- That Japan's Consumer Affairs Agency has answered the placement question directly: a disclosure parked in a reply thread will usually go unnoticed, and a label shown only at the start of a video can be missed by anyone who joins partway
- Why ticking YouTube's "paid promotion" box and satisfying the Japanese rule remain two separate jobs, given that the platform's automatic notice runs for the first ten seconds
- That posts you write of your own accord are explicitly outside the rule, while posts from years ago that are still on display are explicitly not
Quick answers
- Q: If my bio says the account contains PR, can I skip labelling individual posts?
- A: Do not build on that assumption. The operational guideline says whether a business's involvement is clear is judged from the display as a whole, and social posts reach people one at a time — through recommendations and quote posts, from readers who never open your profile. The Consumer Affairs Agency Q&A addresses the closest case: where "advertisement" appears in a reply thread rather than the post body, ordinary consumers will often fail to notice it, so in many cases it will not count as clear.
- Q: Do I have to go back and label posts from several years ago?
- A: The Q&A says that a post amounting to a business's own representation, made before the rule took effect on 1 October 2023, can still be caught if it remains on display afterwards and is not clearly labelled — with administrative action a possibility. It then asks influencers to cooperate with the advertiser. The party facing that action is the advertiser, not you, which is why old posts are better handled together with whoever issues your links rather than decided alone.
This article walks through what Japan's Consumer Affairs Agency has published — the designation, its operational guideline and its Q&A — plus what the platforms say in their own help pages, applied to social posts that carry commission-bearing links. It is informational and educational, not legal advice, and following it does not make anything lawful. Whether a disclosure is clear is judged case by case from the display as a whole, and the programme you have joined and the platform you post on may both require more. For your own situation, read the Consumer Affairs Agency material and consult a professional where needed. The framework itself — what the law covers, what counts as a misleading claim — is set out in our separate guide to Japan's stealth-marketing rule. This piece is about placement only. Note that the Consumer Affairs Agency pages linked here are in Japanese.
The link on a three-year-old post is still live
A notification took me back to one of my own posts. Someone had quoted it, and it turned out to be three years old. The comparison I was using at the time, the link I had attached — all of it still rendering. I tapped the link. It still worked.
The version of me who wrote it had not put "PR" anywhere. Not out of defiance; I was pushing out twenty of those a day and could not have told you what any single one of them said.
What makes this awkward is that the post is not filed away as history. Quote it and it shows. Surface it in search and it shows. Scroll far enough down the profile and it shows. A blog post can be quietly unpublished. A social post has no middle setting — it stays up or it comes down.
So this article covers one question only: for a post carrying a commission-bearing link, which surface, which spot, and in what form. The wider framework — what the law prohibits, what makes a claim misleading — lives in a separate piece. Here I am staying with the parts where the regulator has answered the placement question itself.
The unit being judged is one post, not the account
One piece of groundwork first. The designation that took effect on 1 October 2023 (Cabinet Office Notification No. 19 of 2023) covers, in the words of the notification itself, a representation made by a business about goods or services it supplies where it is difficult for ordinary consumers to tell that it is such a representation. The party the rule is addressed to is the advertiser — not the influencer or affiliate who posts on request. That structure is covered in another article, so I will not repeat it.
What matters for placement is the unit of judgement. The operational guideline says the question of whether a business's involvement is clear to ordinary consumers is decided from the display as a whole.
That is where websites and social platforms part company. The guideline carries this note about comparison affiliate sites:
On an affiliate site comparing prices or details of multiple goods or services, so long as the site itself makes clear to ordinary consumers that this is a business's representation, consumers will not mistake it for a third party's, and it is therefore unnecessary to state that for every single item listed.
A site can rely on having one front door. Readers arrive through the top page or the first screen of an article and go inward from there, so establishing it once, site-wide, is enough. Nobody reaches item seventeen without passing through.
Social platforms break that assumption. A single post drops into the feed of someone who does not follow you. A reader from the recommendation panel never opened your profile; a reader who arrived through a quote post has seen the original body text and nothing else. In practice it is safer to treat the first screen a reader touches as the unit that gets judged.
That first screen is what I will call a surface: one post in an X timeline, a YouTube watch page, a note article page, a single short-form video. Change the surface and you change both where the reader looks first and how much room you have there.
X: put it in the body, not down in the replies
For X, the Consumer Affairs Agency Q&A answers the question outright. Q14:
Where "advertisement" or similar wording appears in the thread (a reply) rather than in the body of the post, ordinary consumers will typically be at risk of not noticing it, and in many cases it will not be considered clear that this is a business's representation.
That lands hard on a common workflow. Links in the body suppress reach, so the body carries the prose and the link hangs off a reply. This article is not arguing against keeping links out of the body. The point is narrower: where the link sits and where the disclosure sits are two separate decisions. Move the link into a reply if you like. Move the disclosure with it and you have built exactly the arrangement Q14 describes.
For wording, start with what the guideline actually lists: "advertisement", "publicity", "promotion", "PR". Or describe the relationship in a sentence — the guideline's own example is along the lines of "posting after receiving a product from Company A". Q19 in the Q&A offers a paid-request version: "posting at the request of Company Z".
Two caveats travel with those words. Q11 confirms that nothing outside that list is forbidden. The guideline's own note runs the other way: using those words does not settle it either, because the display as a whole may still fall short. What is being assessed is not whether you wrote it but where you put it and how big it is.
The failure modes are listed too. Smaller than the surrounding text. A paler colour. Buried inside a long passage. Tucked at the end where it is hard to see. And, named specifically for social platforms, sunk into a wall of hashtags.
Neither the guideline nor the Q&A appears to address what happens when someone quote-posts your post, or when you repost somebody else's sponsored content. I would not read that silence as permission. What Q14 objects to is the risk of the reader not noticing, not the reply feature as such. A reader who arrives through a quote and sees only the original body is in the same position. That is my own application of the reasoning, not a view the agency has expressed. Practically, putting the disclosure where it cannot be stripped off — the opening of the body — saves you from having to think about it later.
X's own policy raises a separate issue
Alongside the Japanese rule, X publishes a paid partnerships policy in its help centre. Affiliates rarely read it, and it matters twice over.
First, X's own definition of a paid partnership includes cases where the goods or services earn you a commission through sales via affiliate links or discount codes. On X's terms, then, a post carrying a commission-bearing link is inside that definition. The policy goes on to require that posts created as part of a paid partnership and published as organic posts clearly and conspicuously carry something signalling commercial content — "advertisement", "promotional content" and the like. Same direction as the Japanese rule.
Second, and this is the part worth your attention: the policy also lists prohibited industries, and among them is financial products, services or opportunities — loans, investment services, cryptocurrency, buy-now-pay-later and other finance-related content. Gambling appears as its own entry. On enforcement, the policy mentions being required to delete the post, a read-only period, and account suspension for repeat breaches.
This is X's house rule, not law, and the policy notes that exceptions may be considered case by case on request. So I am not going to tell you your posts are in breach. What I will say is that if you promote anything financial on X, that list is worth reading with your own eyes, separately from anything the Consumer Affairs Agency says. Search the X help centre for "paid partnerships policy". Policies change, so read the current text rather than my summary of it.
Video: the guideline says the opening alone is not enough
Video is the strictest surface of the set. The guideline gives this as an example of a disclosure that is not clear:
Where, in a video, the indication that this is a business's representation is shown for a period too short for ordinary consumers to register (including, in a long video, showing it only outside the opening — in the middle or at the end — or otherwise only in places ordinary consumers are unlikely to notice).
Q15 in the Q&A then pushes further. Asked whether putting "advertisement" or "this video contains promotion" at the start is sufficient, it answers that viewers may join partway through, so an opening-only label risks being missed and may in some cases fall short — and that keeping "advertisement" on screen throughout, so the video is clear from end to end, is desirable.
Now hold that against the platform. YouTube provides a way to declare paid promotion, and its official help page sets out the steps: in YouTube Studio, open Content, select the video, go to "Show more" and tick the box saying the video contains paid promotion such as product placement, sponsorship or endorsement. According to the same page, ticking that box makes a disclosure message appear automatically for the first ten seconds of the video.
Line those up. YouTube's automatic notice covers ten seconds at the start. The Q&A says the start alone risks being missed. These are not in conflict — they answer different questions. One is a declaration to the platform; the other is Japanese law. YouTube says as much itself: creators and brand partners are responsible for understanding and complying with local legal obligations on disclosing paid promotion, and the page closes by stating that the help article is not intended as legal advice.
In practice: tick the box, then look at how long your video is and how it gets watched, and decide how far the on-screen label should run. On a short-form video the runtime is small enough that keeping a caption up to the end costs almost nothing. On a long video where a permanent label would grate, at minimum put it somewhere beyond the opening as well.
One more thing that gets missed on short video: disclosing by voice alone does not reach anyone watching muted. Neither the guideline nor the Q&A names this case, so I am not stating it as a rule — but since clarity is assessed from the reader's side, putting it on screen as text as well seems like the obvious move.
note and profile fields: outside the wall, and off the path
A note article behaves like a blog post: first screen, not smaller than the surrounding text, not a paler colour, not buried in a long passage, not stranded at the foot of the page. All the same items from the guideline.
What is specific to note is paid articles. Set a post to paid and non-purchasers see only the free portion. Put the disclosure inside the paid portion and readers deciding whether to buy have not seen it. The people the rule is concerned with are ordinary consumers, not purchasers, so the disclosure belongs outside the wall. Neither the guideline nor the Q&A addresses paid articles directly, so this is my reading rather than a stated position.
Q13 in the Q&A gets close to it. Asked whether a line at the top of an affiliate site — something like "this site uses affiliate advertising" — makes things clear, it answers that even with such a line, clarity has to hold once font size and colour are taken into account. The question is less about the sentence existing than about whether the place it sits is somewhere readers actually pass.
Profile fields and pinned posts run on the same logic. A line in your bio saying the account carries PR is far better than nothing. It only works, though, on people who open the bio. Someone who meets one post in a recommendation feed, taps the link and leaves never went near it.
Relatedly, Q12 answers that merely stating an affiliation — "I work at Company X" — does not necessarily make it clear that this is the company's representation. Stating your position and making the post readable as advertising are different things. A bio line may well be treated the same way, and it is worth planning as if it will be.
The guideline says representations made through a business's own social media account fall outside the designation, because it is socially obvious whose account it is. That is about the advertiser's own official account. It does not mean a line in the bio of a personal account promoting somebody else's products gets the same treatment.
Every time? The exemptions are written down
This is where the most confusion sits. For a while the message circulating was that anything you post about a product needs "PR" on it. The Consumer Affairs Agency does not say that. Q21 is exactly this question.
Where a post is made on your own initiative and not in response to a business's request, it does not amount to a business's representation, and there is no need to write "PR" or similar.
Q18 runs the same way. Having once posted a promotion for Company Y's product C on request does not mean that buying product D yourself, liking it and posting about it needs a label — that post is not the business's representation. Doing paid work in the past does not convert every subsequent post on the account into advertising.
So where does a commission-bearing link sit? Closest is Q19. Someone offered payment in exchange for posting their impressions of a product, and accepted. Even without any instruction on content, the Q&A says the consideration and the relationship are weighed together and the post is likely to amount to a business's representation — and that wording making that clear is appropriate.
Put together, the axis is not the subject matter but whether consideration or a request is attached to that post. Which means "every time?" is answered per post, not per account. Posts carrying a commission link get the label. Posts you wrote off your own bat, with nothing attached, do not.
Q17 shows the fuzzy edge: where a product was given free of charge, it turns on the specifics. Someone who already used and liked the product and always intended to post is not thereby required to label it — but if the post shows the item that was supplied, noting that it was supplied is described as useful. For borderline posts, constructing a case for not labelling usually costs more than labelling. That is where I would land in practice.
Three exits for old posts you cannot edit
Back to the three-year-old post. Q20 in the Q&A responds to an influencer told, years later, to add "advertisement" to a post made on request. The answer: where a post amounting to a business's representation was made before the rule took effect but remains on display afterwards, it can be caught as an unfair representation in breach of the designation, with administrative action a possibility. The closing line asks for cooperation with the advertiser.
The party exposed to that action is the advertiser. That is presumably why the sentence is phrased as a request for cooperation rather than an instruction. Read the other way, cleaning up old posts was never meant to be a solo job — better decided with whoever issues the links.
Then the surface-specific bind. On a blog you open the post and add a line. Where posts cannot be edited after publication, you cannot. Three exits:
| Situation | What you can do |
|---|---|
| Body text is editable | Add the disclosure at the top. Not the foot — the guideline lists the end position as hard to see |
| Body text is not editable | Take the post down. There is no middle setting |
| You would rather not delete it | Disabling the commission link is an option — but you cannot conclude that removing a link puts the post outside the rule |
That third row needs its footnote. What the guideline looks at when deciding whether something is a business's representation is not whether a link still resolves; it is the substance — whether there was a request, what consideration passed, how the relationship has run. So "I removed the link, so it is fine" is not a conclusion to reach on your own. That one goes to the advertiser too.
On sequencing: working through everything from the top is not realistic. Start with what is still in circulation. Pinned posts. Posts still pulling traffic from search or recommendations. Posts near the top of your profile listing. Those three alone cover a large share of what people are actually seeing. That is a working order, not a rule.
"Labelling kills performance" needs splitting by surface
Finally, the question everyone actually asks. Does writing "PR" cost you conversions?
I will not claim it does not. I do not have a like-for-like test of labelled against unlabelled under matched conditions. I will not claim it does either — same reason. What I can offer instead is a way of looking at it.
If it does cost you, what you are losing is clicks from people who would not have clicked had they known it was advertising. Those people work out that it is advertising on arrival anyway, so they were never likely to sign up. Which means clicks can fall visibly while signups barely move. To judge it at all, you have to look at clicks and signups separately.
And the friction is not constant across surfaces. Four characters at the top of an X post and a label held on screen through a whole video cannot possibly affect the experience the same way. So do not generalise "labelling costs performance" across surfaces. If something dropped, work out which surface and which placement it dropped on, or you have no next move.
Besides which, the decision here was never whether to label. Labelling is the premise. What you are deciding is placement.
Does the offer still read well once you label it?
An offer that gets harder to explain the moment you write "advertisement" on it was always going to be hard to write about. Kingfin's affiliate programme is free to join and free to use, and the commission terms are visible to you in the dashboard. You can check for yourself whether the copy holds up on plain facts, without exaggeration. Whether any commission arises, and how much, is not guaranteed. Trading with the service being promoted carries the risk of losing your principal.
Sign up freeSix checks before you post
Turned into something you can run just before publishing. Some of it comes straight from the published material; the rest is my own operating judgement, flagged where it applies.
The morning I found that three-year-old post, the first thing I did was start scrolling back through everything. I gave up half a day in, because the order was wrong. Only a fraction of it is still being seen, and that fraction was where to start. Knowing where your posts are still on display probably comes before knowing the rules about labelling them.
Frequently asked questions
Disclosure: This article is informational and educational content from Kingfin's English editorial team and does not constitute legal advice. Statements about Japan's Act against Unjustifiable Premiums and Misleading Representations and about the designation "representations that make it difficult for ordinary consumers to identify them as representations by a business" (Cabinet Office Notification No. 19 of 2023) were checked against Consumer Affairs Agency publications — the notification, its operational guideline and its Q&A on stealth marketing — at the time of writing, but legislation is amended and published material is updated. Whether a disclosure is clear is judged case by case from the display as a whole, and nothing here guarantees that following it makes a representation lawful. Platform terms and the rules of the affiliate programme you have joined may impose more, and both change without notice. For decisions about your own wording or campaigns, read the Consumer Affairs Agency material and consult a professional where needed. Kingfin's affiliate programme promotes OlympTrade, an FX and binary-options trading service that is not registered as a financial instruments business in Japan. Trading always carries the risk of losing your principal, and we cannot promise that anyone will definitely earn or that gains are guaranteed. Affiliate commissions are likewise not guaranteed and results vary from person to person; only ever trade with money you can afford to lose, and make your own decisions at your own responsibility.
