“V me” still violates the rules: 3 Chinese platforms explicitly ban homophone contact tricks

Here is the short answer first: writing your WeChat ID or other contact details as “V me,” initials, homophones, or obfuscated variants does not make the content safer. On platforms such as Zhihu, Douyin, and Kuaishou, public rules already say that homophones, near-sound variants, pinyin, numbers, and symbols used to disguise contact details are still prohibited. Baijiahao does not summarize it in exactly the same language, but its public enforcement examples already include variants such as “威信” and “V信,” which points to the same principle: the platform is not only matching exact strings, it is policing off-platform contact diversion itself.

This is one of the easiest mistakes for content teams to make. People assume that if they avoid writing a standard WeChat ID, the post no longer counts as lead generation. So they replace it with phrases like “V me for the template,” “my profile has it,” or “check the comments for the way.” From the platform’s perspective, though, those are not safer expressions. They are classic evasion patterns. When OmniGoAI’s OmniPost prepares platform-specific rewrites, one of the most common cleanup steps is to remove these disguised contact CTAs entirely instead of gambling on whether a platform misses them this time.

Why platforms still treat contact variants as violations

Because the real enforcement target is not a specific four-character string. It is the act of moving users toward off-platform private contact or conversion.

If a rule only banned the exact word “WeChat,” creators could instantly bypass it with homophones, initials, pinyin fragments, or symbols. That would make anti-diversion rules meaningless. So more platforms are now spelling out the broader target directly:

1. no direct contact details;
2. no equivalent carriers such as QR codes, URLs, or email addresses;
3. no pinyin, homophone, number, or symbol variants used to dodge moderation;
4. no hiding those clues in avatars, profile bios, titles, images, or comments.

For anyone running a distribution workflow, the operational takeaway is straightforward: changing the spelling is not a compliance action. It is a risk action.

Zhihu: public rules explicitly mention homophones, pinyin, numbers, and symbols

Zhihu is one of the clearest examples.

According to the public Zhihu Community Rules and organization-account rules, the platform bans “traffic diversion, marketing, and other promotional behavior.” It also separately states that QR codes, URLs, email addresses, and contact details disguised as homophones, pinyin, numbers, symbols, or similar variants are prohibited. We covered the broader context in our earlier post on Zhihu’s anti-promotion boundary.

That wording matters for two reasons:

1. Zhihu is not only checking whether a literal WeChat ID appears. It is checking whether the post is trying to transmit contact information.
2. Zhihu treats obfuscated variants themselves as an enforcement target. So “V me,” split spelling, or pinyin-number patterns are not safer than the standard form.

In practice, the riskiest placements include:

• titles and opening paragraphs;
• avatars, cover images, display names, and org-account profile fields;
• text embedded in images;
• comments and “supplementary” notes;
• phrases like “DM me for the resources” or “reply with a keyword to get the template.”

If your article is supposed to be knowledge content, the safe Zhihu move is not to disguise the CTA better. It is to remove the contact CTA altogether.

Douyin: public rules explicitly ban homophones and similar character tricks

Douyin is equally direct.

Under the public Douyin Community Self-Discipline Convention, account information cannot contain QR codes, URLs, email addresses, or contact details. The rules also explicitly mention homophones and similar text-number-symbol variants as part of the prohibited pattern, with examples such as “V me” and “+WeChat”-style hints.

That means the following should not be treated as clever workarounds:

1. putting “V me” in a nickname, profile, or bio;
2. hinting at contact through initials or near-sound text;
3. moving the contact cue into subtitles, covers, or screenshots;
4. avoiding a link in the post body but continuing the diversion in comments.

On Douyin, the risk is not limited to text either. QR codes, third-party watermarks, and external-conversion signals are all part of the same broader enforcement logic.

Kuaishou: public rules do not just ban contact info, they also name common variant phrases

Kuaishou’s public rules are unusually concrete.

According to Kuaishou Community Rules and its User Profile Rules, the platform bans off-platform transaction diversion and also bans steering users toward other platforms for content consumption. More importantly, the profile rules explicitly say that account fields cannot contain QR codes, third-party platform signs, watermarks, phone numbers, WeChat IDs, or variant phrases such as “V me,” “my profile has it,” and similar coded prompts.

That matters because it shows the platform is not merely matching one type of string. It is identifying language patterns that imply off-platform contact.

So even if a creator never writes a literal ID, the phrasing itself can still function as a contact signal and be treated that way.

Baijiahao: public examples already include “威信” and “V信” as malicious marketing

Baijiahao is slightly different in wording, but it reaches the same conclusion.

Public handbooks and rule summaries for Baijiahao define advertising and promotional content very broadly: media-account handles, QR codes, promotional URLs, contact information, and traffic diversion to other platforms all fall into risky territory. More importantly, public examples already classify variants such as “威信” and “V信” as part of malicious marketing scenarios.

That tells us two things:

1. Baijiahao does not accept the defense that “I did not write the standard form of the contact detail.”
2. Even if the rule text does not list every possible homophone, the enforcement examples already cover the idea.

That is why trying to outsmart Baijiahao by rotating through coded contact phrases is a bad long-term strategy. We saw the broader pattern already in our post about Baijiahao’s broad ad-detection rules.

Which “not-too-obvious” tricks are still risky?

If you reverse-engineer the logic of these public rules, the following patterns are all bad bets:

1. homophone abbreviations, such as “V me” or near-sound substitutions;
2. pinyin fragments, initials, or mixed letter-number forms;
3. symbol splitting, where characters are broken up with punctuation, spaces, or emoji;
4. suggestive phrases, such as “my profile has it,” “check the comments,” or “message me for the full pack”;
5. image-based hiding, where the contact clue sits in a watermark or footer;
6. comment-section diversion, where the body stays “clean” but the actual CTA happens later.

These are risky not because the platform is overreaching, but because they are exactly the kind of evasive behavior the rules are designed to catch.

Why do people still think these tricks work?

Because moderation is probabilistic, not always immediate.

That creates a dangerous illusion:

• someone else still gets away with “V me,” so people assume it is allowed;
• one of your own posts survives once, so you assume the pattern is safe;
• one platform fails to act this week, so you treat the silence as approval.

For a long-term content operation, though, the real risks are usually broader than a single takedown:

1. lower distribution and weaker recommendation;
2. worse account trust over time;
3. a stronger “marketing account” classification;
4. stricter future review.

That is also why we argued in our overview of external-link policies across 10 Chinese platforms that the better question is not “Can I get away with this once?” but “Is this worth the account-level risk?”

The safer alternative: stop disguising contact info and rewrite the CTA instead

If your goal is repeatable distribution rather than short-term lead capture, this is the safer playbook:

1. Remove all direct and indirect contact details. That includes body text, images, profile fields, and comment templates.
2. Replace “DM me” or “reply for the template” with a knowledge-first ending. Summaries, FAQs, and official references are usually safer.
3. Use weaker CTAs where needed. On stricter platforms, “search the brand name” is often lower-risk than links or contact info.
4. Rewrite by platform. Keep source links on technical communities, remove promotional diversion on stricter platforms.
5. Let your website and technical communities carry the heavier conversion role. Let stricter platforms focus on awareness and knowledge framing.

That is where OmniGoAI’s OmniPost is actually useful. Its value is not helping you disguise contact info. Its value is helping you manage one canonical site version plus multiple platform-safe rewrites without mixing the wrong CTA into the wrong destination.

A practical pre-publish checklist

Before you distribute a post across multiple platforms, check:

1. Does it contain phone numbers, WeChat IDs, email addresses, or QR codes?
2. Does it use variants like “V me,” initials, or profile-based hints?
3. Is any contact cue hidden in a cover image, screenshot, or watermark?
4. Does the CTA ask users to DM, comment for a keyword, or unlock a resource?
5. Are you planning to move the real diversion into comments later?
6. On stricter platforms, has the CTA been rewritten into a no-link, no-contact form?
7. Are source links kept only where the platform context can actually support them?

If any of these checks fail, do not assume a different spelling will save the post.

Frequently asked questions

FAQ 1: If I write “V me” instead of a full WeChat ID, is it still risky?

Yes. On platforms such as Zhihu, Douyin, and Kuaishou, public rules already target homophones, symbol tricks, and similar contact variants. Baijiahao also shows public enforcement examples in that direction.

FAQ 2: Is hiding the contact info inside an image safer?

No. Many platforms treat image text, covers, avatars, and watermarks as part of the same moderation target.

FAQ 3: Is it safer to leave the contact cue in comments instead of the article body?

Usually not. A comment-based CTA is still a CTA, and platforms do not necessarily stop caring just because the placement changed.

FAQ 4: Does this mean I cannot mention my product name at all?

No. Mentioning a product in a knowledge context is not the same thing as leaving contact details or building a lead funnel.

FAQ 5: What is the safest default?

Do not look for the last surviving coded phrase. Remove all contact details and their variants, then rewrite the CTA and linking strategy for each platform.

If you are publishing one article to your site, Zhihu, CSDN, Juejin, CNBlogs, and other destinations, the real asset is not a dictionary of disguised contact words. It is a workflow with one canonical version and multiple platform-safe rewrites. If you want that workflow in practice, start from the OmniPost download page.