Tool AT-W5REV. 2026-A
Are your e-mails landing in spam? Check SPF, DKIM and DMARC
When e-mails land in spam, records are usually missing in your domain’s DNS: SPF, DKIM or DMARC. The e-mail check queries them and tells you in plain words what is missing, what that means for your e-mails and how to fix it with your provider. You only enter the domain, for example your-company.de.
- free
- no sign-up
- 5-minute cache
01 / DNS CHECK
Enter a domain, read the result
DNS check · SPF, DKIM, DMARC
The check reads public DNS records and assesses them. Whether an individual message arrives also depends on its content, the volume sent and the reputation of the sending server, and nobody can see that from outside.
02 / BACKGROUND
Why do e-mails land in spam?
A receiving server initially knows only what the sender field says. And anyone can write anything there. So before it puts a message in the inbox, it looks in your domain’s DNS: may the server that is delivering right now send for you? Does the digital signature match the key you published? And what should happen if neither is right?
If the answers are missing, it has to guess. Gmail and Yahoo have guessed much more strictly since February 2024, Outlook.com since May 2025. You do not see the consequences. The message is in your Sent folder, it never reaches the customer, and nobody calls to tell you.
Then there are mistakes that look like a complete record from outside: two SPF records side by side, an SPF with eleven instead of at most ten lookups, a DMARC that has been on monitoring for years. The check covers exactly these cases. The guide Why do my e-mails land in spam? explains the background at leisure.
03 / THE THREE RECORDS
What do SPF, DKIM and DMARC do?
SPF
TXT on the domain
A list of the servers that may send for your domain, for example v=spf1 include:_spf.google.com ~all. Exactly one record per domain, with at most ten references that trigger a DNS lookup.
RFC 7208
DKIM
selektor._domainkey
Your mail server signs every message, and recipients find the matching public key in your DNS. If the signature matches, the message has not been altered in transit. Gmail requires at least 1024 bit, 2048 is recommended.
RFC 6376
DMARC
_dmarc
Tells recipients what should happen to messages for which neither SPF nor DKIM passes in alignment with your sender domain: only report them (none), send them to the spam folder (quarantine) or reject them (reject). In addition there are daily reports on who is sending in your name.
RFC 9989 (until 2026: RFC 7489)
Many people get stuck at one point: DMARC requires that the domain that passed SPF or DKIM matches the sender domain. A newsletter tool that sends via its own domain passes SPF and still fails DMARC, because the sender field shows your business address. The solution is almost always a DKIM key of the service under your domain. Most newsletter providers guide you through this on a dedicated setup page.
04 / REQUIREMENTS
What do Gmail, Yahoo and Outlook.com require?
The three big mailbox providers have tightened their rules for senders. For business mail in Germany this is noticeable, because many customers use private addresses with these services. The table summarises what the providers themselves have published.
| Provider | Since | Who it is for | Required | Otherwise |
|---|---|---|---|---|
| Gmail, all senders | 1 February 2024 | anyone writing to private Gmail accounts | SPF or DKIM, valid reverse DNS, TLS, spam rate below 0.3% | Spam folder or rejection |
| Gmail, bulk senders | 1 February 2024, tightened from November 2025 | from around 5,000 messages in 24 hours to Gmail accounts; the classification stays permanently | SPF and DKIM (at least 1024 bit), DMARC with at least p=none, sender domain matches SPF or DKIM, one-click unsubscribe | since November 2025 also temporary and permanent rejections |
| Yahoo Mail, bulk senders | February 2024 | Yahoo gives no fixed volume threshold | SPF and DKIM, DMARC with at least p=none (must pass), one-click unsubscribe, unsubscribe requests processed within two days, spam rate below 0.3% | Spam folder or rejection |
| Outlook.com, Hotmail, Live | 5 May 2025 | Domains sending more than 5,000 messages a day to these addresses | SPF and DKIM passed, DMARC with at least p=none, aligned with SPF or DKIM | initially the junk folder; rejection with “550 5.7.515” has been announced |
The threshold of 5,000 messages a day sounds like a large-scale sender. A newsletter to 6,000 Gmail addresses exceeds it with a single mailing, and Google then treats the sender permanently as a bulk sender. So anyone who sends circulars at all is best off setting up all three records straight away. For all other mail, Gmail and Yahoo require SPF or DKIM and a spam rate below 0.3%.
Sources: Google: Email sender guidelines · Yahoo Sender Hub: Best practices · Microsoft: New requirements of Outlook.com for bulk senders
05 / TYPICAL MISTAKES
Which mistakes does the check find most often?
- Two SPF records. Every time a new service came along, another one was simply created. The result is not twice as good but invalid: recipients then evaluate neither of them. Everything belongs in one record, and the check proposes the merged version straight away. In a measurement of 12 million domains, 2.9% of SPF records were faulty or ineffective Czybik 2023.
- More than ten lookups. Every include counts, and the references inside the included records count too. If you include mailbox, newsletter, invoicing software and helpdesk, you quickly reach eleven. The check follows every reference and gives the total.
- Too broad. “+all” lets every server in the world send in your name, “?all” says nothing at all. The right choice is “~all” or “-all”. An ip4 range that releases a whole data centre has a similar effect: 34.7% of domains allow over 100,000 addresses Czybik 2023. The check reports both.
- The newsletter tool is missing. Circulars, shop or accounting were added two years after the mail was set up. It is precisely this mail that then lands in spam, while the normal mail gets through.
- DMARC without reports or on none for years. Without an rua address nobody sees who is sending in your name. And on none the domain stays open to forgers. If the reports go to an external service, it must have agreed to this in the DNS; the check tests that too Ashiq 2023.
- DKIM created but not switched on. Typical for Microsoft 365: the CNAME records are in the DNS, but in the Defender portal DKIM is still off. The check recognises this by the reference that points nowhere. Whether an individual message really passed is shown by its header; the e-mail header analysis reads it.
06 / RESEARCH
What the research says
SPF, DKIM and DMARC are well studied. The measurements of recent years show above all how big the gap is between “record present” and “record works”. Seven findings that have fed into the check’s assessment:
- SPF is widespread, all three together are rare. In a measurement of 12 million domains, 56.5% had an SPF record Czybik 2023. DMARC has grown tenfold in five years but was only at around 11%; about 3% used all three methods Tatang 2021.
- Recipients really do check. Up to 85% of the mail servers examined evaluate SPF, over half all three methods Deccio 2021. A broken record is therefore not without consequences.
- DKIM often fails on maintenance. Of the domains with DKIM, 2.9% were set up incorrectly, and even well-known providers had problems with key management Wang 2022. Weak keys also turned up Tatang 2021. The check therefore estimates the key length.
- Subdomains are an entry point of their own. A measurement of 236 million domains found many faulty SPF and DMARC records and forged mail via subdomains Maroofi 2021. That is why the check warns on sp=none.
- The fear of blocked mail holds DMARC back. Administrators surveyed named it as the main reason for not introducing DMARC Hu 2018. The stepped route via none and quarantine removes exactly this risk.
- Without instructions, recipients decide inconsistently. Of 35 large mailbox providers, many let forged mail through despite having the checking methods Hu, Wang 2018. DMARC is the instruction that changes this.
- Correct records are no guarantee. Several studies show ways to get around the check at the recipient despite correct records, for example via forwarding or header lines read differently Chen 2020Shen 2021 Liu 2023. The providers have closed much of this. For you it means: DKIM in addition to SPF, because with forwarding only the signature remains.
07 / SETUP
How do I set up SPF, DKIM and DMARC?
- Write down all senders. Mailbox provider, newsletter, invoicing and inventory software, shop, the website’s contact form, even the copier with scan-to-mail. Everything that sends with your domain in the sender field.
- Build one SPF record. One include per service, ~all at the end. Then run the check: at most ten lookups.
- Switch on DKIM for every service. The service generates the key and gives you the DNS record. If it gives you the choice, take 2048 bit.
- Set DMARC to none, with a report address. For example
v=DMARC1; p=none; rua=mailto:[email protected]. The address must exist. - Test every sending route separately. A message to a Gmail account, then “Show original” there: SPF, DKIM and DMARC should show PASS. For the newsletter just as for the mailbox.
- Read the reports, then quarantine. After two to four weeks the reports show whether your own mail fails. If not, set p=quarantine.
- Reject last. After a further two to four weeks without your own mail in spam, switch to p=reject. If a new service is added later, it belongs in SPF and DKIM before its first send.
You maintain the records where your domain is managed, usually with the hosting provider. The result above shows where to find the menu for the detected provider. If you would rather not do it yourself: we take it on as a one-off job via IT support or on an ongoing basis as part of ongoing IT support, and what that costs is on the price list. If your website’s contact form runs with us (web design), its sending belongs in the same SPF record.
08 / LIMITS
What the check does not see
- DKIM keys under an unknown name. The search covers 35 common selectors. Not found does not mean none exists.
- Whether a real message passes. Whether the sender domain matches SPF or DKIM and whether the signature is right is shown only by a sent message. That is what the Gmail test is for.
- Reputation and blocklists. Whether your provider’s server is on a blocklist or how high your spam rate is, the check does not examine.
- The MTA-STS policy. The DNS record is checked, not the policy file on the web server.
09 / FOR AI ASSISTANTS
Using this tool from an AI assistant
Assistants such as Claude or ChatGPT that speak the Model Context Protocol (MCP) can call this page’s calculation directly instead of rebuilding it. The server is public, needs no sign-in and stores nothing. The answer always includes the link to this page, the date of the data and the limits of the result.
- Endpoint
https://aureum-tech.com/mcp- Tool
werkzeug_email_check- Transport
- Streamable HTTP, read-only
| Name | Required | Meaning |
|---|---|---|
domain | yes | Domain or e-mail address, e.g. your-company.de or [email protected]. The part after the @ is checked. |
Questions where an assistant should call it
- “Why do the e-mails from mycompany.de land in spam at the customer’s end?”
- “Are SPF, DKIM and DMARC set up correctly for [email protected]?”
- “Does our domain meet the requirements of Gmail and Outlook.com for newsletters?”
- “What do I have to enter at IONOS so that DMARC passes?”
How the connection works and what MCP can do in your own business is explained on the page about MCP integrations.
10 / COMMON QUESTIONS
Common questions about SPF, DKIM and DMARC
What is the difference between SPF, DKIM and DMARC?
SPF is a list of the servers that may send for your domain. DKIM is a digital signature on every message, checked against a key in your DNS. DMARC determines what happens to messages for which neither SPF nor DKIM passes in alignment with the sender domain: only report them (none), send them to the spam folder (quarantine) or reject them (reject). Only all three together make a forgery detectable.
Is DMARC with p=none enough?
For the requirements that Gmail, Yahoo and Outlook.com set for bulk senders, yes, because p=none is enough there. But it does not protect your domain: forged messages are still delivered, and you merely receive reports about them. p=none is the start. Once the reports show that all your own sending routes pass, quarantine follows, and then reject.
Should the SPF record end in ~all or -all?
Either is fine. With -all, recipients are meant to reject other servers; with ~all they mark them as suspicious. Google’s guide uses ~all, Microsoft uses -all for Microsoft 365. If DMARC is set to quarantine or reject, DMARC decides anyway. What is wrong is +all, which lets anyone send, and ?all, which says nothing.
How many entries may an SPF record have?
There is exactly one SPF record per domain. It may contain at most 10 terms that trigger a DNS lookup: include, a, mx, ptr, exists and redirect, counting the references inside the included records (RFC 7208). ip4 and ip6 do not count. At the eleventh lookup the whole record counts as an error, without any message to you.
Why does the check not find my DKIM key?
A DKIM key sits under a name, the selector, and that name appears only in the header of a sent message. The check tries 35 common selectors, including those of Google, Microsoft 365, IONOS, STRATO and mailbox.org. Some providers assign the name per customer. The safe way to find out is a test message to a Gmail account: “Show original” shows DKIM as PASS or FAIL.
Do Gmail’s rules also apply to small businesses?
In part, yes. Since February 2024 Gmail has required every sender to have at least SPF or DKIM, a valid reverse DNS and a spam rate below 0.3%. The stricter rules with DMARC apply from around 5,000 messages a day to Gmail accounts. A single newsletter to enough Gmail addresses is enough for that, and the classification then stays permanently.
How long does it take for a DNS change to take effect?
Usually minutes to a few hours. Every record carries a validity period (TTL), often one hour; for that long other servers may keep using the old version. This check also remembers a result for five minutes. So after a change, do not check straight away but after an hour, and then do the Gmail test.
What happens to the domain I enter?
It is sent to our server, which looks up the public DNS records and sends the result back. The domain and result are not logged. For follow-up checks, both are held in RAM for up to five minutes. To prevent misuse, the server counts queries per visitor using an anonymous check value that is cleared when the server restarts.
11 / STUDIES AND SOURCES
What the tool is based on
The reference values and notes above rest on these 12 academic papers. Each has been checked via its DOI or the publisher’s page (last on 13 November 2026): it exists, it has not been retracted, and its finding still holds for the question today.
Czybik, S.; Horlboge, M.; Rieck, K. (2023): Lazy Gatekeepers: A Large-Scale Study on SPF Configuration in the Wild. Proceedings of the 2023 ACM Internet Measurement Conference (IMC), pp. 344–355
Finding Of 12 million domains examined, 56.5% had an SPF record. Of these, 2.9% were faulty or ineffective, and 34.7% of the domains allowed more than 100,000 IP addresses to send, which makes forgery easier.
In the tool The tool reports invalid terms first and warns if an SPF record itself allows more than 100,000 IPv4 addresses.
doi.org/10.1145/3618257.3624827 · Conference paper
Tatang, D.; Zettl, F.; Holz, T. (2021): The Evolution of DNS-based Email Authentication: Measuring Adoption and Finding Flaws. 24th International Symposium on Research in Attacks, Intrusions and Defenses (RAID), pp. 354–369
Finding SPF is the most widespread method. DMARC has grown tenfold in five years but was only at around 11%, DKIM at at least 13%. Only about 3% of domains used all three, and even large domains had invalid SPF records or weak DKIM keys.
In the tool That is why the check examines all three records together and estimates the length of every DKIM key found.
doi.org/10.1145/3471621.3471842 · Conference paper
Wang, C.; Shen, K.; Guo, M. et al. (2022): A Large-scale and Longitudinal Measurement Study of DKIM Deployment. 31st USENIX Security Symposium, pp. 1185–1201
Finding Of the million most visited domains, 28.1% had DKIM set up, 2.9% of these incorrectly. Problems with key management also turned up at well-known providers.
In the tool Basis for the search among common selectors, the detection of revoked keys and the note on 2048 bit.
www.usenix.org/conference/usenixsecurity22/presentation/wang-chuhan · Conference paper
Maroofi, S.; Korczyński, M.; Hölzel, A.; Duda, A. (2021): Adoption of Email Anti-Spoofing Schemes: A Large Scale Analysis. IEEE Transactions on Network and Service Management 18 (3), pp. 3184–3196
Finding A measurement of 236 million domains showed that a large share do not set up SPF and DMARC correctly, so forged mail ends up in the inbox. Subdomains proved to be an attack route of their own.
In the tool The check evaluates sp= separately and warns if subdomains with sp=none remain unprotected.
doi.org/10.1109/TNSM.2021.3065422 · Journal article
Deccio, C.; Yadav, T.; Bennett, N. et al. (2021): Measuring email sender validation in the wild. Proceedings of the 17th International Conference on emerging Networking EXperiments and Technologies (CoNEXT), pp. 230–242
Finding On the receiving side, up to 85% of the mail servers examined checked SPF, and over half applied all three methods.
In the tool That is why a missing or invalid SPF record counts as red: most recipients really do evaluate it.
doi.org/10.1145/3485983.3494868 · Conference paper
Hu, H.; Peng, P.; Wang, G. (2018): Towards Understanding the Adoption of Anti-Spoofing Protocols in Email Systems. 2018 IEEE Cybersecurity Development (SecDev), pp. 94–101
Finding DMARC was then in use on around 5% of domains. Mail administrators surveyed named as obstacles above all the worry about blocking genuine mail, plus weak incentives and the effort of introduction.
In the tool That is why the check proposes the stepped route none, quarantine, reject and not reject straight away.
doi.org/10.1109/SecDev.2018.00020 · Conference paper
Hu, H.; Wang, G. (2018): End-to-End Measurements of Email Spoofing Attacks. 27th USENIX Security Symposium, pp. 1095–1112
Finding Of 35 large mailbox providers, most had the checking methods but still let forged mail into the inbox in some cases. Only 9 showed a visible warning on unverified mail.
In the tool This explains why a missing DMARC record counts as red: without an instruction from the domain owner, recipients decide inconsistently.
www.usenix.org/conference/usenixsecurity18/presentation/hu · Conference paper
Ashiq, M. I.; Li, W.; Fiebig, T.; Chung, T. (2023): You've Got Report: Measurement and Security Implications of DMARC Reporting. 32nd USENIX Security Symposium, pp. 4123–4137
Finding DMARC reports are often set up without regard to security. Misconfigured servers could be abused in such a way that a single message triggered up to 1,460 times as many reports.
In the tool The check tests whether external report recipients have agreed to accept them, and advises using your own, existing address.
www.usenix.org/conference/usenixsecurity23/presentation/ashiq · Conference paper
Durumeric, Z.; Adrian, D.; Mirian, A. et al. (2015): Neither Snow Nor Rain Nor MITM…: An Empirical Analysis of Email Delivery Security. Proceedings of the 2015 Internet Measurement Conference (IMC), pp. 27–39
Finding The big providers already encrypted and authenticated, but of over 700,000 other mail servers only 35% encrypted successfully, and only 1.1% of domains had a DMARC policy. In seven countries, over 20% of the mail to Gmail arrived unencrypted because of attackers on the network.
In the tool This foundational work explains why the check also queries MTA-STS and TLS-RPT, even though both are only nice to have.
doi.org/10.1145/2815675.2815695 · Conference paper
Chen, J.; Paxson, V.; Jiang, J. (2020): Composition Kills: A Case Study of Email Sender Authentication. 29th USENIX Security Symposium, pp. 2183–2199
Finding Because mail servers and mail programs read the same message differently, the authors found 18 ways to bypass the sender check. All 10 providers and 19 programs tested were vulnerable to some of them.
In the tool That is why the check says openly that correct DNS records are the prerequisite but no guarantee against forgery.
www.usenix.org/conference/usenixsecurity20/presentation/chen-jianjun · Conference paper
Shen, K.; Wang, C.; Guo, M. et al. (2021): Weak Links in Authentication Chains: A Large-scale Analysis of Email Sender Spoofing Attacks. 30th USENIX Security Symposium, pp. 3201–3217
Finding All 30 mail services and 23 mail programs tested could be fooled with at least one new type of attack. Several providers, including Gmail and Yahoo, have since improved.
In the tool Supports the tool’s limit: it assesses the DNS side, not the check at the recipient.
www.usenix.org/conference/usenixsecurity21/presentation/shen-kaiwen · Conference paper
Liu, E.; Akiwate, G.; Jonker, M. et al. (2023): Forward Pass: On the Security Implications of Email Forwarding Mechanism and Policy. 2023 IEEE 8th European Symposium on Security and Privacy (EuroS&P), pp. 373–391
Finding Automatic forwarding at 16 large providers and 4 list services undermines the assumptions of SPF, DKIM and DMARC. In combination, mail could be forged in the name of tens of thousands of well-known domains.
In the tool That is why the check does not treat DKIM as an add-on to SPF: with forwarding, SPF drops out, while the signature remains as long as nobody alters the message.
doi.org/10.1109/EuroSP57164.2023.00030 · Conference paper
Standards, authorities and manufacturer data
- RFC 7208: Sender Policy Framework (SPF) · IETF
- RFC 6376: DomainKeys Identified Mail (DKIM) · IETF
- RFC 9989: DMARC (replaces RFC 7489 since May 2026) · IETF
- RFC 9990: DMARC aggregate reports · IETF
- RFC 8461: MTA-STS · IETF
- RFC 7505: Null MX · IETF
- Google: Email sender guidelines · Google
- Google: Frequently asked questions about the sender guidelines · Google
- Yahoo Sender Hub: Best practices · Yahoo
- Microsoft: New requirements of Outlook.com for bulk senders · Microsoft
12 / MORE TOOLS
Still have a question?
Once the mail is running, the next question is often how quickly the business is back at work after an outage. The IT emergency check answers that in a few minutes.
- AT-W2 · Home & business
Windows 10 check
Does this PC still get security updates, and what needs doing by October 2027? - AT-W3 · Home & business
Wi-Fi planner
How many access points or mesh units does your home, office or warehouse need? - AT-W4 · Businesses
IT emergency check
How quickly could your business get back to work after an outage? - AT-W6 · Businesses
UPS calculator
Which UPS keeps your server, router and till running long enough during a power cut? - AT-W8 · Businesses
XRechnung viewer
Received an XRechnung as an XML file? This makes it readable without the file leaving your computer. - AT-W9 · Home & business
Password generator
A strong password or a memorable passphrase, generated in your browser.
Would you rather not set the records yourself? We check your domain, enter SPF, DKIM and DMARC with the hosting provider and go through the first DMARC reports with you.
+49 3744 4399760[email protected]Mon–Fri 8:00–17:00 · Straße der Jugend 5, 08228 Rodewisch