Multi-ESP Email Deliverability: What Breaks When You Send From More Than One Provider

Most companies don’t choose to run multiple ESPs — it accumulates. Transactional email stays on whatever infrastructure the dev team set up first. Marketing moves to a platform built for campaigns. Cold outbound, if it happens, lands on a third tool entirely. Each is a reasonable decision in isolation. The problems show up only where they overlap — in your DNS.

Run a free audit → — paste your email headers and see every failing check, with the fix for each. No account.

The Three Places Multi-ESP Setups Actually Break

1. SPF’s 10-lookup limit. Every ESP typically adds one include: mechanism to your SPF record — sometimes more, if that ESP’s own SPF record nests further includes. RFC 7208 caps the total resolvable lookups at 10; cross it and the entire record becomes a PermError, not a partial failure. Three ESPs is usually fine. Five or six, especially if any of them nest, is where teams cross the line without noticing — the record still looks valid in a DNS lookup, it just fails at evaluation time. More on PermError specifically.

2. DKIM and DMARC alignment, per ESP. Each ESP typically signs its own DKIM key under either your domain or a subdomain it manages. If two ESPs sign under different domains and your DMARC alignment is set to strict, only the ESP whose signing domain exactly matches your From domain passes — the other silently fails DMARC alignment despite a valid DKIM signature. Full breakdown of alignment failures.

3. Split domain health, with no single view. Google Postmaster Tools and Microsoft SNDS report by sending domain and IP range — not by ESP name. If all your ESPs send from the same root domain, a drop in Gmail’s score for the domain shows up as one combined line with no indication of which ESP caused it. If they send from different subdomains, you get a separate score per ESP, but now you’re checking three separate views to get the full picture, and none of them cross-reference your own campaign or delivery data to explain why the drop happened.

The Audit: One Domain, Multiple ESPs

Step 1 — List every ESP actually sending mail today, not just the ones you remember configuring. Include transactional, marketing, cold outbound, and any tool with an “invite a teammate” or “send a receipt” feature (these often send through a bundled ESP without it being obvious). A quick way to catch ones you missed: check your SPF record’s existing include: list against your actual vendor list.

Step 2 — Count SPF lookups with everything included. The SPF Record Generator shows the running count live as you add each service — if you’re already near 10 with your current vendor list, adding one more ESP without consolidating isn’t safe.

Step 3 — Check each ESP’s DKIM signing domain against your DMARC alignment mode. If any ESP signs under its own domain rather than yours (some do this by default unless you set up “branded” or “custom domain” DKIM), and your alignment is strict, that ESP’s mail may be passing DKIM technically while failing DMARC alignment. The DMARC Record Generator builds a record with alignment set explicitly.

Step 4 — Decide on a subdomain strategy if you need each ESP scored on its own. Splitting transactional and marketing onto separate subdomains (transactional.yourdomain.com, mail.yourdomain.com) means a marketing campaign’s complaint rate can’t drag password-reset emails into spam. Each subdomain needs its own valid SPF and DKIM/DMARC setup — this isn’t automatic once you split.

Step 5 — Confirm the full picture with a live DNS check, since a change that looks correct in an ESP’s dashboard isn’t confirmed until it’s actually resolvable. The DNS Checker queries live via DNS-over-HTTPS, showing the current published state across every mechanism at once.

Frequently Asked Questions

Why do companies end up sending from multiple ESPs in the first place?

Usually not by design — it accumulates. Transactional email (receipts, password resets) often stays on whatever the dev team set up early (SES, Postmark, SendGrid API). Marketing email moves to a platform built for campaigns and automation (Klaviyo, Mailchimp, HubSpot). Cold outbound, if the company does it, typically runs through yet another tool (Instantly, Smartlead) built to keep cold mail out of spam. Each choice is reasonable on its own; the accumulation is what causes problems.

Does using multiple ESPs always break something?

No — plenty of multi-ESP setups run cleanly for years. The risk is specific and checkable: SPF lookup count (each ESP typically adds one include, and RFC 7208 caps the total at 10), and whether each ESP’s DKIM/DMARC alignment is configured correctly for the domain or subdomain it’s actually sending from. Both are a five-minute check, not a guarantee of failure.

Should each ESP send from its own subdomain?

Often yes, and it’s the cleanest fix for split domain health specifically — mail.yourdomain.com for marketing, transactional.yourdomain.com for transactional, giving each its own score with Gmail so a bad campaign send can’t drag password-reset emails into spam alongside it. It doesn’t fix the SPF lookup count on its own, since a subdomain still needs its own valid SPF record; it does mean each one only needs to include the ESPs actually sending from it.

How do I see domain health across all my ESPs in one place?

Google Postmaster Tools and Microsoft SNDS report by sending domain and IP range, not by ESP name — so if every ESP sends from the same root domain, their combined effect shows up together, but you can’t isolate which ESP is driving a drop without checking each ESP’s own delivery dashboard separately. This is the specific problem tools that pull Postmaster and SNDS into one view solve: correlating those signals against which ESP was sending at the time.