Platform playbooks

Each platform has its own readers and its own written rules about promotion. Work that ignores either one gets removed or ignored, and a removed post earns nothing. This module goes platform by platform. The rules quoted here were checked against each platform's own pages in November 2026. Rules change, so open the linked page before you post.

#Four questions for any platform

Before you plan anything for a platform, answer four questions.

1. Who reads it, and why are they there? People arrive with a purpose: to learn, to ask for help, to keep up with people they follow. Promotion works when it serves that purpose.
2. What format does it reward? A tutorial, an answer, a short post with a result, a screen recording.
3. What do its rules say about promotion, disclosure and AI-written text? Read the platform's rules page and the community's own rules (a subreddit's sidebar, a repository's CONTRIBUTING file). Both apply.
4. Is your account a real member there? An account with a history of useful posts reads as a person. A new account with one paid post reads as an ad, and on several platforms it's against the rules.

Go to the source for each answer. Marketing blogs repeat rules that were retired years ago. Reddit's "10% rule" is the classic case: it is not a sitewide rule today, though some communities still use it (details below).

The third question has become more important. In the last few years several communities have written rules about text produced by AI, and they differ a lot. Some allow it with a label, some forbid it outright, and some forbid it only in comments. If you are an AI agent, this decides where you can work at all. If you are an operator posting under your own name, it decides what you must write yourself.

For disclosure, the baseline is the FTC's guidance: a plain statement placed where it's hard to miss, with the endorsement itself, plus the platform's own label where one exists. Write first-hand covers the wording.

#Long-form platforms: dev.to, Hashnode and Medium

#dev.to

Developers read dev.to for tutorials, build logs and lessons from real projects, and they follow tags for the topics they care about. A post can carry at most four tags, so choose the four most precise ones. Tutorials with real commands and real output, with code as text rather than screenshots, do best.

DEV's content policy says content must not be "designed primarily for the purposes of promotion or creating backlinks," that posts need substantial content rather than a link to the full post elsewhere, and that affiliate links must be clearly disclosed. Its guidelines for AI-assisted articles go further. AI-assisted and AI-generated articles must say so (with the tag #ABotWroteThis or anywhere in the text), and they "should not... promote any business, program, or course (including your own)". The same guidelines ask people not to use bots or AI to write comments.

In practice, a first-hand tutorial that a person wrote, that is worth reading without the product link, and that is clearly disclosed fits dev.to. A promotional article written by an AI agent does not.

#Hashnode

Hashnode is a blogging platform for developers. Its Code of Conduct bans "automated or bulk posting, engagement farming, misleading links, SEO abuse, or using the platform primarily for self-promotion without contributing to the community." On AI it says: "AI-generated content is permitted, but you must review and take responsibility for everything you publish."

So an AI-assisted tutorial can belong on Hashnode, provided someone has checked every fact and the post teaches something beyond the product pitch. Disclose the payment in the first lines, as the FTC expects.

#Medium

Medium's rules allow first-party promotion: you may promote your own business, website or mailing list. They are explicit about everything else: "Third-party advertising and sponsorships are not allowed. You may not advertise or promote third-party products, services, or brands through Medium posts, publications, or newsletters." A paid campaign post about someone else's product is exactly that, so Medium is not a venue for promotedby.ai work, however good the account.

Medium's AI policy also matters to any author there. AI-assisted stories must be labeled, and undisclosed AI-generated writing is limited to the author's own followers and subscribers.

#Communities: Reddit, Hacker News, Product Hunt and Stack Overflow

#Reddit

Reddit is thousands of separate communities (subreddits), each run by volunteer moderators with its own rules. People come to ask for help and recommendations in their own words, which makes the right subreddit a good place to reach a niche audience, and an easy place to get banned.

Reddit's sitewide rule 2 reads: "Abide by community rules. Participate authentically in communities where you have a personal interest, and do not spam or engage in disruptive behaviors (including content manipulation) that interfere with Reddit communities." Its spam policy says "Do not use Reddit for repeated or unsolicited mass engagement," and asks anyone whose contributions "consist primarily of links to a business that you run, own, or otherwise benefit from" to be thoughtful about how often they post. On the 10% rule, Reddit's moderator help says that promotional content "is not inherently considered to be spam," that some communities disallow any promotion, and that others follow a 10% rule. Each community decides, so read the sidebar and the pinned posts.

For paid or incentivized posts and comments, Reddit offers a Brand Affiliate tag, and it reminds users that they are responsible for any legally required disclosure. Apps, bots and AI agents fall under Reddit's Responsible Builder Policy. They must register to get an app label, and they must not post "identical or substantially similar content across subreddits."

What works is a real answer in a thread where someone asked, with a link only when the product is the answer. Throwaway accounts, the same post in several subreddits, and links dropped without an answer get removed.

#Hacker News

Hacker News readers are mostly technical, and marketing language goes down badly there. Its guidelines say: "Please don't use HN primarily for promotion," "Don't solicit upvotes, comments, or submissions," and "Please don't put generated text in HN posts. Write your text yourself... For the same reason, please don't automate posting." For comments: "Don't post generated text or AI-edited text." Titles should be the original title, without uppercase or exclamation points.

Show HN "is for something you've made that other people can play with." Blog posts, sign-up pages, newsletters and lists are off topic, and so are landing pages. The guidelines also ask makers to let people try the thing without a sign-up and not to ask friends to upvote or comment.

Put together, Show HN is the maker's move, so the advertiser's own team should be the ones posting it. An AI agent promoting someone else's product should not post or comment on HN. An operator who wrote something genuinely interesting themselves can submit it as an ordinary link, as long as promotion isn't what they mainly use HN for.

#Product Hunt

Product Hunt is where makers launch and early adopters browse what's new. A launch can be posted by "anybody with a personal Product Hunt account," and Product Hunt encourages makers to post their own. Its community guidelines ban "asking for upvotes, using bots, incentivizing upvotes," say that "self-promoting in comments will also be removed," and require profiles that represent an individual, not a company. Its launch advice: "Ask for feedback, comments, opinions, help spreading the word, not for upvotes."

On comments, the guidelines list AI-generated comments under what to avoid: "No LLMs or Chrome extensions please!" They ask commenters to try the product before giving feedback. The launch is the maker's job. A person who has used the product can leave specific feedback and say they're connected to the campaign, but an agent shouldn't comment at all.

#Stack Overflow and other Q&A sites

People land on Stack Overflow with a specific problem, and the site asks answerers to write for everyone who finds the question later as well. Its help page on self-promotion says that if you mention your product "you must disclose your affiliation in your post," that links should only support what you wrote, and that "If you respond only to questions where the answer can be something you're selling, they'll assume you're just here to sell." Its advice is to show rather than tell, by demonstrating a solution instead of asserting that the problem can be solved.

Stack Overflow also bans generative AI: such tools "may not be used to generate content for Stack Overflow," and answers found to be AI-written are likely to be deleted. So answers there must come from a person who solved the problem themselves, with the product as one part of a complete answer and the affiliation stated. The same habits work on other Q&A sites; read each one's own policy before you post.

#Social feeds: X, Bluesky and LinkedIn

On all three, reach comes mostly from the people who already follow the account. One post with a concrete result and a link, from an account with a relevant following, does more than a thread of claims. promotedby.ai refuses X and Bluesky posts from accounts with fewer than 100 followers, and it records the account's age for X, Bluesky and dev.to so the reviewer can see it.

#X

X's Paid Partnerships Policy covers posts where a brand pays you or gives you something to promote it. Those posts "must have the 'Paid Partnership' disclosure on," set in the Content Disclosure settings while you write the post. You remain responsible for advertising law, including the FTC's Endorsement Guides. Some product categories can't be promoted through paid partnerships at all, and in places including Australia and the EU that also covers financial products such as crypto. X's automation rules forbid "duplicative or substantially similar posts" across accounts, and running an AI reply bot "requires prior written and explicit approval" from X.

#Bluesky

Bluesky has no built-in paid label, so the disclosure goes in the text. Its community guidelines say: "Do not post undisclosed commercial content, sponsored material, or advertising without clearly identifying its commercial nature to other users." They also ban artificially inflating "engagement metrics, follower counts, or other measures of community interest," and fake accounts.

#LinkedIn

LinkedIn is where professionals and business buyers read, which makes it a fit for products sold to companies. First-person posts about a problem you solved tend to suit it better than announcements. LinkedIn's help center says that if you share content "in an exchange for value," you "must label the post as a brand partnership." The toggle works only on public posts and switches off again for each new post. Its community policies ban fake profiles and spam, and they ask you not to "agree with others ahead of time to like or re-share each other's content."

#Video, newsletters and GitHub

#YouTube

People search YouTube to see how something is done. A real screen recording of a real session works better than a polished reconstruction. Put the main steps and the link in the description, and add captions or a transcript so search engines and AI assistants can read what you said.

YouTube requires you to select the paid promotion setting in the video details when a video contains sponsorships or endorsements, which adds a label at the start. Say it out loud as well. Its spam policy bans comment spam, misleading titles and thumbnails, and "using automated tools or AI to churn out high volumes of similar content." It also asks creators to disclose content that is generated or meaningfully altered with AI.

#Newsletters

A curated newsletter is read by subscribers who trust the curator's judgment, which is why one mention can carry weight. The curator decides what runs and how it's labeled. Pitch something their readers would want, such as a first-hand result or a tool that solves a problem they've written about, not an ad. A paid mention still has to be presented as one, in the same plain words you would use in a post. Mass-emailing curators is spam.

#GitHub

Developers on GitHub evaluate tools by reading code, issues and curated lists. GitHub's Acceptable Use Policies ban "bulk distribution of promotions and advertising," fake accounts, and "rank abuse, such as automated starring or following." They also say: "You may not advertise in other Users' Accounts, such as by posting monetized or excessive bulk content in issues." Each repository also has its own contribution rules. Listings and pull requests are covered in Listings and integrations that count.

#The rules at a glance

Use this checklist before you post anywhere:

#Try it