the tag graveyard
I was auditing a client’s Kit account recently.
He’s been building his email list for years. Careful about it. He’s the kind of person who stores data the way some people save receipts.
You know, in case you get audited by the tax man in 6 years, you got them receipts!
So by the time we jumped on our strategy call, he had a graveyard of a hundred or more tags.
“Clicked This Course”, “Interested in That Course”, “Clicked but didn’t sign up to XYZ” — these are typical behavioural tags people add in their ESPs.
Sound familiar? I’m talking tags for courses they’d visited but not bought, topics they’d browsed, lead magnets they’d downloaded eighteen months ago.
He walked me through it and said, “I thought we might need this data at some point, but I’m open to just starting fresh.”
If you’re capturing every click, view and thought in tags? You need to understand that most of that data will lie to you over time.
tags don’t forget. that’s the problem.
I get it though. You feel productive when you tag every action. You tell yourself you’re building a rich profile of each subscriber… who they are, what they want, where they are in the buyer journey.
Your instincts aren’t wrong! Understanding your audience is genuinely the difference between a list that trusts, buys and refers, and a list that ignores and marks you as spam.
The problem is that tags are static. They don’t expire. They don’t forget. They sit on a subscriber’s profile from the moment they were applied and stay there, claiming to describe a person who may have moved on, leveled up, lost interest, or become your best customer since then.
A tag that says “interested in pricing” doesn’t know that it was added fourteen months ago, and that person has since raised their rates, bought your course, and is now asking totally different questions. The tag just sits there, proud of itself, and completely wrong.
So you build your marketing, your content plan and your launches on top of bad data. You create segments based on it. Maybe you even personalize based on it! Congratulations, you’ve gone and built a highly sophisticated marketing machine for sending the right message to who someone used to be.
Kind of reminds me of that song, “Somebody that I used to know” by Gotye…
When I look at a successful creator’s or newsletter operator’s tag list and see hundreds of behavioural tags, I’m not impressed. I just know I’m going to have to have the talk with them. You know… the “Tag” Talk. Wherein I reveal I will be deleting all of it.
what to tag (and what not to)
Look, the instincts aren’t bad! The point is to personalize a reader’s experience, relationship and buying journey. But relying on Tags in Kit, or whichever ESP you use, is the wrong architecture for personalization and segmentation is not the right approach. Tags are great for things that are permanently true and that a person can have multiples of, and when a good tagging strategy is paired with a consistent, canonical naming convention? “{brand} – {action} – {meta}”, i.e. “Acme – Purchased – Consulting A”, or simply, “{action}: {context}”, i.e. “Downloaded: Lead Magnet Y”
Purchased courses. Completed programs. Products they own. Those things don’t expire. Tagging for them makes sense.
And if you use tags for segmenting or interest, things that naturally expire or decay, anticipate that and build that into your system, with an automation that triggers when “Interested: Course A” is added -> wait 90 days -> remove it. If they click through your Link Trigger and the tag gets added again? Great. Now it accurately mimics real interest. It comes and goes, just like your tags should.
And for even better ESP account organization, you could adopt Jason Resnick’s “review, remove, retain” strategy once every few months, where you review your tags, remove what you don’t need, and retain what’s useful. Nothing like the good ol’ fashioned way of keeping your house (ahem, Kit account) clean. Not everything needs to be automated.
Other signals that decay are intent data, topic affinity, where someone is in their journey (i.e. beginner, advanced, expert) — that data needs to breathe. It needs to decay when it’s no longer true. It needs to update when the person updates.
That’s not what tags do.
the hot air balloon approach to lead scoring
We take more of a “Hot Air Balloon” approach when lead scoring.
If you visit a sales page, the lead score goes up! Registered for a webinar? Up! Booked a Calendly meeting with you? Up and up!
But what if they go quiet for four months, stop opening emails, or visiting your site? That lead score will steadily deflate (or just flat out reset)!
That’s what’s called lead score decay, and the point is your reader has to work to keep that score up in the air, which means when it’s high, you can actually trust it.
So if we don’t use tags to collect this data, you might be wondering what we use. We use Custom Fields (or Properties for you HubSpot folks). In a custom field, only one thing can be true, not many things at once. For example, their current stage, their exact position in your funnel, timestamps that tell you when they moved from welcome sequence into pitch, into evergreen, into a seasonal promo. You can’t be both Advanced AND Beginner — the custom field stores either or. Whereas with a Tag, you can have “Purchased Course A”, AND “Purchased Course B”. Make sense?
what we actually keep
Now back to my client.
He thought he was storing good data, but it was just digital hoarding, and guess what? It all got deleted anyway.
On our call, he asked me what Tags we should keep.
I told him we’d keep the purchase and completion tags, but we’ll cull everything else.
You’re capturing things. You’re building something. The list gets longer and you feel like you know more.
And look, we still collect and track behaviour, intent and signals — we just do it in a more real time way with our in depth RightMessage strategy, using intelligently designed System Flows, and Surveys that sync the data between the ESP and RightMessage, removing tags, updating custom fields, segmenting, lead scoring, decaying interest, and polling for signals that meaningfully could personalize their experience.
Most email systems I see are closer to the first thing than the second. And the owners usually know it, somewhere in the back of their minds, which is probably why the question of what to do with all those tags keeps getting deferred.
the right design problem
You don’t need to know everything about your subscribers.
You need to know the right things, at the right time, with enough confidence to act on them.
That’s a different design problem than the one most people are trying to solve.