Start a new channel or fix the one you have

What actually transfers when you start over

Be concrete about what you keep and what you lose, because the fantasy of a fresh start is usually built on getting this backwards.

Lost: every follower and subscriber, the entire back catalogue, and — this is the underrated one — your ability to compare. Diagnosing a channel means comparing your best posts against your median. A new account starts at n=0, which means for the first two months you cannot tell a bad idea from a bad week.

Kept: your topics, your taste, your editing habits, your hook instincts, your posting rhythm and your assumptions about who you are talking to. If the honest diagnosis is "my content is not landing", every single one of those comes with you to the new account, and the new account is worse because it also has nobody watching.

That asymmetry is the whole argument. A restart moves the things that were not the problem and keeps the things that were.

The one case where starting fresh genuinely helps

There is a real mechanic behind it, and it is worth stating precisely rather than as folklore.

When you post, the people most likely to be shown it first are the people who already follow you. Their response is the earliest and cheapest signal available about whether the post is worth showing to anyone else. If you built an audience around one thing and you now make a different thing, your own followers become the wrong test audience: they scroll past, and the post looks weak in exactly the moment it most needs to look strong.

That is the situation a new account actually fixes. Not "bad reach" in general — a mismatch between the audience you accumulated and the content you now want to make.

Two related cases are also legitimate:

• Two genuinely different audiences. You want to make cooking videos and finance videos, and no single person wants both. Two accounts is the honest structure, not a growth hack.
• A different language or region. Same reasoning — the first-viewer pool has to plausibly want the thing.

Everything else on the usual list — "the algorithm has decided I'm not worth showing", "the account is burned", "I need a clean slate" — is a description of how it feels, not a mechanic either platform publishes.

"My account did well a year ago and now it's dead"

Almost always this is a change in what performs, not a change in your standing. Recommendation systems weight recent performance, so an account that stops producing posts people finish is treated as a new unknown quantity rather than an established one. That is recoverable by construction, because the input it reads is your next post.

Before assuming decline, check the two things that produce identical symptoms and have completely different fixes.

Check the published state first. Instagram shows account standing and any content restrictions in the app, and YouTube shows restrictions and strikes in Studio. If something is flagged there, no amount of content strategy is the answer. The full walkthrough of both, including what each platform will and will not tell you, is on am I shadowbanned or just unlucky.

Then check whether it is a drop or a plateau, because they diagnose differently. A drop has a date; a plateau has a level. The ladders are Instagram Reels views dropped and YouTube views dropped: what to check first.

One thing worth clearing up because the search results for this question are misleading: the "reset the algorithm" control that Instagram offers is a viewer-side setting. It clears the recommendations *you* are shown in your own feed. It does nothing to how your posts are distributed to other people, and no equivalent creator-side reset exists on either platform.

"Can an account recover after years of bad reach, or is it dead?"

Neither platform publishes a permanent account-level penalty, and there is no mechanism either company describes by which an old bad stretch keeps suppressing new posts indefinitely. What is published, repeatedly, is that recommendations are driven by how recent posts perform with the people they are shown to.

The practical translation is unromantic but freeing: your last ten posts matter far more than your worst year. Recovery does not require undoing history. It requires one post that performs well enough with a cold audience to widen the next one.

That is also why the recovery *feels* impossible from inside — you are judging by follower count, which moves slowly, while the thing that actually needs to move is the performance of a single upload.

Should I delete the old posts?

No, and specifically because you need them.

Neither platform publishes a mechanic by which a low-view post from last year drags down a new one. What deleting definitely does is destroy the comparison set — the only evidence you own about what your audience responds to — and remove whatever residual reach those posts still bring in.

If the old posts embarrass you or misrepresent the account, archive or hide them rather than deleting. You keep the data, the profile reads the way you want, and you have not paid anything for it.

The decision, as a table

If you are fixing, this is the order

1. Establish what your account is for, in one sentence a stranger would understand. Not a niche label — a reason to come back. It can be a topic, a format or a voice, but it has to be one of them and it has to be legible from three posts in a row.

2. Read your own catalogue before making anything new. Your best-performing posts are the most specific evidence you own, and most creators never study them systematically. The method is on what to post next.

3. Change one variable at a time and give it a run of posts, not one post. A single upload cannot tell you anything — variance between posts on the same channel is enormous, which is unpacked in why one video blows up and the next one dies.

4. Expect a slow patch while the audience re-sorts. If you change direction on an existing account, the followers who came for the old thing will not engage with the new thing, and reach dips before it recovers. That dip is the cost of the change, not evidence the change was wrong. It is also cheaper than starting from zero.

Where a tool helps, and where it does not

Disclosed interest: we make a channel diagnostic, so read the following as an interested party being specific rather than as neutral advice.

The part of this that is genuinely hard alone is step 2 — you cannot see your own pattern from inside your own catalogue, because you remember the intent of each post rather than its result. Channel X-Ray reads your recent uploads and names the single bottleneck, and the free Instagram Reel analyzer and free YouTube channel audit give a lighter version of the same read with no signup.

What no tool can decide for you is the restart question itself, because it turns on something none of them can see: whether the audience you have is the audience you want. That one is yours.