Definition
Review aggregation is the data-engineering work of collecting customer reviews from the places they live and landing them in a single store the rest of a feedback system can query. For consumer brands, the sources are mostly retail: Amazon, Walmart, Best Buy, Costco, Lowe's, Target, and Home Depot, plus the thousands of Bazaarvoice-powered retailer and brand pages that syndicate reviews across the web. Each source has its own structure, identifiers, and access model.
Aggregation is more than scraping. A working aggregator handles per-retailer identifiers (ASIN, Walmart Item ID, Best Buy SKU, Costco Product#), schema differences (star scale vs. 5-star, title vs. body, photo attachments), rate limits and politeness rules, syndicated duplicates (the same Bazaarvoice review appearing on multiple retailer sites), language detection, and incremental updates so a new review posted this morning shows up in the store today. Verified-purchase flags, reviewer metadata, photo and video attachments, and helpful-vote counts each have to be preserved, not flattened, if downstream analysis is going to mean anything.
Why it matters
For a consumer brand, the review corpus is the largest piece of structured feedback available — often dwarfing tickets, surveys, and returns combined. A product team that reads only Amazon reviews is missing the Walmart audience, the Costco audience, and the retailer-specific patterns that reveal channel issues. A brand that reads only its own e-commerce reviews is missing the vast majority of what customers write.
Aggregation is also where most review projects stall. Maintaining 20+ ingestion paths, handling retailer schema changes, and keeping identifier maps fresh is a full-time engineering effort. Most brands do not have that capacity in-house, and most feedback tools stop at one or two sources. The result is a partial view dressed up as a complete one — and product decisions made on the subset of feedback that was easiest to collect, not the subset that mattered most.
Example
A sporting-goods brand sells the same treadmill on Amazon, Walmart, Best Buy, Dick's, and Academy. Amazon shows 1,247 reviews under one ASIN. Walmart shows 412 on an Item ID. Best Buy runs Bazaarvoice and shows 188, many of which are syndicated copies of reviews originally posted on the brand's own Bazaarvoice-powered site. Dick's and Academy each show a handful under their own product IDs. A working review-aggregation pipeline pulls all five sources daily, normalizes the schema, deduplicates the Bazaarvoice syndicates, resolves every identifier to one SKU, and lands ~1,800 unique reviews in the Indellia store. The product manager queries by SKU, not by source. Retailer-specific slices are still available — Costco buyers may comment differently than Amazon buyers — but the default view is the product.