you grow a meme account by posting things people forward to someone else, and by doing it on a schedule you can hold for a year. that is the whole method. everything written about how to grow a meme account faster than that is describing volume, and volume is the specific behaviour every platform has spent the last two years building filters against.
the reason is worth understanding before you post anything. a recommendation system is trying to work out whether a piece of content is worth showing to strangers. it has two signals available: what people do when they see it, and what the account has done before. an account that posts forty reposts a day is answering the second question for the system, and the answer is no.
the numbers platforms pay on
it helps to know the actual bar. youtube's partner programme requires 1,000 subscribers with 4,000 qualified public watch hours in the last 12 months, or 1,000 subscribers with 10 million qualified shorts views in the last 90 days. those are the two doors, and there is a detail inside them that catches meme accounts specifically: qualified watch hours from shorts views in the shorts feed do not count towards the 4,000 hours.
read that twice. a shorts-only account cannot back into the watch-hours door by accumulating short views. it has to clear ten million of them in ninety days, which is a little over a hundred thousand a day, sustained. that is a real number and it tells you what kind of account the format supports. it is achievable with a format that people rewatch and share. it is not achievable by uploading more.
other platforms run the same logic with different arithmetic, and none of them count an upload as an achievement. what gets counted is the second view, the one the algorithm gave you because the first person did not scroll past.
the reposting problem
most meme accounts start as aggregators, and aggregation is exactly the behaviour policy language now describes. google's spam policies define scaled content abuse as many pages generated for the primary purpose of manipulating search rankings, and the same page counts scraping feeds, search results or other content to generate many pages, including automated transformations such as synonymizing and translating. change pages to posts and you have a description of a repost account with a caption generator bolted on.
the enforcement pattern is consistent across platforms and it is not a ban. when spotify tightened its rules on 25 september 2025 it said it had removed over 75 million spammy tracks in the past 12 months, and described a spam filter that identifies uploaders using mass uploads, duplicates and seo hacks and then stops recommending them. stops recommending. the account stays up, the numbers just flatten, and the person running it concludes the algorithm changed.
there is a second cost to reposting that has nothing to do with policy. an account made entirely of other people's work gives nobody a reason to follow it over the accounts it takes from. the follow is a bet that the next post will be good, and nothing about a repost feed supports that bet.
one format, repeated
the accounts that grow have a shape you can describe in a sentence. same crop, same caption position, same joke structure, same voice. a follower who sees the third one recognises it before they read it, and recognition is what converts a viewer into a follower.
that constraint is also what makes daily posting survivable. inventing a new idea every day is impossible; filling a format you already built takes twenty minutes. pick the format by making thirty posts and looking at which five travelled, then throw the other twenty-five away and do the five again.
the material has to come from somewhere real. macquarie dictionary named ai slop its word of the year for 2025 on 24 november 2025, defining it as low-quality content created by generative ai, often containing errors, and not requested by the user. audiences now have a word for filler and they use it fast. a model can help you cut a caption down. it cannot supply the observation, and an account with no observations in it reads as slop within a week.
rhythm over volume
a workable schedule for a starting account looks like this:
- one post a day, at the same time, seven days a week
- one longer or stranger post a week that would not work daily
- one week in eight with nothing new, reposting your own best work
the last one is the part people skip. your own archive is the only feed you are allowed to mine freely, and a post that worked six months ago will work again for the people who joined since.
why posts spread covers the mechanics under this, and the anatomy of a viral reddit post is a useful case study in how much of a post's performance is decided by its first line.
a stalled account
flat numbers point to one of three situations, and each needs a different response. if reach dropped suddenly across every post, something changed on the platform side and the answer is to keep posting the same thing for three weeks before concluding anything. if reach declined slowly over months, the format has worn out and needs a variation. if reach was never there, the format was never distinct enough to be recognised.
the diagnostic is share rate. a like costs nothing and means nothing. a share means someone put their own name next to your post in front of people they know, which is the behaviour every recommendation system is proxying for.
give a format six months before you judge it. an account that posts daily for six months has made about 180 attempts, which is enough to know whether the thing works, and almost nobody gets there because the first eight weeks are silent by design.