How Substack Recommendations Actually Work

Published: September 2026

Recommendations are the single largest growth channel on Substack by platform-wide volume, yet most writers set them up once during onboarding and never think about them again. Understanding the actual mechanism - and its real limits - changes how deliberately worth managing they are.

The mechanism: a post-subscribe endorsement

When a reader subscribes to a publication, they're shown a screen suggesting other newsletters to follow - curated specifically by the writer they just subscribed to, not by a generic platform algorithm. This is the entire mechanism: it's an endorsement, not an advertisement. The underlying logic is that if a reader already trusts Writer A enough to subscribe to them, that trust transfers partially to whoever Writer A vouches for. One description frames it as "borrowed trust" - when a creator recommends another publication, "they are placing their reputation beside yours," and a new subscriber arriving via recommendation starts the relationship pre-qualified rather than at zero, because someone they already trust did the vetting for them (Grow Like Crazy).

The scale of the effect, platform-wide

Recommendations aren't a niche tactic - they're reportedly Substack's single biggest growth lever. Across the platform, recommendations drive 50% of all new subscriptions and 25% of new paid subscriptions. That's not one channel among many for the average publication; for a large share of writers, it's the largest single source of new subscribers, ahead of search, social, or word of mouth. It's part of what's described as having solved newsletter writers' historic "cold start problem" - a brand-new publication with genuinely good positioning can now get real distribution from writers who already have an audience, rather than needing to build one from nothing.

What this looks like at very different scales

The effect scales, but not predictably. At the high end, one prominent writer, Lenny Rachitsky, reported that 78% of his subscribers came from other Substack newsletters recommending his, at a scale of roughly 250,000 subscribers. At smaller scales, reported figures vary widely - some writers attribute 28% of growth to recommendations, others report the feature accounting for just over 10% of their subscriber base. The range itself is the useful data point: recommendations compound with scale, since being recommended by other well-recommended writers puts you in front of more potential recommenders in turn, but even a small or new publication can see a meaningful share of growth from the feature if positioned well.

Why alignment beats raw audience size

A specific and important nuance: a recommendation's value depends far more on audience overlap than on the recommender's size. As one breakdown puts it, "a recommendation from a publication with 2,000 highly aligned subscribers may be considerably more valuable than one from a publication with 50,000 people who have little interest in your subject." This mirrors the same audience-overlap logic that makes restacking effective - in both cases, the platform (and the reader) responds to actual topical relevance, not raw reach. Chasing a recommendation from the biggest newsletter in your orbit is less valuable than getting one from a smaller newsletter whose readers are genuinely likely to care about your specific topic.

Being "recommendable" is a positioning problem, not a luck problem

Writers can't force other people to recommend them, but they can make it far easier to be recommended in the first place - and the biggest lever here is clarity of positioning. A vague description like "Chris writes interesting things" gives another writer nothing concrete to say when introducing you to their audience. A specific value proposition - something closer to "Chris helps independent writers turn their newsletters into sustainable creator businesses" - gives a potential recommender an actual sentence they can use. If your niche is blurry, other writers genuinely don't know how to categorize you when deciding who to recommend you to, which quietly removes you from consideration regardless of how good the writing itself is.

Recommendations grow from relationships, not swaps

The healthiest version of this system runs on actual reading, commenting, and engagement between writers - not transactional "I'll recommend you if you recommend me" arrangements. A recommendation someone gives because they've genuinely read and value your work reads as more credible to their subscribers (and produces better-matched new readers) than one given as a favor. This is worth remembering alongside the broader point that Substack's growth mechanics reward genuine engagement over volume - recommendations are one more place where showing up authentically in other writers' communities pays off more than gaming the system.

The honest counterpoint: recommendation subscribers aren't automatically better subscribers

It's worth being clear-eyed here rather than treating recommendations as an unambiguous win. At least one analysis found the inverse of what you might expect at first: a newsletter with a smaller share of recommendation-driven subscribers performed significantly better on open rates, engagement, and paid conversion than one more heavily reliant on recommendations (How to Grow a Newsletter). The likely explanation is the same one that applies to any low-friction acquisition channel: a subscriber who tapped "yes" on a suggestion screen during onboarding hasn't necessarily made the same deliberate choice as someone who actively sought your publication out, and that difference in intent can show up later as lower engagement. The practical takeaway isn't to avoid recommendations - it's to watch your subscriber sources and engagement by source, the same way you'd track engagement for any other channel, rather than assuming every recommendation-driven subscriber is equally valuable.

The bottom line

Recommendations work through a genuine trust-transfer mechanism, and at the platform level they're the single largest driver of new subscriptions on Substack. The lever you actually control is positioning - being specific enough about what your publication does that other writers have an easy, credible reason to vouch for you - plus building the kind of real relationships with other writers that produce recommendations because they mean it, not because it's a swap. Just don't assume every recommendation-driven subscriber behaves like one who found you deliberately; the data suggests they often don't.

NotesIQ helps you see which of your growth channels - including recommendations - are actually converting into engaged readers, not just headline subscriber counts.