Creative Strategist & Growth Marketer

I find the signal behind what works and turn it into the next test.

I read the customer, the account, and the creative together—then turn the signal into concepts, sharp briefs, and systems clear enough for a team to ship.

A visual census of video ads reviewed for the BoomBoom case study
72 assets mapped into one testing picture
A lifestyle creator ad analyzed in the EffyDesk account
EffyDesk creator creative

Evidence, not decoration

~80% lower Google Ads spend after I rebuilt the account
$201K commission-qualified booked revenue across 40 accounts I sourced or reactivated
62% higher ROAS found in creator-whitelisted ads versus brand-owned prospecting
~491K CAD in DTC Meta spend analyzed—about $70K a month across seven months

01 Selected work

Five cases. Five usable decisions.

Live-account results and independent work, labeled clearly and built to be inspected.

Case 01 Live-account analysis 2022–2023

Creator-whitelisted ads returned 7.47x ROAS. Brand-owned prospecting returned 4.60x.

Same account, same seven months, $491K CAD in spend—about $70K a month. I ran weekly and monthly performance and creative reporting on EffyDesk’s live Meta account, campaign by campaign. During BFCM, whitelisted ads hit 12.28x, ahead of both retargeting and catalog. When whitelisting spend wound down, blended ROAS declined alongside it.

Performance reportingCreative analysisFunnel comparison

The decision: protect creator whitelisting as a performance lane and test creator identity against brand-owned delivery with offer and format held as constant as possible.

See the account read

Client account. I owned the reporting and findings; the consultancy ran execution.

Back-pain expert creative account creative
Lifestyle UGC account creative
Case 02 Independent case study 2026

72 Meta assets. 54 creative units. 8 angles. One gap nobody had tested.

TikTok winners were short, one-shot identity moments that bought reach. Meta ads leaned on mouth tape as the persuasion carrier. The open test: move the proven TikTok shape across the portfolio without carrying the unsafe claims with it.

Public ad-library pullCreative censusGap mapSpec concept

The next test: move TikTok’s proven short, one-shot identity container across the portfolio while keeping claims compliant. Kill it if hold rate and CTR trail the platform control.

See the gap map

Independent strategy and spec creative built from public ad-library research.

BoomBoom creative census showing grouped public ad-library assets
Frame from a produced BoomBoom spec ad
Second frame from a produced BoomBoom spec ad
platform shape
portfolio test
Case 03 Independent case study + spec work 2026

~460 active video ads. Five angles shared the same ten-beat structure. Nobody made probiotics the villain.

Five angles shared the same ten-beat structure. Nobody made probiotics the villain. That gap became the hook: “Your probiotic may be making the problem worse.” I took it from ad-library census through voice-of-customer research, script, storyboard, and a finished vertical hook.

Competitive researchVOCScriptProduction

The next test: run the probiotic-villain opening against the current hook. Keep it only if three-second hold and outbound CTR beat the control.

See the research + script

Independent strategy and spec creative.

Capsules frame from the gut-health spec ad
Two-room mechanism frame from the gut-health spec ad
Case 04 Historical client work 2023

Spend down ~80%. Paid subscriptions up.

Within roughly one month, the account looked efficient because brand demand was masking a broken acquisition engine. I separated conflicting conversion goals, fixed cannibalizing keywords and weak negatives, and rebuilt the account structure.

Paid searchGrowth diagnosisAccount rebuild

The decision: separate brand protection from acquisition, optimize to paid subscriptions, and stop PMax from making captured demand look like new demand.

See the rebuild logic

Client engagement. I directly rebuilt Google Ads; other channel work was collaborative.

auditseparate intentrebuildmeasure
remembered comparative result ~80% lower spend with paid subscriptions slightly higher
Case 05 B2B demand generation 2022–2025

$201,146 in commission-qualified booked revenue across 40 accounts I sourced or reactivated.

Three years running B2B demand generation across LinkedIn, email, newsletter, research, and direct outreach. After leaving, I turned an insight the role surfaced into two spec ads and a full creator-content operating system: brief, capture, rights, scoring, and test loop.

Demand generationCreator briefEditingTest system

The system: every activation produces usable proof in under ten minutes—then rights, score, and batch it into a repeatable test library.

See the newsletter See the creator brief

Sales closed the opportunities. The creator system and edits are my own spec work.

Spec edit 01
Spec edit 02
the collection loop One activation becomes a library.
01Briefone clear ask
02Captureunder 10 min
03Rightsapproval gate
04Scoreformat signal

proof content · same frame · different city · every weekend

02 What I do

The useful part is what the team makes next.

I connect the analytical answer to the creative and operational next move.

01

Audience & creative research

Voice-of-customer mining, comment analysis, competitor teardowns, angle maps, and market pattern recognition.

VOCanglescreative census
02

Performance diagnosis

Creative and channel analysis that separates the visible metric from the business problem underneath it.

03

Concepts, hooks & briefs

Research-led creative territories translated into scripts, storyboards, references, guardrails, and clear tests.

“What would have to be true for this to work?”
04

Testing & operating systems

Repeatable workflows that define what changes, what stays fixed, how creative gets scored, and what happens next.

05

AI-directed production

I direct image, motion and voice models the way I would direct an editor—storyboard first, animate only what must move, and keep an iteration log of what I kept, what I killed, and why. The spec films in this portfolio were built that way.

KeptStoryboard-first economics
KilledThree takes that read wrong, not broken
See the production log
Meta Ads LibraryGoogle AdsMotion AtriaForeplayActiveCampaign ApolloFigmaCapCutPremiereCanva Claude CodeHiggsfieldKlingElevenLabsPerplexity

03 The working loop

Read. Explain. Brief. Ship. Learn.

Read the customer, account, and creative. Explain why the pattern matters, not just what happened. Brief the hypothesis, hero reference, variable, and kill condition. Ship with a clear owner and next action. Then kill, iterate, or scale.

04 About

Evidence first. Then the idea.

I work the same way on every account: read the customer, the account and the creative together, find the pattern that explains the result, and turn it into a brief someone can build from. The useful output is not the analysis—it is the next test, written clearly enough that a team can ship it without me in the room.

530+ adsreviewed across two category censuses
2 spec filmsresearched, scripted, storyboarded, produced
3 languagesItalian · French · English
2025–NowGoogle Ad Grants Manager · PACE
2022–2025Demand Generation Manager · Scratch Event DJs
2023Growth Marketing Consultant · Venice Music

05 Open to roles

Looking for a strategist who connects performance evidence to what ships next?

I’m seeking creative strategy and growth marketing roles with teams that care about rigorous thinking, strong briefs, and accountable testing.

fulviasgi@gmail.com Click to send an email