How to automatically find new products in supplier price lists
Product range expansion is a key driver of e-commerce sales growth. However, manually searching through hundreds of supplier price lists for new products takes tens of hours. Automating this process allows us to identify promising products in real time and add them to the catalog in minutes.
In this guide, we'll cover modern methods for automatically searching for new products, matching algorithms, selection criteria, and quick import tools.
Why expand your product range?
Constantly updating your catalog is critical to the competitiveness of your online store:
Increase traffic and sales
Each new product offers additional entry points through search engines. Expanding your product range by 20-30% can increase organic traffic by 15-25%. More products mean more opportunities for cross-selling and upselling.
Capturing new niches
Early introduction of trending products allows you to establish leadership positions in new categories before competitors arrive. This is especially important in fast-growing markets, where first movers gain a significant advantage.
Reducing dependence on individual positions
A diversified product range protects a business from the risks associated with supply interruptions or falling demand for specific products. A broad product line creates a more sustainable business model.
Customer retention
Customers prefer stores with a wide selection, where they can find everything they need in one place. Regularly updating the catalog encourages repeat visits and increases customer lifetime value.
Statistics: Shopify research shows that stores that add 10+ new products weekly experience 40% higher revenue growth than those that update their inventory less frequently.
Methods for finding new products
There are several approaches to automatically identifying new products in supplier price lists:
1. Comparison by SKU and SKU
Basic method based on matching unique product identifiers:
- Algorithm: The system loads the supplier's new price list and compares the SKUs with the existing product database. Any items not in the database are marked as potential new arrivals.
- Advantages: High accuracy (95-98%), fast processing of large volumes of data, minimum false positives.
- Restrictions: Does not work if the supplier has changed the article number system or uses non-standard formats.
2. Semantic analysis of names
Intelligent comparison of product names using natural language processing algorithms:
- Technologies: Lemmatization, stop word removal, semantic similarity calculation using TF-IDF or word embeddings methods.
- Application: Identifying duplicate products with different names ("Apple iPhone 15 Pro 256GB Smartphone" and "iPhone 15 Pro 256Gb Black").
- Accuracy: 85-92% with correct similarity threshold settings.
3. Analysis of categories and attributes
Multifactorial comparison by combination of characteristics:
Comparison parameters: - Product category - Brand/manufacturer - Model - Key characteristics (size, color, volume) - EAN/UPC/ISBN codes
This method is especially effective for products with clearly structured attributes (electronics, household appliances).
4. Machine learning and AI
Modern systems use machine learning models to identify new products:
- Training on historical data: The system analyzes patterns of past product additions and learns to recognize truly new items.
- Relevance Prediction: AI evaluates how well a new product fits your product range and the needs of your target audience.
- Automatic improvement: The models self-learn based on feedback (which products were added and which were rejected).
Automatic price comparison
An effective automatic comparison system includes several components:
Automatic download of price lists
Setting up sources
The system receives price lists through various channels: email (automatic attachment extraction), FTP/SFTP servers, supplier APIs, and web scraping. Multiple formats are supported: Excel (XLS, XLSX), CSV, XML, and JSON.
Data normalization
Bringing price lists to a unified format: unifying column structures, converting units of measurement, standardizing currencies, clearing data of unnecessary characters and formatting.
Field mapping
Intelligent recognition of price list columns: product names, SKUs, prices, inventory levels, and specifications. The system remembers each supplier's price list structure for subsequent downloads.
Identification of new products
Applying the search algorithms described above: SKU comparison, semantic analysis, and attribute matching. The result is a list of potential new products with a confidence score.
Update frequency
The frequency of price comparisons depends on the specifics of the business:
- Daily: For rapidly changing niches (electronics, fashion, products with a short life cycle).
- Weekly: Optimal for most categories, balance between relevance and system load.
- Upon receipt of the price list: Instant processing upon receipt of a new price list from the supplier.
- According to the schedule: Run the comparison at night to minimize the impact on site performance.
Advice:
Use a differentiated approach: check critical suppliers daily, and others weekly. This optimizes resources without losing data relevance.
Product selection criteria
Not every new product from a supplier's price list should be added to your catalog. An automatic filtering system should consider many factors:
Commercial criteria
Marginality
Minimum allowable markup: 20-30% for high-volume items, 40-60% for niche items. The system automatically calculates the potential margin based on the purchase price and average market prices (competitor parsing).
Average bill
Filtering by price range: exclude products that are too cheap (low absolute margins) or too expensive (long sales cycles, risks). The optimal average order value depends on your niche.
Availability in stock
Priority is given to items with confirmed supplier availability. Excludes "made to order" items with long delivery times (more than 7-14 days) if this doesn't fit your business model.
Minimum order quantity
Filtering out products with unacceptable MOQs (Minimum Order Quantities). For example, if a supplier requires an order of 10 units for a product priced at $500, this may not be practical for testing.
Marketing criteria
- Competitive analysis: Checking competitors' product availability. If the product is already sold on 10+ platforms, a price advantage is needed. If no one else has the product, it's a first-mover opportunity, but the risk is higher.
- Trending: Search query analysis in Google Trends and Yandex Wordstat. Growing interest is a signal to add. Integration with Amazon Best Sellers and eBay Trending for international markets.
- Seasonality: Accounting for seasonal factors (adding swimsuits in spring rather than fall). The system can use historical data on sales seasonality.
- Brand Compliance: The product must fit within your store's positioning. AI analysis can assess compliance based on your existing catalog.
Technical criteria
Description quality: Adequate product information (at least 200 characters, including key features) is essential. Products without adequate descriptions can create promotional challenges.
Availability of images: Filter products without photos or with low-quality images (resolution below 800x800px). High-quality visuals are critical for conversion.
Uniqueness of the article: Check for conflicts with existing SKUs in the database.
Correctness of categorization: The ability to accurately determine the category for placing a product in the catalog.
Scoring system
Instead of rigid filtering based on the "yes/no" principle, it is more effective to use a scoring model:
Example of calculating the final score (0-100): - Marginality (0-25 points) - Competitiveness (0-20 points) - Trendiness (0-20 points) - Data quality (0-15 points) - Availability in stock (0-10 points) - Compliance with assortment (0-10 points) Thresholds: - 80+ points: Automatic addition - 60-79 points: Adding for moderation - 40-59 points: Requires analysis - Below 40: Automatic rejection
Quick add to catalog
Once you've selected suitable products, you need to quickly add them to your catalog. Automating this process is key to scaling:
Automatic filling of product cards
Basic information
- Product name (with automatic SEO optimization)
- Article and SKU
- Category (automatic distribution by category tree)
- Brand and manufacturer
Pricing
- Purchase price
- Retail price (subject to markup rules)
- Promotional price (if automatic promotions are set up)
- Price for different customer groups
Description and characteristics
- Importing a description from a supplier's price list
- Automatic description enhancement with AI (expansion, SEO optimization)
- Parsing technical specifications into a structured form
- Generating SEO tags (title, description, keywords)
Images
- Downloading photos from the supplier's servers (using the URL in the price list)
- Automatic processing: resize, optimize, remove watermarks
- Creating thumbnails and additional sizes
- Adding your logo (if required)
Batch import
Batch processing is used for bulk adding of products:
- Forming a queue: All selected goods are placed in a priority import queue.
- Parallel processing: The system processes several products simultaneously (5-20 threads depending on server resources).
- Error handling: Products with import errors are marked for manual review without blocking the import of others.
- Logging: Detailed report on import results (successfully added, skipped, errors).
Speed optimization:
When properly configured, the system can import 100-500 products per hour. Critical factors include connection speed to suppliers' servers, your server's performance, and the complexity of image processing.
Integration with Elbuz
The Elbuz platform provides a ready-made solution for automatically searching and adding new products:
- Automatically download price lists from 1,000+ suppliers
- Intelligent product matching (96% accuracy)
- Customizable filtering and scoring rules
- Batch import with automatic filling of product cards
- AI-powered generation of improved descriptions and SEO metadata
- Integration with popular CMS and marketplaces
Testing new products
Adding a product to the catalog is just the first step. Its effectiveness must be systematically tested:
A/B testing
Experimental approach to the implementation of new products:
Price testing
Launching a product with different prices for different audience segments. Analyzing the price point that achieves the optimal ratio of margin to sales volume. Test duration: 2-4 weeks.
Testing descriptions
Comparison of the effectiveness of different product description options. Option A: technical description with specifications. Option B: emotional description with customer benefits. Success metric: conversion to purchase.
Image testing
Different main photo variations: product on a white background vs. product in an interior setting. The impact on CTR and conversion can be significant (up to a 50% difference).
Placement testing
Experiment with categories: sometimes a product sells better in an unobvious but more popular category. Also, test placement in recommended blocks.
Key metrics for analysis
- Views: The number of times the product page was opened during the period. Low views may indicate issues with search or placement.
- CTR (Click-Through Rate): The percentage of product clicks out of the number of impressions on category and search pages. Reflects the attractiveness of the image and price.
- Cart conversion: The percentage of product page visitors who added it to their cart. This is an indicator of the quality of the description and the price/value ratio.
- Conversion to purchase: The final metric is the percentage of purchases from the number of views. Target: 1-5% depending on the niche.
- Refunds: Return rate for new items. A high rate (more than 5-10%) indicates quality issues or a discrepancy with the description.
- Reviews and rating: Early reviews are critical. Monitor them and respond promptly to any negative feedback.
Testing period
Recommended timeframes for evaluating the effectiveness of new products:
- Express assessment (7-14 days): The initial assessment of interest is whether there are enough views and additions to cart.
- Full cycle (30-60 days): Collection of complete statistics on sales, repeat purchases, and returns.
- Seasonal products: Evaluation throughout the season (3-6 months).
Success criteria
Based on the testing, a decision is made on the future fate of the product:
Successful product (keep in stock):
- Conversion rate is 20%+ higher than the category average
- Marginality is in line with the planned level (deviation no more than 10%)
- Positive reviews (average rating 4.0+)
- Low return rate (less than 5%)
Promising product (requires optimization):
- High views but low conversion - optimize price or description
- Low views but high conversion – improve SEO and promotion
Unsuccessful product (remove from catalog):
- Conversion rate is 2x lower than average for 60+ days
- Regular returns and negative reviews
- Actual marginality is 30%+ below target
International experience
Large marketplaces like Amazon and eBay use advanced algorithms to predict product performance. Amazon uses machine learning to predict sales of new items based on the analysis of millions of products. eBay launched the "Trending on eBay" program, which automatically identifies trending products and promotes them on the homepage. These practices can be adapted for your own store.
Conclusion
Automating the search and addition of new products is a critical process for e-commerce business growth in 2025. Key findings:
- Systems approach: Use a combination of new product discovery methods, from simple SKU comparisons to AI analysis. Don't rely on a single algorithm.
- Smart filtering: Implement a scoring system to automatically select truly promising products. This will save time on manual review and improve the quality of your product range.
- Quick import: Automate the entire cycle from product discovery to catalog publication. Speed to market with new products gives you a competitive advantage.
- Continuous testing: Treat every new product as an experiment. Systematically analyze metrics and optimize based on them.
- Use ready-made solutions: Platforms like Elbuz allow you to implement a full automation cycle without developing from scratch, saving months on integration.
A properly configured automatic search system for new products allows you to add hundreds of high-quality products monthly with minimal effort. This creates the foundation for sustainable sales growth and business scalability.
Need help setting up?
Explore the complete guide to assortment and inventory management for a deeper understanding of all processes.
Platform Elbuz Provides a turnkey solution for automating supplier price list management, including intelligent search for new products and product import.
Save a link to this article
In the world of automation, I am a translator of ideas into the language of effective business. Here, every dot is a code for success, and every comma is an inspiration for Internet prosperity!
Discussion of the topic – How to automatically find new products in supplier price lists
How to automatically find new products in supplier price lists
There are no reviews for this product.
Write a comment
Your email address will not be published. Required fields are checked *