How Lens reads
a page


A short explanation of how the system works.
In plain language.

Throughline · Lens · Calidescope LLC
The big picture

Three steps

Step one Inputs A page, plus three rivals
Step two Engine Six classic, well-tested ideas
Step three Output A read, in five sections

We'll walk through each one.

What's in this deck · click any slide

Start anywhere

Act I · Inputs · what you put in

You give Lens
a URL.

That's it. Plus three competitors — we pick them for you.

https://lancea.ai
https://lancea.ai
+ persado.com
+ signalai.com
+ quid.com
Brand URL

Your page is the brand

homepage
homepage
/product
product page
/launch
launch post
readsone URL

Any page. Pick the one that represents what you want read.

Competitors

Three competitors
come along

You can't judge a page by itself. You need three other pages trying to do the same thing. Lens picks them — swap any you like.

Brandlancea.ai
Rivalpersado.com
Rivalsignalai.com
Rivalquid.com
Brand · then the field
Why comparison matters

You can't tell heavy
without holding light

Your brand alone ?
Brand + the field NOW YOU CAN WEIGH IT

Without the field, every brand looks normal.

What Lens reads

It reads the page
the way an editor would

lancea.ai
TextEvery word actually on the page
StructureWhich line is the headline, which is body, which is footer
PositionWhat came first. What came last
VoiceFormal or casual. Direct or hedged
The harvest

Four pages become
organized data

lancea.ai
persado
signalai
quid
Organized data · one shape per page
headlinePredict message performance before you spend
subheadAdaptive, specialized AI agents
features45+ agents · MARA · 5-axis testing
proof120,000 campaigns
footerFrom insight to influence

Same questions of every page. Same shape. Now the engine can begin.

Act II · Engine · what the engine does

Six classic,
well-tested ideas

The engine isn't magic. It's six time-tested ideas about how arguments work. You'll meet each one.

Genres Stasis Appeals
The engine
Pyramid Ideas Rhythm
The six frameworks · map

One picture, six tools

Aristotle's genres
What KIND of argument is this?
Stasis
What's the argument staked on?
The three appeals
Logic, credibility, or feeling?
Minto's Pyramid
Does the argument stand up?
Adler's Great Ideas
What words is it leaning on?
Essential Rhythm
What beats does the story walk?
The engine · the one-sentence read

Four jobs every page is doing

This page speaks tomarketing teams, asksthem to predict message performance before they spend, promisesadaptive, specialized prediction, and backsit with 120,000 campaigns.

Click any verb to see where on the page it came from

The read · five sections, in order

Five sections

§01The AppealsWhat kind of argument is this?
§02The BreakdownHow does each appeal do its work?
§03The ThroughlineHow does the story sequence?
§04The IdeasWhat vocabulary is it leaning on?
§05The RecommendationsWhere do you act first?

Scan it in five seconds. Or read it in twenty minutes.

§01 · The Appeals · Aristotle's genres

What KIND of argument

Deliberative
consulting deck
Argues what you should do next.
Judicial
defense brief
Argues what happened, and who's right.
Epideictic
product launch
Shows how something works, and praises it.

Walrus.dev shows how something works → epideictic. Lancea.ai argues what you should do → deliberative.

§01 · The Appeals · the question at issue

Five rungs of an argument

05ProcedureIs this the right way to do it?
04PolicyShould we do it at all?
03QualityIs it good? Is it worth it?
02DefinitionWhat kind of thing is it?
01FactDoes it even exist?

Every argument is staked on a question — at one of these five levels.

§01 · The Appeals · logic · credibility · feeling

Logic, credibility, feeling

Logic
Reasons, proof, numbers. Speaks to the head.
Credibility
Who's saying it, and why trust them.
Emotion
What it makes the reader feel.
9Emotion
25Credibility
66Logic
Lancea.ai · the mix This is what we measure.
§01 · The Appeals · the Core Sample

The Core Sample

9Emotion
25Credibility
66Logic
Lancea.ai

Like a geologist cutting through rock: feeling on top, credibility in the middle, logic as the bedrock. Each layer's height is how much of the argument runs on it.

The field, for comparison
persado
signal
quid
lancea
Logic share across the four. Lancea sits high — but not alone.
§01 · The Appeals · the read

What kind of argument

9Emotion
25Credibility
66Logic
Lancea.ai
Logic-led

This page argues, first and most, through logic — two-thirds of the weight sits in reasons and proof.

Credibility plays a real supporting part. Emotion is nearly absent — the page reasons with you; it doesn't reach for feeling.

§02 · The Breakdown · Minto's Pyramid

One claim. Three pillars. Proof beneath

Governing thoughtAdaptive prediction beats static testing
Pillar 01Specialized agents
Pillar 02Stress-tested rigor
Pillar 03Proven at scale
45+ specialized agents
MARA · 5-axis testing
120,000 campaigns

A clean argument: one claim up top, three reasons holding it, proof under each.

§02 · The Breakdown · three deep looks

How each appeal does its work

Logic
The Pyramid
Does the claim have three real reasons, each with proof?
Credibility
The moves
Whose authority is named — and is it the kind a reader can check?
Emotion
The words
Which words were chosen, and what feeling do they carry?

Three appeals, three deeper looks.

§02 · The Breakdown · the read

How each appeal works

The dominantLogicTwo-thirds of the argument. Three pillars, each carrying its own proof.
The secondaryCredibilityNames the scale — 120,000 campaigns — but not the people behind it.
The absentEmotionAlmost no feeling-words. Nothing here is asking the reader to feel.
§03 · The Throughline · the five beats

Every story walks the same beats

1
Setup
Static messaging is obsolete
2
Problem
AI tools lean on biased synthetic personas
3
Claim
Lancea orchestrates 45+ specialized agents
4
Proof
MARA stress-tests across 5 axes
5
Payoff
From insight to influence

Miss a beat, or play them out of order, and the argument breaks.

§03 · The Throughline · reading the order

Reading the beat sequence

1
Setup
Static messaging is obsolete
2
Problem
Biased synthetic personas
3
Claim
Arrives before credibility is earned
OUT OF ORDER
4
Proof
MARA · 5-axis testing
5
Payoff
From insight to influence

Lens marks what's there, what's not, and what's where it shouldn't be.

§03 · The Throughline · the read

How the story sequences

1
Setup
Static messaging is obsolete
2
Problem
Biased synthetic personas
3
Claim
45+ specialized agents
4
Proof
MARA · 5-axis testing
5
Payoff
From insight to influence

Five beats, in order. Each one shows what the page said for it.

§04 · The Ideas · Adler's Great Ideas

The shared vocabulary of every argument

Adler mapped the 102 ideas people argue about — Quality, Change, Cause, Time. Lens uses his list so every brand is read on the same words.

Ideas this page is in
Quality17 Change13 Means / End11 Knowledge7
Canonical — but absent here
Cause Time Relation

What ideas the page is in — and which it's not in — both matter.

§04 · The Ideas · matching the canon

Matching your words to the canon

From the page

Purpose-built, specialized AI agents deliver adaptive, intelligent messaging — purpose-built for teams who need it to keep getting better.

In
Quality Change Means / End
Absent canon
Cause Time Relation
§04 · The Ideas · the read

What vocabulary you're leaning on

Top three in play
Quality17 Change13 Means / End11
Canonical gaps — the field uses these, this page doesn't
Cause Time Relation
§05 · The Recommendations · worth trying or holding

Worth trying, or holding — never invented

Worth trying
Pick one anchor per idea
“Adaptive” for Change, “specialized” for Quality — so the page stops naming one idea three ways.
Holding
The proof stack is already balanced
Logic, credibility, and emotion sit near the field median. Nothing to copy — the pencil stays in the drawer.

Honest is more useful than complete.

§05 · The Recommendations · where to act first

Not advice. The new copy.

LANCEA.AI § Ideas Worth trying
Semantic consistency move

Pick one anchor per idea — “adaptive” for Change, “specialized” for Quality.

From

Class-leading, top-tier, best-in-class messaging prediction with proven results across 120,000 campaigns.

To

Adaptive messaging prediction. Specialized AI agents. Proven across 120,000 campaigns.

WhyThe reader stops re-deciding whether “class-leading,” “top-tier,” and “best-in-class” name different things — and “adaptive,” already the tagline, becomes the one word that stays.

BasisPersado, Signal AI, and Quid each hold a single anchor for their core idea; Lancea’s six-way scatter on Quality reads as the less-settled page in the set.

That's the whole system

InputsOne URL, plus three rivals
EngineSix well-tested ideas
OutputA read, in five sections

One URL in. Five sections out. About two minutes.

Try it · throughline.builders ↗
Throughline · Lens · Calidescope LLC
01 / 30
Jump to any slide · Esc to close

Thirty slides