30% = 750 hours.
At the same rate, the new operating costs would be:
- 750 hours × PLN 150/h = PLN 112,500.
The direct savings thus amounted to:
- PLN 375,000 minus PLN 112,500 = PLN 262,500.
In addition, the automatic generation of names ensures their consistency and compliance with guidelines, which improves brand perception and reduces the number of potential errors. We estimate that the minimization of errors saved another approximately 10% of the original cost (or about PLN 37,500).
In total, the implementation of the AI system has resulted in savings of about PLN 300,000. In addition, by automating the process, employees can focus on higher value-added tasks, which translates into improved operational efficiency and a stronger positive brand image in the food market.
Assuming that it takes an average of 10 minutes to prepare one technical description, the total workload would be:
- 250,000 descriptions × 10 minutes = 2,500,000 minutes, or about 41,667 hours.
At the standard rate of PLN 150 per hour, the cost of the manual process would reach:
- 41,667 hours × PLN 150/h = PLN 6,250,050.
The use of AI algorithms, which automatically generated detailed technical descriptions based on data from the PIM system, made it possible to automate the process. This reduced the team's time by 80%, meaning that only 20% of the original workload had to be done manually. In practice, this leaves:
- 41,667 hours × 20% = 8,333 hours.
New operating costs with this efficiency were:
- 8,333 hours × PLN 150/h = PLN 1,249,950.
The direct savings thus amounted to:
- PLN 6,250,050 minus PLN 1,249,950 = PLN 5,000,100.
In addition, the automatic generation of descriptions by the AI system allowed for a significant reduction in errors that could have generated further costs in the traditional process. We estimate that the elimination of errors saved an additional 10% or so of the original manual costs, viz:
- PLN 6,250,050 × 10% = PLN 625,005.
In total, the implementation of the AI system contributed to savings of the order of:
- PLN 5,000,100 + PLN 625,005 = PLN 5,625,105.
Implementation of this solution allowed the client not only to significantly reduce operating costs, but also to ensure high quality and consistent product information. By automating the process of creating technical descriptions, employees can focus on higher-value-added tasks, which contributes to the company's overall operational efficiency.
Three more tasks AI takes over in a PIM
Finding errors and gaps
AI compares product records and finds missing values, typos and values that contradict each other, so the team fixes them before the data reaches a channel.
Classifying from descriptions and images
Natural language processing reads product descriptions and computer vision reads product photos. Together they suggest attributes and the right category for each product, which also makes products easier to find.
Spotting where the process loses time
An analysis of the PIM workflow shows repeated manual steps and the places where products wait longest. That tells you what to automate next.
Start with one product group and one task, compare the time before and after, then extend. Our AI consulting works the same way: a short workshop, then a pilot on one data set. See how LemonHub.AI works on PIM data and the LemonHub.AI project for Ceramika Paradyż.
An example in Pimcore: descriptions from attributes
We built a description generator for Pimcore that follows this model:
- The prompt is built from attributes already in the PIM, such as colour, material, dimensions and intended use.
- An editor starts it on demand with a button in the Data Object view.
- Generation runs in the background with Symfony Messenger, and the field is filled when the text is ready.
- The editor uses, edits or rejects the text before it is published.
More about the background queue: how Pimcore works with Symfony Messenger.
Summary
Integrating artificial intelligence with PIM systems brings tangible benefits in the form of automating complex processes, reducing errors, and saving time and costs. With such solutions, companies can focus on the strategic aspects of their business while providing consistent and compelling product information to their customers.
At LemonMind, we are constantly developing our tools to meet the growing needs of the market and support our customers in effectively managing product information.
Read more: how AI is changing PXM and AI-generated content and SEO.
