After completing the distribution analysis of the review ratings, you can start the distribution analysis of the number of reviews. The following will take the review rating number data in the “cleaned product portrait data” Excel table as an example. Select the “Rating Quantity” column in the table and insert a line chart through the Excel chart (the specific operation is similar to the operation of inserting a chart).
The operator can understand the fluctuation trend of the number of review ratings from the rating number line chart. Overall, the number of reviews for highly ranked products (products that appear in the first few pages of the search page) is generally higher (more than 5,000). And the review distribution also reflects the historical sales distribution to a certain extent (because reviews in the same category come from a certain proportion of sales, and the review conversion ratio of similar products will not be very different). Operators will find that even if some listings have nearly 5,000 reviews (ranked around 4,000, 7,000, and 10,000- and above), their exposure ranking values are high, which means that under the keyword “dress”, the products have a short life cycle, that is, the competition between products is fierce, because some products have high historical sales (high number of reviews) but the exposure ranking is not ideal, which is consistent with the experience of most apparel industry practitioners. Therefore, using product review quantity distribution analysis can help operators analyze the overall life cycle of products under different keywords.
In addition to judging the life cycle, review quantity distribution analysis can also help operators judge the differences in consumer demand in different categories. In order to contrast with the search results of the keyword “dress”, the keyword “compression springs” is used for search and review quantity distribution analysis.
After the review quantity of all listings under the search results of “compression springs” is captured by a crawler program, it is stored in an Excel table called “compression springs Listing Catch” (readers can download and view it by themselves), and the data in the “rating quantity” column in the table is visualized.
Compared with the review distribution under the search term “dress”, the review distribution under the search term “compression springs” does not have an obvious “28 distribution” phenomenon. The review quantity distribution is more like an “average distribution”. There are both listings with high review quantities and listings with low review quantities in different search ranking intervals.
The reason for this distribution difference is the difference in consumer demand. For “dress” related products, consumer demand is large and mainly “emotional”, that is, most consumers will buy the current hot-selling products that they think are suitable; while for “compression springs”, the demand is relatively high, and the demand is mainly “emotional”, that is, most consumers will buy the products that are currently popular and suitable for them; For products related to “springs”, there is less demand, and the demand is mainly “rational”, that is, most consumers will carefully look for matching products based on their own needs for hardware products, so they will be more patient to turn pages to find products of interest, and the review distribution generated tends to be more evenly distributed.