Substack growth strategy 2026: 4 channels, hours and payback

Most Substack growth advice is a list of tactics with no arithmetic attached. Post more Notes. Be consistent. Collaborate. All true, all unfalsifiable, none of it tells you whether the thing you did last week was worth doing.

This guide is the other version: the four channels that actually move a Substack in 2026, what each one costs in hours, the number that tells you it is working, and the order to do them in. Where we have measured something ourselves we say so. Where we have not, we give you the ratio to measure instead of a number to believe.

What changed going into 2026

Three shifts matter for how you spend your week.

• Notes is the discovery surface, not the newsletter. Recommendations still bring subscribers, but the feed is where a publication with no audience gets its first hundred readers. A publication that only ships posts is invisible by default.
• Reciprocal growth stopped paying. Subscribe-for-subscribe rings, engagement pods and follow-back threads still inflate counts, and Substack's own surfacing leans on engagement rates that those subscribers destroy. The result is a bigger list that reaches fewer people — see why we stopped.
• Search became viable for publications, not just posts. Custom domains, real pages and original guides now bring a slow trickle that compounds — and unlike the feed, it does not stop when you stop posting.

The four channels, honestly compared

Read the second column and the fourth together. Replies pay fastest and compound least on their own; search pays slowest and is the only one still working while you are on holiday. A strategy is a choice about that trade-off, not a to-do list.

Channel 1 — Notes, as a system

One to three original Notes a day beats ten Notes on Monday and silence until Friday. That is the whole cadence rule, and it is boring for a reason: the feed rewards presence, and presence is a frequency, not a volume.

What to actually publish, in rough proportions over a week:

• ~50 % original thinking — an argument, a number, a failure, a question. These build the reason to follow you.
• ~30 % archive-derived — one sentence of a published post, expanded into a standalone Note. Never a link with "new post!" attached; that is an announcement, and announcements do not travel.
• ~20 % conversational — the reply-shaped Notes that exist to start threads.

The metric: median replies per Note, tracked weekly. Not impressions, which mostly track how often you posted. Going from a median of 0 to 3 is the single largest available improvement for a small account, and it is a writing change, not a volume change — how to write Notes that get replies is the mechanics.

Channel 2 — Replies, the underpriced channel

Reply space is uncrowded because it is unglamorous. Nobody screenshots their reply. But the arithmetic is favourable: a substantive reply under a Note with two thousand impressions puts your name in front of an audience you did not have to earn, and the author — who reads every reply — now knows you exist.

A workable routine, 30 minutes a day:

1. Find five to eight Notes from people whose audience overlaps yours and who are slightly bigger than you. Much bigger and you are one of two hundred replies; much smaller and there is no reach to borrow.
2. Reply with something that could not have been written without reading the Note: a counter-example, a number, a specific disagreement, an extension of the point.
3. Never pitch. Never link. The profile is the click.
4. Come back to the ones that got answered — the second exchange is where the relationship actually forms.

The metric: profile visits per reply session. If you write eight replies and get no profile visits, your replies read as agreement, not as thinking.

Channel 3 — Posts, for retention and money

Posts are what people subscribed for, and the newsletter is where paid conversion happens — but posts are a poor discovery channel on their own. Treat them as the thing that makes the audience worth having rather than the thing that builds it.

• Cadence beats length. A predictable weekly post outperforms an unpredictable brilliant one, because retention is a habit.
• Every post is at least three Notes. Not "here is my new post" — three arguments from it, each standing alone. The Note wins the reader; the post keeps them.
• Fill the SEO fields. Substack defaults to generic titles and descriptions if you leave them empty, which costs clicks on every channel that shows a preview.

The metrics: open rate (the honest measure of whether the list is real — this is what reciprocal subscribers wreck) and posts → Notes conversion, i.e. whether your archive is actually feeding the feed.

Channel 4 — Search, the one that keeps working

The slowest channel and the only one that is an asset. Two things make it work:

• Own a domain. A custom domain makes your publication a site rather than a page on someone else's, and it lets you build pages that are not posts.
• Publish original pages, not mirrors. Guides, comparisons, templates — content answering a question people type into a search box, hosted on your own domain, canonical to you. A mirror of your Substack archive is a duplicate; a guide is an asset.

Expect nothing for the first six to twelve weeks. That is not failure, that is the shape of the channel: crawl, then index, then impressions, then — last — clicks. The metric that tells you it is working long before traffic does is non-brand impressions: searches that were not your own name or publication title. Zero non-brand impressions after two months means your pages are indexed but not competitive for anything. First non-brand impressions means the machine has started.

The 90-day plan

Weeks 1–4 — presence. One Note a day, 30 minutes of replies a day, weekly post. Nothing else. The only goal is that the feed learns you exist and you learn which Note shapes your readers answer. Checkpoint: median replies per Note above zero; five to ten people you now recognise by name.

Weeks 5–8 — leverage. Keep the cadence. Add archive-derived Notes so posts start feeding the feed. Start the second exchange with the people from weeks 1–4 — restack them, quote them, answer their questions properly. Checkpoint: restacks from people you did not ask; subscribers arriving from Notes rather than from recommendations alone.

Weeks 9–12 — assets. Same cadence, plus two owned pages: the guide only you could write, and the comparison your niche has no honest version of. Submit the sitemap, request indexing, link each new page from two existing ones. Checkpoint: pages indexed; first non-brand impressions. Not clicks — impressions.

At day 90 you should be able to answer: which Note shape works for my readers, which five people reliably amplify me, and which query my site could plausibly own. If you cannot answer those three, more volume will not help.

Vanity metrics vs. the ones that predict

The left column moves when you are busy. The right column moves when you are effective. Most publications that feel stuck have a healthy left column — that is exactly why the situation is confusing.

One caveat that breaks all of this: if some of your engagement is reciprocal, your metrics are lying to you in the direction of optimism. A reply from a subscribe-for-subscribe thread and a reply from a stranger who found you look identical in the Substack interface. Separating them is the difference between knowing what works and repeating what happened. It is the specific job Subshack does — it reads your own Notes stats locally and splits the growth that was organic from the growth that was reciprocal, so the top-five-vs-bottom-five comparison actually means something.

Where to start this week

If you do one thing: pick the channel with the shortest time-to-signal that you are currently not doing. For most people that is replies — thirty minutes a day, no pitching, for two weeks. It is the cheapest evidence you will get about whether your ideas travel, and it costs nothing but the half-hour.

Whichever channel you pick, all four end at the same page: your profile. Before you send traffic to it, run the free Substack profile audit — a score out of 100 across positioning, conversion, trust, activity and discovery, with the fix list ordered by weight. It is the cheapest hour in this whole plan.

If you want the tooling map first, the best Chrome extensions for Substack writers covers what each category does and where drafting help ends and automation begins. And before you spend a week optimising for the feed, read the Substack Notes algorithm: what we know — most of the reachable levers turn out to be upstream of ranking.