Naver Shopping Best Seller Ranking, Keyword Trend & Brand Chart

Naver is where South Korea shops. This Actor reads Naver Shopping's own published charts and hands them back as clean rows: the products Koreans viewed and bought most, the search keywords that are climbing, the brands they are looking for, and the themed collections Naver builds every day. Korean titles are kept exactly as Naver writes them, prices are in won, and every row carries the date Naver computed it.

It is ranking and trend data. Read the What this cannot do section before you buy, because it is short and it matters.

What you get

Product charts. Up to 100 best sellers per run, ranked either by how many people viewed the product or by how many bought it, daily or weekly. Each row carries the Naver product id, the Korean title, list price and sale price in won, the discount percent, the star rating, the review count, the store selling it and that store's link, the delivery charge, and whether the listing carries Naver's arrival guarantee.

Eleven categories. Everything combined, or one of fashion clothing, fashion accessories, cosmetics and beauty, digital and appliances, furniture and interior, maternity and baby, food, sports and leisure, life and health, or leisure and convenience.

Twelve audience segments. This is the part you will not find elsewhere. The same chart can be ranked for men in their twenties, women over fifty, or any of the ten other age-and-gender bands Naver tracks, as well as everyone at once. The top-selling product for men in their twenties is usually nothing like the top-selling product overall, and each segment is one run.

Trending keyword charts. Twenty rising Korean shopping searches per chart, with the rank, Naver's own movement label (NEW, UP, SOAR, DOWN, STABLE), the full category path Naver files the keyword under, and related search terms. Three different orderings: most searched, rising fastest, and newly charting.

Keyword history back to February. The newly-charting keyword chart has roughly seven months of weekly snapshots and eight monthly ones. Ask for a past date and you get that week's chart. Ask for a date Naver does not hold and the run tells you so, and names how many dates it does hold, instead of handing back an empty sheet.

Brand charts. The twenty brands Koreans search for most, on three separate charts, with related search terms for each brand.

The products behind a keyword or a brand. Turn on one checkbox and every keyword or brand entry is expanded into its top twenty products, so a single run turns twenty trending searches into four hundred ranked, priced products.

Today's themed picks and the four vertical lists. Naver's daily curated collections, each with its Korean theme name and the copy Naver wrote for it, plus the fashion town, best deals, arrival-guaranteed and shopping-live top twelves.

What this cannot do

Three limits, stated plainly, because a product sold without them would be sold dishonestly.

It is charts, not search. There is no way to type in an arbitrary keyword or a product code and get matching listings. The keyword charts cover keywords that are already trending; the product charts cover whole categories. If your question is "find every Naver listing for this SKU", this Actor is the wrong tool and no amount of configuration will change that.

The prices are chart prices, not checkout prices. Every row carries data_date, and on the ranking charts that date is one or two days behind. Some surfaces are labelled with the current Korean date, but on every product that appears in both a dated chart and a current-dated surface the price was identical, so the fresher label has not been shown to mean a fresher number. Treat every price here as the price Naver charted, not the price at the till. If you need a live checkout price, take the product URL from a row and check it yourself.

Review counts stop at 99,999. Naver does not print an exact figure above that; it prints 99,999+. Rather than hand you 99,999 and let you believe it is exact, those rows carry an empty review_count, the raw string in review_count_display, and review_count_capped set to true. Six of ninety-nine rows on the all-category chart were capped when this was measured, so it is not a rare edge case.

One more thing worth knowing rather than discovering: rank_change_reported is Naver's own movement number and it does not survive arithmetic. Checked keyword by keyword against the immediately preceding chart on Naver's own list of dates, four of fifteen agreed and eleven did not, including a keyword that climbed from second to first and reported no movement at all. It is passed through unchanged because substituting our own figure under Naver's name would be worse. Read it as direction, not as a measurement, and compute your own movement from two dated runs if you need it exact.

Reading the output

Each row carries row_type (product, keyword or brand) and source, which says which chart produced it. Columns are not padded: a chart that does not carry review data produces rows with no review columns at all, rather than rows with empty ones, because an empty column reads as a fact about the product when it is really a fact about the chart. Check source before assuming a column exists.

The run summary in the key-value store separates the outcomes that look identical in an empty dataset. It says whether the request was refused before it was sent (and why), whether Naver declined to serve the chart, or whether Naver served a valid chart that happens to be empty today. It also reports the dates the data was computed, how many review counts were capped, and how many rows Naver marked as paid placement.

Who this is for

Korean market analysts tracking what is selling and for whom. Brands watching their own rank against competitors week by week. Sellers hunting product ideas by reading what is climbing in their category. Trend researchers who want Korean demand data segmented by age and gender, which is hard to get anywhere else. Agencies building Korea market reports who need the chart, the date it was computed, and an honest note on what the number means.

Notes

Prices are in Korean won. Korean text is passed through exactly as published, not transliterated or translated. Charts refresh daily; the same run tomorrow gives a different chart, and storing yesterday's run is how you build a movement series.