Form submissions from your site fail to reach your inbox for one of three reasons: the form handler never sent the email, the mail service rejected the send, or the email arrived but landed in spam or under a filtering rule. To the visitor all three scenarios look identical - they see a “Thank you” screen and leave. You can only find the point of loss by checking the whole chain in order, not by guessing at DNS settings.
As of Q3 2026, major mail providers filter incoming mail more strictly than they did five years ago. Since February 2024, Google has required configured domain authentication from all senders, and from senders of more than 5,000 emails per day - the full set of checks and a spam complaint rate below 0.3% (Google’s sender requirements). Outlook introduced comparable rules for high-volume senders starting in May 2025 (Outlook’s requirements).
On Exceltic.dev support projects we regularly work through the same scenario: the form works, the handler log shows a successful response, and there is no email in the inbox. Meanwhile the site owner is convinced there simply were no leads.
Below is a walkthrough of the whole chain from the button to the “Inbox” folder, a 15-minute diagnostic, and the argument for why email should not be the only place a lead lives.
The failure is silent. The visitor sees the thank-you screen and leaves confident that the request was received. Sales never gets the notification and has no idea who to call back. The loss surfaces weeks later - when someone opens the “Spam” folder or a customer comes back with a complaint.
Where exactly a form submission gets lost on its way to your inbox
A submission’s path from the button to your inbox has five links, and a failure at any one of them looks identical to the visitor.
- Form submission. The browser sends a POST request to the handler’s address. This breaks least often: a wrong
action, a CORS rejection, a JavaScript error before the send. - Form handler. A serverless function, a third-party form service, or a script on your hosting receives the data and triggers the email send. If that call fails, the visitor never sees it: the “Thank you” screen is often shown before the response arrives.
- Sending mail service. Your host’s SMTP, a transactional email service, or a built-in send function. Here the email can be accepted and silently deferred, rejected, or queued with no retry.
- The recipient’s spam filter. Gmail, Outlook, or a corporate mail gateway decides what to do with an email from your domain. This is where most submissions are lost.
- The recipient’s mailbox. The email was delivered, but it is sitting in “Spam” or “Promotions”, or an old filtering rule moved it to the archive.
How a form passes data to a handler in the first place is a separate topic, covered in the article on accepting a form on a static site without a backend. That one covers where the submission goes; this one covers why the notification never arrived even though the submission went through. The job of the diagnostic is to find the last link where the email still existed.
Why your site sends email “as itself” and why that does not work
The main cause of silent loss is a mismatch between the domain in the sender field and whoever actually sent the email.
A typical form puts an address like noreply@yourdomain.com in the From field, while the email itself goes out from your host’s server or from a third-party service. The recipient sees the claimed domain, checks whether that server is allowed to send on its behalf, and finds no confirmation.
The worst variant is a form that puts the visitor’s own address in From so that hitting “Reply” is more convenient. To a mail provider this looks like sender forgery. Some providers reject such emails, others drop them into spam.
The correct setup is simpler. From holds an address on your own domain, the right to send from that domain is confirmed in DNS, and the visitor’s address goes into the Reply-To header and into the body of the email.
Three DNS records that solve most cases
Three DNS records confirm to a mail provider that email from your domain was sent with your knowledge.
SPF: a DNS record that lists the servers allowed to send email on behalf of your domain.
DKIM: a cryptographic signature in the email header that lets the recipient verify the email was sent by the domain owner and was not altered along the way.
DMARC: a DNS rule that tells the recipient what to do with an email that fails the checks, and where to send reports.
What to check in practice:
- SPF lists every real sender: your hosting, your mail service, your newsletter service, your third-party form handler. A forgotten service is the most common finding.
- DKIM is enabled in your mail service’s panel and the public key is published in DNS. Without a signature the email rests on SPF alone, and any forwarding breaks that check.
- DMARC is published with at least a
p=nonepolicy. It blocks nothing, but it turns on reporting: you can see which emails fail the checks and where they are sent from (the protocol overview on dmarc.org).
A note on volume. Form notifications are a handful of emails per day, so the requirements for senders of 5,000 or more emails per day do not formally apply to you. But the minimum of SPF or DKIM is mandatory for everyone, and an unsigned email carrying someone else’s domain in From is filtered harshly regardless of volume.
DNS record examples
; SPF - one TXT record per domain, lists every sender
; ~all - soft fail, -all - hard fail; starting with soft is safer
example.com. TXT "v=spf1 include:_spf.google.com include:sendgrid.net ~all"
; DMARC - handling policy and the address for reports
_dmarc.example.com. TXT "v=DMARC1; p=none; rua=mailto:dmarc@example.com"
Why an email notification is an unreliable way to receive leads
Even with correctly configured DNS records, the email remains the most fragile link in the chain.
It can land in spam because of domain reputation, vanish under a former employee’s filtering rule, or drown in a shared mailbox nobody looks at. A deleted email cannot be recovered, and there is nowhere left to learn that the lead ever existed.
The architecturally correct order is the reverse of the familiar one. The form handler first writes the lead to a CRM or your own database, gets confirmation of the write, and only then sends the notification. In this setup the email is a secondary “something new arrived” signal, not a data store.
The difference shows up at the moment of failure. If the email never arrives, the lead is still sitting in the database with its timestamp, source, and UTM tags. If email was the storage, the lead simply does not exist.
The practical side of this setup is connecting your site to a CRM, payments, and analytics: the form handler writes the lead through an API, and the email and the messenger notification follow, each with its own retry.
Numbers here are more honest given as a range. On a typical project where notifications went to email only with no database write, connecting a CRM reveals that somewhere on the order of 5-15% of leads were passing the sales team by. Naming the exact share before the fix is impossible precisely because lost leads leave no trace anywhere, and the estimate is built by comparing with the period after the switch to writing into a database.
How to diagnose your form in 15 minutes
The check runs from the start of the chain to the end and needs only access to the site, DNS, and email.
1. Send a test submission and read the handler log
Submit the form from a real device and open the handler log right away: the serverless function’s dashboard, the third-party form service’s history, or your hosting logs. What you need is the mail service’s response to the send attempt. A successful response means the email went out and you have to look further down the chain. An authorization error, a timeout, or a rejection means the problem is on your side. If there is no log at all, that is the first thing to fix: without one, diagnosis turns into guesswork.
2. Check the headers of the email you received
If the email made it anywhere at all, open its raw source - in a webmail interface this is usually the “Show original” or “Show source” item. The Authentication-Results header shows the outcome of each check: pass, fail, or none. A value of none means the record is simply not configured. While you are there, compare the domain in From with the domain in Return-Path: the mismatch is exactly what causes the filtering.
3. Check domain reputation and DNS records
Make sure SPF, DKIM, and DMARC are published and visible from the outside - through a public DNS checker or from the terminal. A common finding is two SPF records on one domain, which makes the check invalid. The second is a service you started using six months ago and never added to SPF. If DMARC has a reporting address set, the reports will show every source sending email on behalf of the domain.
Checking the records from the terminal
dig +short TXT example.com
dig +short TXT _dmarc.example.com
dig +short TXT selector._domainkey.example.com
4. Check the “Spam” folder and your mailbox rules
The last step is the mailbox itself. Check “Spam”, “Promotions”, and the archive, then open the list of filters and rules: a rule created long ago for one specific sender can keep messages out of sight for years. On corporate email, also check the quarantine at the gateway level - the message may never reach the user’s mailbox at all.
Common mistakes
- The visitor’s address in the
Fromfield. It looks convenient and reads to the provider as forgery. Its place is inReply-To. - Sending through a built-in function on shared hosting. The IP address is shared with hundreds of other people’s sites, and the sender reputation is not yours.
- Notifications going to one employee’s personal mailbox. When the employee leaves the company, the lead intake channel leaves with them.
- Setting DMARC straight to
p=reject. Without an observation period based on reports, the policy cuts off your own legitimate email. - No delivery testing. Regular testing of the form and its notifications is part of ongoing site maintenance, not a one-time setup at launch.
Who this matters most for
The loss hits hardest at companies that pay for traffic: every lead has a known cost, and silently losing some of them distorts the whole ad payback calculation. For a landing page built for paid traffic, a notification that never arrives is a paid click turned into zero. The second vulnerable group is B2B companies with rare but expensive leads, where one lost form costs more than a year of site support.
Frequently asked questions
Why do form submissions never reach my inbox even though the form shows “Thank you”?
The thank-you screen is triggered by the form handler’s response, not by the fact of email delivery. These two events are separate: the handler returns a response to the browser before the mail service even attempts to send the notification. After that the email can be rejected, deferred, or filtered on the recipient’s side, and none of that affects the visitor’s screen. That is why a thank-you on the site does not confirm the lead reached the sales team. There is only one reliable confirmation - writing the lead into a CRM or database and checking the result of that write.
Emails go to spam only on Gmail, is that normal?
No, that is a typical symptom of missing or incomplete domain authentication. Gmail checks the sender more strictly than many corporate gateways, so the problem shows up there first. Under Google’s requirements, all senders must have SPF or DKIM configured, and the alignment of the domain in the From field with the check result affects the filter’s decision. If emails from the same handler arrive in one mailbox and not in another, start with the DNS records and the Authentication-Results header, not with the text of the email.
Will sending through an external mail service instead of your hosting help?
In most cases yes, but not automatically. An external transactional email service gives you separate sending infrastructure and a log of attempts for every message, which a built-in send function on shared hosting does not have. But you still have to verify the sender domain: add the service to SPF and publish the DKIM key it issues. Without that, you are swapping one unverified source for another. The main benefit of the move is the log itself: you can see whether a message was accepted, deferred, or rejected, and with what wording.
Do you need DMARC for a small site with a single form?
Yes, at least in p=none mode. This policy blocks nothing and cannot break working email, but it turns on reports that show every server sending email on behalf of your domain. For a site with a single form, this is the cheapest way to find a forgotten service that never made it into SPF. As of Q3 2026, having DMARC is mandatory for high-volume senders at Google and Outlook. Tightening the policy to quarantine or reject makes sense after the reports show a clean picture.
What to do next
- Send a test submission and make sure the form handler has a readable send log.
- Check SPF, DKIM, and DMARC on the domain and compare the list of allowed senders with the real one.
- Remove the visitor’s address from the
Fromfield and move it toReply-To. - Rebuild the flow so the lead is written to a CRM or database first and the email is only a notification.
If leads from your site go to email only and you have no idea how many of them are lost - describe the task to Exceltic.dev: which site, which form handler, and where the notifications currently go. We will walk the chain, show you which link the email disappears at, and scope the work of moving lead intake into a CRM.