Tradly MemoryBlog · Growth
Which visitors are going to buy? Read the intent signals.
Most analytics report what happened. Intent scoring tells you who's close. Here are the behaviors that separate serious buyers from casual browsers.
The idea
Not all visits are equal
A visitor who lands on your homepage and leaves in 30 seconds is noise. A visitor who returns next week, searches for your product, reads pricing, and clicks your trial button is revenue in motion. The difference isn't luck — it's intent, and it's readable.
Intent signals are behaviors that correlate with conversion. Capture them, score them, and you can focus every growth lever on the people most likely to buy — while the analytics dashboard just watches everyone.
The signals
Six behaviors that predict buying
| Signal | Why it matters | Strength |
|---|---|---|
| On-site search | Reveals problems they're trying to solve in their words | High |
| Return sessions | Came back on purpose — awareness turned into evaluation | High |
| Deep page reads | Pricing, features, comparison — pages closer to decisions | High |
| CTA clicks | Trial, demo, add-to-cart — commitment moments | Very high |
| Referral source | AI search and curated lists bring qualified traffic | Medium |
| Dwell time | Time on page separates reading from bouncing | Medium |
Scoring
Turn signals into a score
The simplest scoring model is a weighted sum. Each signal adds points; the visitor's total estimates how close they are to buying.
// intent score for maya-84f2
searches: 2 × 10 = 20
return_sessions: 1 × 25 = 25
deep_page_views: 3 × 8 = 24
cart_add: 1 × 30 = 30
intent_score: 99 / 100 (high)
You don't need to perfect the weights on day one. The threshold — "who's above the line?" — matters more than the exact number, and you can tune it as you observe who actually converts.
Capturing
Where the signals come from
Pages
Tag pages by role: awareness (blog, guides), evaluation (features, pricing), decision (trial, demo, checkout). A visitor's page mix tells you their stage.
Searches
On-site search is typed intent — the shopper literally tells you what they want. Store it, count it, and score it.
Interactions
Clicks on primary CTAs, form focus, cart adds, and trial starts are the strongest signals. Capture them with a few lines of event tracking.
Referral source
Where a visitor came from predicts their quality. AI search referrals and curated lists tend to convert better than cold display traffic.
Using it
What to do with a score
Prioritize outreach
Contact high-intent visitors before they go cold.
Build segments
Split high vs. low intent for email and ads.
Trigger at the moment
Recover and personalize while intent is fresh.
In practice
The intent profile, live
Every signal lands in a per-visitor memory card. High-intent visitors surface automatically — no dashboard digging required.
Customer memory / maya-84f2
High intent
score 99Pricing × 2
evaluation
delivery software
typed query
Trial CTA
clicked
2 sessions
returning
Know who's about to buy
Install the pixel and watch intent profiles build themselves. Your next customer is already on your site.
Install the pixel