ESP matching (Instantly calls it "provider matching") routes each email through a sending mailbox on the same provider as the recipient: Google mailboxes send to Gmail and Google Workspace recipients, Microsoft mailboxes to Outlook and Microsoft 365. Buyers care because same-provider mail is widely believed to land better, and because it sells mixed Google and Microsoft infrastructure (Google vs Microsoft vs SMTP).
How it works
- Classify recipients. Look up each lead's domain MX records to identify the receiving provider (Google, Microsoft, Zoho, "other"). A company domain on Google Workspace counts as Google even if it does not end in gmail.com.
- Classify mailboxes in the campaign by provider.
- Allocate. When scheduling, prefer a mailbox on the recipient's provider that still has daily capacity. When those run out, fall back.
The fallback rule is where vendors differ. Smartlead's example: matched mailboxes are used first, and once Outlook leads finish, "those 8 Outlook accounts will be used to send emails to the remaining Gmail leads". Salesforge defers to other providers when matching mailboxes are full and sends unknown-provider leads from any mailbox.
Who does it well
| Tool | Providers | Default / gate (Oct 2026) |
|---|---|---|
| Smartlead | Gmail, Outlook, Zoho | Campaign setting; API docs describe "AI ESP Matching" that also weighs account deliverability scores and history (API docs); can also isolate sending to one provider (help) |
| Instantly | Google, Outlook, "Others" | Off by default; campaign or workspace-wide (help) |
| Salesforge | Gmail, Outlook, etc. | Growth plan, $80/mo (pricing) |
| PlusVibe | Not stated | "Basic" on Personal, "advanced" on Business (pricing) |
| Woodpecker | Not stated | Included (pricing) |
EmailBison's homepage mentions "ESP identity matching", but in the context of merging duplicates, not routing (emailbison.com). Its public docs cover only the API. Whether EmailBison routes by recipient provider is unverified, as is support at lemlist, Saleshandy, Reply.io and the suites.
Where implementations differ
- Fallback vs throughput. Instantly's article contradicts itself. It says the system "proceeds with Gmail without waiting" when there is no Outlook mailbox, and also that matching "may not reach your full daily limit" while it waits for matched accounts. Either way, buyers trade some volume for matching.
- Signals beyond provider. Smartlead's API documentation describes matching that also weighs account deliverability scores and sending history, which moves toward true routing. Instantly separately detects Secure Email Gateways (Barracuda, Mimecast, Proofpoint, Cisco), but only as a lead filter, not as a routing input.
- Granularity. Matching is by provider family only. No vendor documents routing for regional providers such as IONOS, Strato, GMX/WEB.DE, T-Online or OVH, which host a meaningful share of European SMB mail. That is an inference from the absence of documentation, not from a market-share study.
Impact numbers are unaudited. A third-party post claims "most teams see a 5-15% improvement in inbox placement", and Prospeo's blog cites Google Workspace→Gmail at 94–96% versus Microsoft 365→Gmail at 85–88%. Neither publishes methodology.
Risks and abuse
The mechanism is benign: choosing which of your own mailboxes sends a message. The second-order effect is what matters. ESP matching is the reason agencies buy both Google and Microsoft mailboxes in bulk, often through resold or "DFY" accounts that sit in a policy grey zone at both providers (Provider rules: Google, Microsoft and the ESPs, Infrastructure strategy: build or partner). It also hard-wires a dependency: if Google tightens rules on Workspace accounts used for cold email, the Google half of every matched fleet is exposed (Google and Yahoo sender rules, Microsoft rules: Outlook.com and Microsoft 365 limits).
Matching is a routing problem that vendors have solved only for the two largest providers. Routing that accounts for corporate gateways (SEGs), regional providers and each mailbox's live placement by destination, and that publishes its measured lift, is an open lane, especially for a DACH-first entrant.
What this means for an entrant
- Ship Google/Microsoft matching on by default, with a clear fallback policy the user chooses: "wait" or "send anyway". Instantly's ambiguity is a small trust gap you can close.
- Generalise it into destination-aware routing: recipient provider, SEG presence, and per-mailbox placement history by destination. Smartlead is moving this way, and nobody exposes it clearly.
- Add European providers (IONOS, Strato, GMX/WEB.DE, T-Online, OVH) as first-class destinations if you lead with DACH (Non-English markets, The wedge).
- Measure and publish uplift with holdouts. A randomised "matching on/off" split per campaign is cheap and would be the first credible number in the category.