← All articles

Email list and pop-ups: the bots damaging your deliverability

For a shop we support, 76% of sign-ups in September 2026 came from bots, and almost none from the pop-up. What they cost deliverability, and how to protect every form.

Dots Papers cover for the article on bot newsletter sign-ups and deliverability, Email Marketing section

In brief

  • To grow an email list without damaging deliverability, the first danger is not the pop-up: it is the bots that sign up through any unprotected form.
  • For an online shop we support, in September 2026, 76% of newsletter sign-ups came from bots: 5,332 out of 7,050. The pop-up had received only 16 in total: they came through the footer form.
  • One family of these addresses had received 2,928 sends from 3 to 6 October 2026, for 5,978 soft bounces and not a single open, against 17% among real subscribers.
  • After blocking at the entrance and flagging the records, send failures fell from about 2,450 a week to 0 or 1 per send.
  • The method: protect every form, confirm sign-ups, refuse impossible addresses at the entrance, flag rather than delete, and keep in your lists only people whose consent you can prove.

Growing your email list with pop-ups does not harm deliverability if every sign-up is real and consented. What harms it is bot sign-ups, which come through every unprotected form. For a shop we support, they made up 76% of sign-ups in September 2026, and almost none came from the pop-up. Once they were blocked at the entrance, send failures fell from about 2,450 a week to almost zero.

How we measured The figures come from an online shop we support, measured from 6 to 12 October 2026 in its sign-up database and in its email platform’s events: origin of each sign-up, address pattern, sends, bounces and opens. The post-fix figure comes from the weekly newsletter tracking report of 12 October 2026. No address was deleted: records were flagged and removed from sends.

Where do fake sign-ups come from?

Not necessarily where you expect. For this shop, at 6 October 2026, we counted 7,865 bot sign-ups out of 24,122, a third of all recorded sign-ups. Their pace had picked up: 1,598 in August, 5,332 in September out of 7,050 sign-ups that month, and still 184 out of 286 over the last 24 hours, about eight an hour, day and night.

The sign-up pop-up had received only 16 of them. The bots used the native footer form, present on every page. That form had neither double confirmation nor a bot test: the site’s captcha was only connected to account creation. Every fake address was therefore subscribed immediately, pushed to the email platform, and even received the sign-up confirmation email.

The addresses could be recognised by their shape. A first family added random characters mixed with digits after the domain extension, like “gmail.com5k…”: yet a domain extension never contains a digit. A second family added letters only, giving extensions of 11 to 13 letters, when the real addresses in the database had 2 to 7. The sign-ups came from 734 IP addresses, highly concentrated in two ranges: blocking one IP address at a time would have been useless.

What do bots cost your deliverability?

More than skewed statistics. They degrade the sender reputation that mailbox providers consult to decide whether your emails land in the inbox or in spam.

What we measured What it damages
1,165 addresses failed at every send, without ever being flagged as invalid They were re-sent every newsletter, and skewed every rate
A second family of 1,700 addresses: 2,928 sends accepted from 3 to 6 October, 5,978 soft bounces, 0 opens Sends to mailboxes that do not exist, which mailbox providers see
17% open rate among real subscribers, 0% among these bots The average open rate falls, and you think customers are losing interest
A confirmation email sent to every fake address Extra sends to invalid addresses, from the moment of sign-up

Mailbox providers judge a sender on bounces, complaints and recipient engagement. The Gmail and Yahoo requirements, one-click unsubscribe and a low complaint rate, are set out in our article on compliant email collection. A list swollen with bots degrades exactly those signals.

How do you protect a sign-up form, pop-up included?

  1. List all your entry points. Pop-up, footer, checkout, contact page, competition: bots take the least protected, not the most visible.
  2. Refuse impossible addresses at the entrance. An extension containing digits, or abnormally long, matches no real address. For this shop, the rule was set on the form itself: the address is refused before it is saved.
  3. Confirm the sign-up. With double confirmation, the address only enters the list after a click in the email received: a bot does not click. It is one more step for real people, to be decided with your eyes open.
  4. Add bot protection to every sign-up form, not only to account creation.
  5. Do not push a sign-up to your email platform before it is validated.
  6. Flag, do not delete. The bot records were flagged and removed from sends, not erased: you can go back, and a real person misclassified can still be found.
  7. Flag failures too. An address refused by the email platform must be flagged so that it is not retried every newsletter.

After the entrance block, set on 6 October 2026, and the flagging of 28,294 bot records, send failures fell from about 2,450 the previous week to 0 or 1 per send from 7 October.

Is an incentive pop-up still compliant?

Yes, if it collects genuine consent. The pop-up must say what the person will receive and how often, pre-tick nothing, and not make sign-up a condition of buying. A welcome discount in exchange for signing up is common practice; what matters is that the person knows what they are signing up for and can unsubscribe in one click. The detailed rules, and the proof to keep, are in our article on the form, proof and hygiene of email collection.

Consent also serves deliverability. For the same shop, on 8 October 2026, we removed from the lists accounts with no order and no sign-up, and contacts imported from files with no proof of consent. The main shop’s audience fell from 50,788 to 22,805 addresses. Nothing was deleted: a record comes back to the lists on its own as soon as proof exists.

The open rate trap A list polluted by bots lowers the open rate, and you conclude that customers are losing interest. Conversely, opens triggered by mailbox providers themselves push it up. We explained this in your open rate no longer measures readers: track real clicks and orders instead.

What this says about the bigger picture

An email list is judged by what it brings in, not by its size. A database that grows fast may be one filling up with bots, and every fake address costs a little reputation at every send. Healthy growth is slower, but every address in it is a person who chose to read you.

That work, from protecting forms to measuring sends, is what we do through our email marketing expertise.

FAQ

How do you grow an email list with a pop-up without harming deliverability?

By making sure every sign-up is real and consented: bot protection and refusal of impossible addresses on every form, pop-up included, double confirmation if you accept it, nothing pre-ticked, one-click unsubscribe, and removal of addresses that fail.

How do you recognise a fake newsletter sign-up?

Often by the shape of the address. For a shop we support, bots added random characters after the domain extension: an extension with digits, or of 11 letters or more, when real ones had 2 to 7 letters. They never open and often bounce.

Do bots come through the pop-up?

Not necessarily. In our case, the pop-up had received only 16 bot sign-ups, against 7,865 through the footer form, present on every page and unprotected. Protect every entry point.

Should you delete bot addresses?

No: flag them and remove them from sends. You keep the ability to go back if a real person was misclassified, and a record of what happened.

Is double opt-in mandatory?

No, the GDPR does not require it. But it is the surest way to keep bots out and prove consent, at the cost of one more step for real subscribers.

Can Dotsland help?

Yes. We measure the share of bots in your database, protect your forms and clean your lists without deleting anything. Start by listing your entry points, then get in touch.

Want to apply this to your own business?

Ask us about this topic →

Further reading

Leave a comment

Your email address will not be published. Required fields are marked *

two × one =