Honeypot links...


GM+ Member Britt just wrote the below in the community (and coined the term "honeypot links")...

"I knew there would be fake clicks, but there could be a lot. I’ll know soon because I set up the system today and from tomorrow, I’ll know."

After a few recent "viral giveaways" I noticed something weird with my email engagement.

The little "follow me on social" buttons above my signature all recieved the exact same number of clicks.

Weird, right?

When I looked into it, the clickers were all addresses from brands like McKinsey, Blackrock etc.

If you've ever seen this, this is what's happening.

Big brands have email security which clicks every link in your email to check it for spam and malicious content before the end recipient gets to see it.

Your ESP logs it as a click, even though it's not a real person.

Which throws off your tracking.

It looks good in the ESP, but the time on site and engagement post click is seconds.

On the surface, it looks liek you're doing well, but really nothing is actually happening with those clicks and there's no genuine engagement.

You don't even know if the real person is actually reading anything you send.

Here's the thing.

Cleaning this with a normal re-engagement campaign that has a "click this to stay subbed" isn't going to work because these systems click EVERY link.

So I needed to come up with a better system.

I adapted the normal re-engagement campaignt to focus on a specific, on-page action rather than a click.

Here's an overview of how it works...

  1. I include a secret, hidden link in the emails only a bot will click
  2. That tags them as a bot
  3. There's a new sequence that starts (similar to a re-engagement email)
  4. The user has to resubmit their email to stay subbed (not based on simple clicks)

I've done a more detailed breakdown in the GM+ Community you. can see by joining here.

I dropped that hidden link in an email yesterday and it got 232 link clicks.

That's 232 clicks per link my emails were getting not from people, but from bots.

So, if you had 5 links, your ESP would record 1160 clicks.

You can see how this would throw off your tracking, right?

In 7 days I'll know if these people are engaged or if they're simple bots, and I'll have a cleaner list at the end of it.

And I'll once again get a better idea of actual link clicks for engagement.

If you're getting suspect clicks and a regular re-engagement sequence isn't cutting it, I recommend trying this.

And if you want to see a video of how I set it up, check out the GM+ Community here.

Pete "Bot killer" Boyle

Vagrants, Vagabonds, and Villains Ltd, Unit 16535, 13 Freeland Park, Wareham Road, Poole, Dorset BH16 6FA
Unsubscribe · Preferences

Growth Models

I've spent ~10 years helping digital brands grow. I share what I know and what I'm experimenting with in this newsletter.

Read more from Growth Models

I’m reworking a bunch of offers and ideas at the minute. And I’d love some feedback on what you’d find most valuable and useful. First though, I’m gonna explain the reasoning behind this little shift - I broke this down in detail in GM+ recently, but want to expand on it a little here. So I’m looking to create 12 offers so I can easily promote one per month to the email list. An easy cycle where, every month, we can offer something that helps people grow their biz. Also, these will double up...

I’ve lost the battle, but not the war with building custom AI Agents to create assets as products. I just can’t get it to replicate the same kind of output as the Custom GPTs. I’ve recorded a video below, but I’ll kick off with the difference in one area of one of these AI Assistants. So, for the $1 Product Strategy Advisor Custom GPT, the output for each stage is like the below… “1. Front-End Offer – $1 Title: “Client Magnet Scripts” What it solves: Freelancers struggle to start sales...

I'll be honest and say 99% of the applications of AI out there are little more than flashy demo reels which offer no real value. Yeah, it's fun to create an image of your dog dancing in a top hat, but it's not exactly useful. I've been working out how to make AI more useful for what we all do. And I've based what I've been doing all on one core psychological driver that's stronger today than it ever has been. People's impatience and need for immediate results. We are, as people, impatient. We...