Korean marketers spent the start of October arguing about two things. One is whether Naver quietly patched blog search again. The other is whether Naver’s own ad products are really so different from the “search abuse” it accuses Toss of. Neither question has an official answer yet, but both are new since our last Naver community recap on AI Briefing citations, and both matter if you report Naver performance to a client.
“They patched Latest again”
The blog-visibility grumbling has been building since September. A column published on 5 October by Kim Min-ji of the agency Sungo Marketing Lab (syndicated on News N Yonhap) pulls together what practitioners have been sharing on i-boss and elsewhere. The cases include:
- posts that used to appear suddenly disappearing;
- posts visible under Latest (최신순) but missing from integrated search (통합검색);
- new posts indexed differently from before, or appearing for a while and then dropping out;
- older posts ranking higher than expected.
On 1 October, the column notes, an i-boss thread went up under the title “최신순 또 패치했네요”, roughly “they patched Latest again”. This isn’t a one-off complaint. Back in May, an i-boss user asked why posts ranked first or second in the Blog tab but didn’t show in integrated search. That symptom keeps coming back.
The column is careful about one point, and so should you be: Naver hasn’t announced a blog search algorithm change for any specific date. What exists is observed volatility that practitioners call a patch. Naver’s help centre says relevance and Latest sorts use different logic. Relevance blends time, content quality and popularity, not keyword counts. New or edited posts can take up to 24 hours to appear, and Naver doesn’t guarantee any individual post’s exposure. Spam, mass-produced posts, write-ups with no real experience behind them and keyword stuffing can all be restricted.
The practical advice is the right advice. Before you call a blog 저품질 (“low quality”, the Korean shorthand for a penalised blog) and start rewriting or deleting posts, log publish time, visibility after 24 hours, and visibility under Latest, Relevance and integrated search, keyword by keyword. Rewriting everything the morning after a drop just destroys the evidence.
Naver vs Toss: where does “search abuse” start?
The second thread is louder. Naver asked police in February to investigate fintech app Toss for business obstruction and breach of the Network Act. The target is Toss’s rewarded “one-week visit mission” product, which pays users points (about five won) to search a keyword on Naver and prove they did. Naver says this inflates search volume regardless of real interest and can distort Place and Store results. Yonhap reported the row escalating in mid-September.
Then on 5 October, Kukmin Ilbo published an exclusive on Naver’s own CrossMedia product. It uses TV, digital and outdoor ads to get consumers to search an advertiser’s chosen keyword on Naver. Naver’s case studies tout search-volume (QC) increases of 2,156% for a confectionery brand and 6,543% for a health-supplement brand.
Naver’s answer is that rewarded marketing “induces artificial consumer behaviour, from entering the search term to proving the result, in exchange for money”. CrossMedia, it says, simply invites interested consumers to search for themselves. Toss has asked Naver to show evidence of actual ranking distortion. Naver says that is a trade secret.
What the community is taking from it
- Search volume isn’t a clean metric. If both paid missions and TV campaigns can move a keyword’s QC by thousands of per cent, treat volume spikes in keyword tools with suspicion before you build a content plan on them.
- Behavioural signals are the battleground. The row lands just as Naver Place starts ranking restaurants on saves, routes, bookings and reviews. Whatever line Naver draws on Toss will tell you how it plans to police those signals too.
- Don’t chase phantom patches. Without an official note, the right response to blog volatility is measurement, not panic.
Our take: Naver’s position is legally arguable but rhetorically weak. The difference between “we pay people five won to search” and “we pay TV networks so people search” is real, but both turn the search box into a marketing KPI. The more Naver’s own rankings depend on behaviour, the harder it gets to say whose behaviour counts as real. Korean SEOs should watch the police case closely. Its outcome may say more about Naver’s ranking philosophy than any Search Advisor guideline.