If you’re still segmenting by “VIP customers (2+ orders)” or “engaged in last 30 days”, you’re not segmenting.
You’re filtering.
And filtering is how good brands stay mediocre.
Filtering gives you buckets.
Segmentation gives you leverage.
Filtering tells you where your customer has been.
Segmentation tells you where they’re going — and what to do about it.
Let’s Define the Real Problem
Most segmentation is built for internal clarity, not customer behavior.
You’re optimizing for neatness — not outcomes.
This happens because most ESPs make it easy to filter:
- Bought X = Add to List Y
- Opened Campaign A = Flag “Engaged”
- Not active in 30 days = Add to “Winback”
What does that actually tell you about the customer?
Nothing.
You don’t know:
- How fast they’re moving
- Whether they’re deepening or shallow in your product stack
- Whether they’re early in exploration or mentally checked out
- If they’re price-sensitive, outcome-driven, or just curious
Filtering shows state. Segmentation shows trajectory.
Why Trajectory > Tags
Imagine you’re running a skincare brand.
Filtering logic says:
- User A: Ordered the Glow Cleanser 2x
- User B: Clicked the promo email
- User C: Signed up for the newsletter
So what? Each of these tells you what happened. Not what’s next.
But trajectory-based segmentation tells you:
- User A reordered in 14 days and is clicking content about layering routines → fast mover, habit forming
- User B clicked 3 promos, bounced from cart, and opened a competitors tab → price shopper, on the edge
- User C hasn’t engaged but spent 12 minutes on your skin quiz → high-intent, wrong medium
Same data source, completely different insight.
This is how you stop guessing and start nudging people based on momentum, not static identity.
Why Most ESPs Lead You Astray
Klaviyo, Drip, Omnisend — they all train you to use last event logic.
“If someone did [action] in [last timeframe], drop them into [flow].”
That sounds fine. But here’s the trap:
It builds a lifecycle system based on isolated triggers — not interconnected states.
You end up with:
- Welcome flows that assume everyone who signs up is ready to buy
- Post-purchase flows that ignore whether someone’s satisfied or confused
- Campaigns that hit both your best repeat buyers and the guy who bounced after using a 20% coupon
Your list is full of false positives — people who look engaged, but aren’t.
And worse? You’re missing the people who don’t look engaged, but are quietly warming up.
Trajectory-based segmentation fixes that.
Build Segments That Model Movement, Not Membership
Here’s how.
Segment: High-Velocity Users
What you look for:
- <20 days between first and second order
- Multiple product page views
- Email click-through >35% across 3 sends
- Viewed UGC or “how-to” content
What they’re telling you:
- They’re getting results
- They’re building a habit
- They’re open to expanding the stack
What to send:
- Subscription prompt tied to effortlessness, not price
- Stack upgrade education
- Identity-focused messaging (“You’re not experimenting anymore. You’ve chosen this.”)
This is where your real LTV lives. If you’re not proactively locking in this segment, you’re leaving money on the table.
Segment: In Decay / Drag
What you look for:
- Last order was 45+ days ago
- Last email click >30 days
- Opened FAQ/support email but didn’t reply
What they’re telling you:
- Something broke in the loop: delivery, expectation, usage
- They’re not gone — but they’re not engaged
- They’re in drag state — and dragging users become churned users unless nudged carefully
What to send:
- “Need a reset?” messaging
- “Still using X? Want help getting more out of it?”
- Education-heavy email with a “no-purchase” CTA
Don’t discount here. Don’t push urgency. You’re not winning them back — you’re clearing friction.
Segment: Category Expanders
What you look for:
- Bought from category A
- Viewed or clicked category B (but didn’t purchase)
- Skimmed routine-building content or bundles
What they’re telling you:
- They trust you — enough to explore
- They’re trying to expand their self-assigned identity (e.g. “I’m becoming a matcha person.”)
What to send:
- Product comparison emails (“Not sure which blend is right for you?”)
- Personalized quiz → product recommendation
- Stack logic (“You’re using X. Here’s what makes it work better.”)
The opportunity here isn’t upsell. It’s alignment. Show them a path that makes their behavior make sense.
Segment: False Positives (Openers, No Clicks)
What you look for:
- 4+ opens, 0 clicks
- Email activity but no sessions
- Repeated interaction with discounts/promos only
What they’re telling you:
- They’re mentally gone, but technically “active”
- They’re skimming, bored, possibly price trained
- They’re a cost center in your ESP
What to send:
- Repermission flow: “Still want to hear from us?”
- Channel switch: SMS-only, monthly digest
- Optional: sunset them
This is your silent list killer. Filtering says “they’re engaged.” Segmentation says: stop wasting time.
How to Implement Trajectory-Based Segments
Here’s what an operator builds in their ESP:
| Trigger Type | Logic Example | Outcome |
|---|---|---|
| Purchase velocity | Order 2 date – Order 1 date < 21 days | Add to High-Momentum segment |
| Engagement slope | Email clicks declining over last 3 campaigns | Route to “Decay” handling |
| Category depth | Viewed >2 product types without purchase | Push bundle/cross-sell content |
| Time since event | Last engaged with support/FAQ > X days ago | Trigger usage check-in |
| Intent score | Time on site + add-to-cart + quiz = intent > threshold | Trigger conversion content |
Then you build flows that start based on trajectory, not triggers.
What Happens When You Don’t
Let’s make it painfully clear.
Without trajectory segmentation:
- You send winback emails to people who haven’t left
- You upsell cold leads who haven’t seen value yet
- You push bundles to people who haven’t even finished their first unit
- You burn the attention of your best users with irrelevant reminders
- You train your list to ignore you
You don’t just lose conversions. You lose the ability to learn from behavior.
Stop Sending Email Based on Events. Start Sending Based on Energy.
Filtering is event-based. Segmentation is energy-aware.
Energy = where your user is in their journey — and how far they’re drifting toward or away from a goal you can support.
That’s what good email marketing actually is:
Not a flow. Not a campaign. But a conversation that responds to energy.
If your ESP doesn’t reflect that? You’re flying blind.
Want Real Segments That Actually Convert?
At Ghost Goat, we design segmentation logic that goes way beyond “opens and tags.”
We build behavior-based campaigns that don’t just hit — they guide.
DM to book a strategy call — and let’s turn your list into a retention engine.
Leave a Reply