Picture a new client site that goes live on Monday. The signup form sits on the homepage, and the new sending domain started its warmup the week before. By Thursday the list has 600 new contacts, and 400 of them came from scripts. The welcome sequence goes out to all 600. Hard bounces come back, a few addresses turn out to be spam traps, and someone whose address was typed in by a bot marks the email as spam. All of that lands on a domain that hasn't built any reputation yet, and nobody has seen a real lead.
This is why the security work on a lead capture form belongs before launch, not after. It's also why the timing matters this week. It's 20 February 2024, and the large mailbox providers have just started enforcing stricter rules for bulk senders. You now need authenticated sending, one-click unsubscribe, and spam complaint rates kept well under 0.3%. A burst of bot signups on a fresh domain is one of the quickest ways to break those limits.
Below is the order of work we use. Certificate first. Then invisible bot protection, double opt-in, lean fields and verification on arrival. Last, you attack your own form in staging. Every step gives you the action, what you should see, and the mistake that catches people at that point.
By the end you will have a form that rejects bots, encrypts every submission and feeds a list you can send to on day one
Here's the finished state. The form loads over HTTPS on the client's final domain, with no browser warnings. Bot protection filters scripted submissions without making real visitors solve puzzles. You've decided on double opt-in for this form and written that decision down. Every real address lands in the right client list, verified, with suppression applied. You also have a short staging test log that shows what you tried and what happened. That's what it means to secure signup forms before site launch. You end up with a list you're happy to send a welcome email to on the first morning.
Before you start, gather these. Without them you'll stall halfway through.
- Access to the client's DNS records or hosting control panel, or a named person on their side who can make changes the same day
- The final production domain, plus every subdomain the campaign will use, such as www and any landing page subdomain
- The staging URL where you'll test the form before cutover
- A launch date and the date the new sending domain began warming up
- Test inboxes on at least three different mailbox providers, including one personal webmail account and one business mailbox
- Your SendTo client account, with the destination list already created
Step 1: Put the certificate on the final domain before the form exists on it
Issue and install the TLS certificate on the production domain before you point DNS at the new site, or at the moment you do. A lot of people ask whether you can do this before the site is publicly live. You can. If you validate through DNS, you add a TXT record that proves you control the domain, and the certificate is issued while the old site is still running. Hosts that validate over HTTP need the domain to resolve to them first. On those hosts, schedule the cutover for a quiet hour and watch the certificate get issued in the first few minutes.
Cover every name the campaign uses. That means the bare domain, the www version and any landing page subdomain such as offers.clientname.co.uk. Then force HTTPS with a permanent redirect, so anyone who types the address without https gets moved across. Fix mixed content too. A single image, font or script loaded over plain http breaks the padlock and can stop the form's own scripts from running.
Be clear about what the certificate does. It encrypts the name and email address while they travel from the visitor's browser to the server, and it removes the "not secure" warning that drives people away from a form. It doesn't stop bots. A script will submit junk over HTTPS just as happily as over HTTP. HTTPS keeps the data private. It does nothing to make the data real.
What you should see is a padlock on every page variant, no browser warnings, and a form whose submit action posts to an https address. Open your browser's developer tools and check the network request when you submit. If it goes anywhere over http, it isn't finished yet.
Step 2: Decide whether the form is embedded or hosted, because that decides who owns the certificate
There are two ways to put a SendTo form in front of visitors. Each one changes who is responsible for the certificate. If you embed the form on the client's own site, the client's site certificate covers that page, so Step 1 is on you or the client's web host. If you put the form on a SendTo hosted landing page, the platform serves the page and handles the certificate.
| Embedded on client site | SendTo hosted landing page | |
|---|---|---|
| Who manages the certificate | The client's host or web team | The platform |
| Best for | Homepage and blog signups, forms inside existing pages | Campaign pages, ad traffic, launches before the main site is ready |
| Main risk | Certificate gaps on www or subdomains | Visitors land on a separate page, so the brand needs to match |
Choose one per campaign and write it in the launch doc, with the exact URL the form will live on. That one line means nobody argues later about who should have renewed what. If the client's new site is running late, a hosted landing page lets the campaign start on time while the main site catches up.
The mistake here is embedding the form on a page builder's preview domain, which changes at launch. Your tests then ran on the wrong origin. Some bot protection and browser security rules care about which domain the form is loaded from, so a form that passed on the preview address can behave differently on the real one. Test the embed on the domain it will actually go live on.
Step 3: Switch on bot protection that real visitors never notice
Bots fill signup forms with three kinds of garbage. Some is plain junk. Some is addresses made to look like typos of real domains. Some is spam traps, which are addresses mailbox providers run to catch senders who mail people who never asked. There's a nastier version called list bombing. A script signs a real person's address up to hundreds of forms at once, often to bury a security alert in their inbox. Your client's welcome email becomes one of those hundreds, and that person marks it as spam.
Follow the chain through. Bot signups put bad addresses on the list. Bad addresses produce hard bounces, spam trap hits and complaints. Those signals land on a sending domain that's still in warmup, when every message counts heavily toward its reputation. Our guide to automated domain warmup schedules for new sending accounts explains why those first weeks are so sensitive. Small client lists don't get a pass either. The bulk sender rules switch on at a volume threshold, but complaint rates and bounces count against every sender, whatever the list size.
Switch on SendTo's bot-protected signup form rather than building your own. The layers that work without bothering anyone are a hidden honeypot field that only scripts fill in, a timing check that rejects forms submitted faster than a person could type, and behaviour checks that look for signs of a real browser and a real visitor. A visible CAPTCHA still has a place. Save it for a form that's already under attack, because every puzzle costs you some real signups, especially on mobile.
What you should see is a form that looks exactly as it did before. A test submission with the hidden field filled in should be refused quietly, and a normal signup should go straight through.
Step 4: Turn on double opt-in for any form that will carry paid or shared traffic
Double opt-in sends a confirmation email, and the contact only becomes active once they click it. In SendTo it's optional and set per form, so make the decision form by form. Our rule is simple. Turn it on for paid ads, giveaways, partner placements and any high-traffic launch. Those are the sources bots and list bombers target. Single opt-in is fine on a low-traffic form, a form behind a login, or a list that came from offline sources such as a shop counter sign-up sheet the client collected in person.
The trade-off works both ways, so be honest with the client about it. Double opt-in loses some real people who never open the confirmation, and on many forms that's a noticeable share of signups. In return, every contact on the list has proved they own the inbox, which shuts down list bombing and spam traps almost completely. Single opt-in keeps every signup but trusts every address. On a new domain under the new bulk sender rules, a smaller list that's real beats a bigger one that's guesswork.
What you should see is the confirmation email arriving in each test inbox within a minute, and the contact showing as pending in SendTo until you click. After the click, the contact turns active and the welcome sequence starts.
Step 5: Ask for fewer fields and validate the ones you keep
Cut the form down to what the first campaign needs. For a welcome sequence that's usually an email address and a first name. Phone number, company size, job title and a free-text "tell us about your project" box can come later through a preference link or a follow-up form, once the person trusts the brand. Every field you remove makes the form quicker to finish and gives a script less to play with.
There's a security reason and a GDPR reason, and they point the same way. Data you never collect can't leak, and it doesn't need protecting, storing, or deleting when someone asks. Collecting only what the stated purpose needs is the plain-English version of data minimisation. Pair that with a clear consent line next to the button, such as "Send me the weekly offers from Harper Garden Supplies. Unsubscribe any time." Leave the checkbox unticked by default if you use one. That's good practice, not legal advice, so have the client's own adviser sign off the wording if they're unsure.
Free-text fields are where bots do their dirtiest work. They paste links, scam copy and code into open boxes, and some of that ends up in a merge tag or an internal notification email. If you don't need a message box on a signup form, don't have one.
Validate the email field as the visitor types. What you should see is a submission like "sarah@gmial" or "sarah.example.com" refused politely, with a short message under the field asking them to check the address. Nothing should be submitted, and the visitor shouldn't lose what they've typed.
Step 6: Route every submission into a list that verifies and suppresses on arrival
Connect the form to the right client list in SendTo, and check it twice. On an agency account with a dozen clients, a form pointed at the wrong list is an easy slip, and it sends one client's leads to another client's welcome email. Name lists after the client and the source, such as "Harper - homepage signup", so the right one is obvious in the dropdown.
Then confirm two things are in place. First, addresses get verified before any send, so dead mailboxes and invalid domains are caught before the welcome email tries to reach them. Second, platform-wide suppression applies. An address that bounced hard last year, or that unsubscribed from this client, can't sneak back onto the list through the new form. Without that, a bot replaying old addresses, or a person signing up a former subscriber out of spite, puts known-bad contacts straight back into the mail queue.
This is the safety net behind Step 3. Bot protection decides who gets to submit. Verification and suppression decide who gets mailed. Something will slip past the form now and then, and this step makes sure it never reaches the new domain as a bounce or a complaint. If a client asks whether checking addresses is worth the effort on a list this fresh, our piece on whether automated email list cleaning is worth it sets out the case in numbers.
What you should see is a test address you know is invalid flagged and held back from sending. A test address you added to suppression earlier should stay suppressed after you submit it again through the form. If either one turns up as an active contact, stop and fix the list settings before launch.
Step 7: Attack your own form on staging before visitors do
You don't need a security team for this. You need about forty minutes, the test inboxes from your setup list, a phone, and a short log where you write down what you tried and what happened. Run the script on staging, then run it again on the production domain within an hour of cutover.
- Submit real addresses from three different mailbox types, such as a personal webmail account, a business mailbox and one more provider. Expected result: each confirmation email arrives within a minute and lands in the inbox, not spam.
- Submit an obviously fake address such as test@nowhere-at-all.invalid. Expected result: the field refuses it, or verification holds it back and it never becomes an active contact.
- Submit the same real address ten times in under a minute. Expected result: one contact in the list, not ten, and no more than one confirmation email in that inbox.
- Open developer tools, fill the hidden honeypot field, then submit. Expected result: the submission is refused quietly and nothing new appears in the list.
- Turn off Wi-Fi and submit from your phone on mobile data. Expected result: the form loads fast, the padlock shows, and the signup goes through without a puzzle.
- Type the page address starting with http instead of https. Expected result: an instant permanent redirect to the https version, with the form intact.
The mistake is testing once, from the agency office, on a desktop, on fast broadband. That's the one setup no real visitor has. Mobile browsers, slower connections and different mailbox providers all treat forms and confirmation emails differently, and the bugs live in those differences.
Check it worked in launch week, and know what to do if it did not
Five days after launch, four signals tell you whether the work held. First, a healthy share of signups confirm their double opt-in, and the rate stays steady from day to day. Second, hard bounces on the welcome send stay close to zero, because verification already caught the dead addresses. Third, there are no bursts of signups at odd hours, like 80 new contacts between 2am and 3am from a local garden centre's homepage. Fourth, rejected and held addresses aren't piling up faster than real ones.
Judge the welcome email by people, not scanners. Corporate security tools open and click links automatically, which inflates numbers and hides a weak list behind good-looking stats. SendTo reports leave out security scanner opens, so the engagement you see on the welcome email comes from real readers. Our guide to tracking real human open rates covers how to read those reports for a client.
If bots got through, work in this order
Pause anything still sending to the new contacts first, because every bad send costs the warming domain. Then look at the signups that arrived in the burst. Same time window, similar-looking addresses, no confirmations? Remove them. Turn on double opt-in if it was off, and add a visible CAPTCHA to that one form for a week or two while the attack runs out. Accept the lost signups for now. Once the bursts stop, take the CAPTCHA off again.
If someone writes in to say they never signed up, take it seriously. That's often the first sign of list bombing. Suppress their address, apologise in one line, and check whether other signups from the same minutes look the same.
Agencies that launch clean forms sell results, and affiliates who refer them earn from it
Every step above protects one thing, which is the client's first real leads reaching the inbox from a domain that still has its reputation. When you secure signup forms before site launch, the welcome email goes to people who asked for it, and the client's first report shows real readers.
On SendTo, the bot-protected forms, address verification, platform-wide suppression and automated domain warmup are already part of the platform. An agency runs this checklist under its own brand, at its own prices, without stitching four tools together. If you write for agency owners or recommend software to them, that's a platform you can point them to with a clear conscience.
Here are the terms of the programme. You earn 20% of the partner subscription each agency you refer pays SendTo, every month that agency pays it. You never get a share of what the agency's own clients pay the agency. Attribution lasts 90 days from the first click, and the first touch wins. Commission is paid monthly by PayPal.
Refer agencies to the platform that ships these protections, and earn 20% of their partner subscription every month they pay.
Join the SendTo affiliate programme


