CV GIANCARLODINARDO.6@GMAIL.COM
Frameworks / Home Services Lead Gen

Before I write a single ad, I build this map

Most accounts have a creative problem disguised as a performance problem. Here's the framework I built to fix that, worked through a roofing lead generation account.

Vertical
Home Improvement · Roofing
Channel
Meta (Facebook & Instagram)
Goal
Lead Generation
Creative Strategy Map · Roofing
01
Product: Residential roof replacement
02
Persona: 3 distinct buyer types
03
Awareness: 5 levels per persona
04
Pain point: 1 primary angle each
15
Creative directions
3
Personas mapped
0
Audience overlap

The problem hiding inside most Meta accounts

At some point Meta advertising quietly changed, and most accounts didn't adjust.

For years the job was targeting. Find the right audience, build the right interest stack, dial in the lookalike. Creative mattered, but the audience was what you optimized. That's shifted. Meta's move to broader delivery means the algorithm now decides who sees your ads based on what those ads actually signal. Run an ad about storm damage and Meta finds people worried about storm damage. Run one about home value and it surfaces a completely different person. Same campaign settings, different creative, different audience.

So the creative decisions you make before launch are, in practice, audience decisions. Most accounts haven't caught up to that yet.

Here's where it falls apart. When you run similar ads (same angle, same emotional hook, different headline), Meta groups them under the same entity ID. They share delivery, burn through the same audience segment together, and research on ad fatigue points to a roughly 45% drop in conversion likelihood after four exposures to the same ad. Tweaking a headline doesn't reset that. A different concept does.

Every account I've audited that can't scale past a certain point has the same thing going on: lots of ads, very few distinct signals. The framework below is how I solve that before a single ad goes into production.

Four reasons your creative pool is smaller than you think
01 · Creative is now targeting
The ad itself signals who sees it. Same angle = same audience reached, regardless of your targeting settings.
02 · Andromeda collapses duplicates
Meta groups similar ads under the same entity ID. They share delivery and tap out the same segment together.
03 · Fatigue compounds fast
Research on ad fatigue points to a roughly 45% drop in conversion likelihood after four exposures to the same ad. Variations of the same concept share that counter.
04 · The algorithm needs distinct signals
Meta optimizes across genuinely different creative. Without distinct concepts, there's nothing new for the system to learn from.

Four layers.
Infinite directions.

I call it the creative strategy map. Four layers, worked through in order: Product, Persona, Awareness Level, Pain Point. The output is a matrix of creative directions that don't overlap. Not a list to pick from. A structure that makes running out of angles impossible.

01
Product
Pick your strongest product or service. Build everything around that one thing first.
02
Persona
Identify 3–4 buyer types defined by why they buy, not demographics. The specific moment that pushes them toward a decision.
03
Awareness level
For each persona, map what they're thinking at every stage, from unaware through most aware, in the first person.
04
Pain point
Find the primary angle per persona based on review frequency and competitive signal. One angle each. That's your creative brief.

A persona isn't a demographic

Age, income, homeowner status: that's not a persona. A persona is the specific thing pushing someone toward a decision. The moment that makes them search. Two people with identical demographics can be completely different buyers if the reason they're looking is different. For this roofing account, I mapped three personas that consistently drove volume.

🔒
The anxious procrastinator
The roof has issues and they know it. Cost, disruption, not knowing which contractor to trust. There's always a reason to wait. By the time they call, they've been sitting on this for months.
"The shingles look rough but it's probably fine… right?"
Primary angle: Fear of interior damage
The crisis responder
Storm came through, water is coming in, and the call needs to happen today. There's no research phase, no comparison shopping. They need someone fast and they need to trust them immediately.
"I need someone out here today, not next week."
Primary angle: Urgency + trust gap
📈
The value-conscious planner
Listing in 12 to 18 months. Not in crisis, but knows a flagged roof either kills offers or forces price concessions. This is a financial decision, not an emotional one.
"Will a new roof actually increase what we get for the house?"
Primary angle: Investment justification

What they're thinking at every stage

For each persona, I map first-person thoughts across all five awareness levels, from completely unaware through ready to convert. Each cell is a distinct creative direction, each triggering a different entity ID on Meta.

The anxious procrastinator The crisis responder The value-conscious planner
Unaware "The shingles look a little rough but it's probably fine." "That was a bad storm last night." "We've been thinking about listing the house in the spring."
Problem aware "There's a stain on the ceiling. I keep ignoring it." "There's water coming in. This is bad." "The inspector flagged the roof. Buyers will notice."
Solution aware "I know I need a roofer. I just keep putting it off." "I need someone out here today, not next week." "A new roof could help us get asking price."
Product aware "I got one quote but haven't called them back." "I've seen their truck around. Are they legit?" "I've seen 3 quotes. Who gives me the best value?"
Most aware "Free inspection, no pressure? I'll do that." "Same-day response, licensed. I'm calling." "Financing available, warranty transferable. Easy decision."
3 personas
×
5 awareness levels
×
1 angle each
15
distinct creative directions, before a single format is chosen

Why this produces real diversity

Three personas, five awareness levels, one primary angle each: that's 15 creative directions before format is even a consideration. Layer in format variation: UGC video, static, carousel, testimonial. You have more combinations than any campaign exhausts in a quarter.

What matters is that each combination is structurally different. Not a different headline on the same concept. A different emotional entry point, a different assumed starting point, a different person. Meta sees them that way too: separate entity IDs, separate delivery, separate audience segments.

In practice I focus on two personas and their strongest angles rather than testing all 15 at once. Around 12 concepts with three hook variations each gives a serious testing pool without creating a production problem.

The process, step by step

Building this map manually used to take the better part of a week: pulling reviews, going through competitor creative, mapping audience psychology by hand. I now run most of the research phase through an AI workflow that cuts that down dramatically. The judgment calls are still mine. The legwork doesn't have to be.

Step 01
Product selection
Identify the strongest product or service (usually by conversion volume or margin). Build the entire map around that one offer. Everything else supports it in the funnel.
Step 02
Persona extraction from review data
Pull customer reviews from Google, Yelp, or third-party sites. Feed them into an AI workflow to cluster recurring motivations. Look for distinct buying triggers, not demographics.
Step 03
Awareness-level thought mapping
For each persona, generate first-person thoughts at every awareness stage. This is the creative brief. An ad for an unaware procrastinator and an ad for a most-aware procrastinator are completely different pieces of creative.
Step 04
Pain point prioritization
Identify the primary angle per persona based on review frequency and competitive signal. Which pain points come up most? Which angles are competitors testing? One angle per persona. That's the creative focus.
See this framework applied to a live roofing account at scale
The roofing case study shows what active account management looks like once this map has been built: deconstructing one winning video into nine variants, giving five markets their own campaign architecture, and building the creative system to absorb a 5x spend increase. Read it →
Before this framework
The same ad with a different headline, running to the same audience, being called a test
Meta grouping similar creatives together, sharing delivery, burning through the same segment
CPMs climbing as the pool dried up with nothing new to feed the algorithm
Creative decisions made reactively: what performed last week, iterated, repeated until it stopped working
No visibility into which audience segments were never reached at all
With the creative strategy map
15+ creative directions mapped before anything goes into production
Each one gets its own entity ID: separate delivery, separate segment, no overlap
Clear picture of which personas and awareness stages are uncovered going into launch
The creative brief exists as a document you can hand to a team, not an instinct that lives in your head
Research that used to take days now takes hours, without losing the strategic layer

Want this built for
your account?

I apply this framework to every account I work with, lead gen or eCommerce.
If your creative is running but not really testing, let's talk.

Get in touch →