How to Use MailReach to Find Why Your Emails Land in Spam

The short answer: when a mailbox has a high reputation and your campaign still lands in spam, MailReach's own help centre puts the fault in the email or the way it is sent, and about six spam tests will find it; MailReach's smallest Spam Tester bundle is 20 tests for $15 a month (pricing page, checked on 26 October 2026), and the first test costs nothing. The steps: run the free test, take a reference test from your real setup, send a plain control email, put the original back together one element at a time until it lands, fix and retest, then let an agent run the tests while a person decides what changes.

Disclosure: I may earn a commission through some links on this site. It does not change what this page says.

MailReach's own help centre is plain about the limit of its warmer: you can have a perfect sender reputation and still land 100% in spam because of your content or your sending setup, says its troubleshooting article. This guide follows that article's method with the spam test. What the product includes and costs is in the MailReach glossary entry.

Count the tests before you buy any:

1. The budget. For example, the method below takes six tests: one reference, one control and four variations (no tracking, no links, new wording, no images). Six credits fit in any bundle on the pricing page: 20 pay-as-you-go tests cost $19.5, and the All-in-One plan includes 20 a month. A test that fails to collect results is not charged, per the API guide.
2. The email you already send. Keep the real campaign email, the real sending tool and its tracking settings exactly as they are. The troubleshooting article is firm that a test sent under other conditions gives another result.
3. The score that matches your list. Every test reports one score for professional inboxes (Google Workspace and Microsoft 365) and one for personal inboxes (Gmail.com, Outlook.com and Hotmail), according to the API guide. Read only the one your recipients use.

What you need

• The campaign email, the tool that sends it and access to its tracking settings.
• DNS access on the sending domain, in case the answer is a custom tracking domain.
• After step 1, a MailReach account with spam test credits; for step 6, an assistant that supports MCP.

Step 1: Run the free test

On the free spam test page, copy the code MailReach shows into the body of your email, send it to the addresses listed and ask for the score; no account needed. The report shows where the email landed across more than 30 mailboxes and checks SPF, DKIM, DMARC, links, HTML and spam words. If the email lands in the inbox at the providers you target, stop here: the content is fine, and the question is reputation.

Step 2: Take the reference test

Send the campaign email exactly as your recipients get it: same tool, same sending address, tracking on if it is normally on. If it lands mostly in spam on the inboxes you target, keep that result as the reference for every later test. If it lands mostly in the inbox, the troubleshooting article reads that as no spam problem, or as earlier sends that ran under different conditions.

Step 3: Send the control email

The troubleshooting article calls it the "100% innocent email". One or two plain professional sentences. A natural subject in small letters, such as "quick intro", with no punctuation. Plain text with no images or formatting, no signature, no links at all (the unsubscribe link included), no first name, and open and click tracking switched off in your sending tool. Switch off any browser extension that tracks opens as well.

Then check the Content Analysis tab of the result: it should detect zero links and no tracking pixel. If the control lands in the inbox, the content is the cause; go to step 4. If it still lands in spam, the cause is outside the copy: a sending IP, an unusual setup such as a self-hosted server, or a consumer list, where MailReach's warming network is thinner. Compare another sending tool or IP before touching the words.

Step 4: Rebuild the email one element at a time

Start from the original and remove one thing per test, in the order the article suggests:

1. Tracking. If the version without tracking lands, set up a custom tracking domain, using the domain of the sending address, as MailReach's tracking domain article explains.
2. Links. If the version without links lands, drop those links or move them to another domain, never use a URL shortener, and take any email address out of the signature.
3. Wording. Rewrite sentences about money, loans, health, security or insurance. Spam words are learned from complaints, so any word can become one; test with it and without it.
4. Images. If the version without images lands, remove them, change them or host them on another domain.

The troubleshooting article gives away the usual ending: in 80% of cases, links and a missing custom tracking domain are the cause.

Step 5: Fix, then retest under the same conditions

Apply the change in the real tool and run the reference test again, aiming for mostly inbox at the providers you target. Keep the report as the baseline for the next campaign.

Step 6: Hand the test runs to an agent

MailReach runs a hosted MCP server at https://api.mailreach.co/mcp. In Claude, open Settings, then Connectors, add a custom connector named MailReach with that URL and approve the project on the MailReach page that opens. The connection acts with your role on one project and expires 30 days after consent.

Ask the assistant to create a spam test. Its create_spam_test tool uses one credit and returns the seed list to send to. The agent cannot send for you, and that is the useful part: a person sends the campaign email from the production setup, with the test identifier in the body. Then ask it to read the result and list placement per provider, the score for your audience and the warnings. The agent summarises; a person decides which element to change and makes the change in the sending tool. Its tools are limited to what your role allows, and disconnecting a mailbox for good requires an explicit confirmation.

Knowing it worked

The reference test lands mostly in the inbox at the providers you target, and the first campaign after the fix gets more replies per 100 sends than the last one did. To keep it that way, the paid plan runs automatic spam tests with alerts in Slack or through webhooks, per the pricing page.

What MailReach's articles warn about

• Reading inboxes you do not target. Google and Microsoft can file you as spam for irrelevant targeting alone, so a B2B sender ignores the personal inboxes and the other way round.
• Open trackers in the browser. The troubleshooting article calls a Chrome extension that tracks opens terrible for deliverability in 90% of cases.
• Shortened URLs. They send an email straight to spam.
• Aliases. MailReach's alias article recommends separate inboxes and domains instead, because spam on one alias spreads to every address on that mailbox.

One more thing to weigh before upgrading. MailReach is paid per warmed mailbox, $25 a month for each of the first five on its pricing page, while its own article blames links and tracking for most of these cases, and no warming removes a tracking link; buy the warmer when the warming graph shows a reputation problem, not when the control email has already pointed at the copy.

FAQ

Is MailReach's spam test free?

The test on its free spam test page needs no account. Inside the app each test uses one credit, and the All-in-One plan includes 20 a month, per the pricing page.

What counts as a high Reputation Score?

The troubleshooting article treats 85 and higher as high. The score first appears 72 hours after a mailbox is connected.

Do I need a custom tracking domain?

If your tool tracks opens or clicks, yes. By default the tracking link uses a domain shared with the tool's other customers, spammers included, and the tracking domain article says to set your own on the sending domain.