Blibli Scraper for Product, Price and Seller Data

Blibli product data fields we extract

A Blibli scraper works on two levels because the catalogue does: ps-- is the product, is-- is one item, and the pickup point code decides which price and stock you see. Item RIS-70054-00084-00003 belongs to product RIS-70054-00084 and merchant RIS-70054, and search cards join all three into one default key shaped like RIS-70054-00084-00003-PP-3156917.

• Identity: product SKU, item SKU, merchant code, pickup point code, the internal MTA product and item codes, EAN where a seller supplied one, and the url-friendly name in the address.
• Blibli price data: offered price and listed price in rupiah, list discount and total discount, the struck-through card price, discount percent, minimum price and the flag that says variants span a range.
• Installments: starting monthly amount, interest rate and the Cicilan 0 percent label, with tags separating card and non-card plans.
• Warranty: type, duration and the sentence shown to shoppers, for example Garansi Resmi stated as one year.
• Availability: stock quantity, the threshold that flips a listing to limited stock, ready stock against pre-order, out-of-stock and coming-soon states.
• Fulfilment: delivery and pickup availability, how many other pickup points hold the same item, Click and Collect tags, and the ships-from province, city and postal code.
• Reception: rating with its decimal value, review count, seen and sold counters, and the rounded sold range cards show as 2,1 rb or 2.1 k.
• Merchandising: campaign name and code such as Flash Sale, promo end time, seller voucher count, wholesale minimum quantity and discount percent.
• Content: title, brand and whether that brand is official, image list, unique selling point bullets, the HTML description and specification rows such as Tipe Garansi or Kelengkapan Paket.
• Taxonomy: three parallel category paths - sales, master and multi-sales - with level, id and url.

Variants arrive as attributes with image swatches, usually Warna for colour, and as an option list where each option carries its own pickup point and availability flag. We extract Blibli data at that level, so one row is one item, in one place, at one price.

Blibli seller data, warranty flags and pickup points

The seller is a first-class object, not a name string. Each store carries a merchant code, a page at /merchant/name/CODE, an official flag, a top-rated badge in the Diamond, Gold, Silver and Bronze scale, a merchant score in percent, and the components behind it: positive reviews, on-time fulfilment, active response and successful transactions. Alongside those sit the store city, the join date, opening and closing hours per weekday, an open or closed status, and an international flag for cross-border sellers, whose listings a facet labels as shipped from abroad.

Reviews are structured. Every review carries a star value, free text, a verified-buyer flag, edit history, photos, the variant that was bought, and canned reasons such as ORIGINAL_PRODUCT or SHIPPING_SAFE_PACKAGING, while the summary block returns the one-to-five star histogram and sort orders for popular, latest and oldest. Reviewer names are personal data, so we leave them out unless a client has a lawful basis for them.

Coverage is a question of slicing. Search answers 40 products at a time and returns page and item totals with every response, so a large category is covered by cutting it along its own facets - category tree, brand, price band, store location, discount tier, warranty type and warranty length from under six months to over three years - rather than paging until the tail runs out. Blibli also publishes its own map: in early 2026 the product sitemaps ran to 383 files of about 10,000 addresses, keyword pages under /jual/ to 507 files, store pages to five files of about 5,000, plus a separate list of 176 brand flagship stores.