Substack: can you even test deliverability? Yes, here is how.

Substack is an odd beast in the deliverability world. It has a million writers, tens of millions of subscribers, and almost no visible controls for either side. Writers cannot set DKIM keys, cannot rotate IPs, cannot warm their own domain, and often cannot even see per-send bounce detail. Readers cannot easily move a Substack out of Promotions other than marking it important in Gmail.

This frustrates writers used to Mailchimp, ConvertKit or Beehiiv. Where are the settings? Where is the reputation dashboard? The honest answer is: Substack manages it for you, and you do not get to peek. The good news is that you can still do the single most useful thing any newsletter operator can do — send a real issue to real mailboxes you own, and see where it lands.

What Substack will and will not let you do

Substack abstracts away most of the deliverability stack. What that means in practice:

• No custom DKIM. Your posts sign with substack.com.
• No custom sending IP. You share the Substack pool with every other publication.
• No per-send speed or send-time customisation beyond choosing a publish time.
• No custom tracking domain. Clicks resolve through Substack's own infrastructure.
• Basic stats: opens, clicks, unique subscribers. Folder placement is not reported.

Substack's position is reasonable — most writers do not want to touch DKIM, and a carefully managed shared pool usually outperforms an amateur-configured solo setup. The problem comes when your opens drop and you have no visibility into why.

Subscribing seed addresses to your own publication

The one lever Substack leaves you is your subscriber list. Anyone can subscribe to any Substack, including you. That means you can add a handful of seed addresses — mailboxes you own on Gmail, Outlook, Yahoo, Mail.ru, ProtonMail and so on — and those seeds will receive every post you send, at the same time, from the same infrastructure as your real subscribers.

1. Get seed addresses. If you maintain your own (old spare accounts) that works. Otherwise the free Inbox Check tool provides a fresh batch of 20+ on request.
2. For each seed, go to your Substack URL in a browser logged into that seed's email. Enter the seed address in the subscribe box. Confirm the double-opt-in email that lands in the seed mailbox.
3. Alternatively, use Substack's Settings → Subscribers → Add subscribers flow to import the list of seed addresses in one go. Note: some regions require double-opt-in regardless.
4. Publish as normal. Seeds receive the post alongside your real subscribers.

What Gmail actually does with Substack senders

Once seeds are confirmed and receiving, you finally have visibility. The interesting patterns show up in Gmail:

• Most Substacks land in Gmail's Promotions tab by default. That is normal, not a failure.
• Established publications with consistent engagement graduate to Gmail Primary for some readers. Whether this happens is partly up to Google — you cannot force it.
• New Substacks, and those that ship sudden volume surges, sometimes land in Spam for a few issues until the pool re-calibrates them.
• Outlook and Yahoo typically deliver to Inbox without tab filtering. Mail.ru is the strictest of the big providers for Substack senders.

my-substack.substack.com — placement over last 6 issues
                gmail      outlook  yahoo   mail.ru  proton
  issue 12      Promo      Inbox    Inbox   Inbox    Inbox
  issue 13      Promo      Inbox    Inbox   Spam     Inbox
  issue 14      Promo      Inbox    Inbox   Inbox    Inbox
  issue 15      Spam       Inbox    Inbox   Spam     Inbox  <- bad
  issue 16      Promo      Inbox    Inbox   Inbox    Inbox
  issue 17      Primary    Inbox    Inbox   Inbox    Inbox  <- graduated

The issue-15 Gmail-Spam / Mail.ru-Spam pair is the moment you want to catch. Something — subject line, link pattern, image ratio, a complaint from a subscriber — pushed that issue over a line. Without seeds you would only see a bad open-rate number and guess.

Implications: what you can actually change on Substack

Because Substack hides so many dials, your response to a bad seed result is mostly editorial rather than technical:

• Subject line. Spammy-adjacent words ("free", "limited time", excessive punctuation) hurt your Substack as much as any ESP. Seeds will show you the difference.
• Image-to-text ratio. A post that is a single big image on top of a short paragraph reads as promotional. Add text.
• Link count. Dozens of clickable links per issue tilt classifiers. If your links are all Substack-internal, it is less bad; if they are a bundle of outbound affiliate URLs, much worse.
• Publish cadence. Sudden triples after silence look like list-dumping. Your seeds will see you in Spam on the comeback issue.

Things seeds will not fix on Substack

Seeds measure placement. They do not give you back controls Substack does not expose. If your results show persistent Spam in Mail.ru and you run a Russian-language publication, you can change subject lines, cadence and content — you cannot change the sending domain from substack.com to your own. If that is a blocker, the answer is to move to a platform that lets you use a custom sending domain (Beehiiv, Ghost, Buttondown, ConvertKit) rather than to keep grinding on Substack.

Frequently asked questions