Policy risk, broken tracking, or a weak offer — most underperforming accounts have one of these problems, not all three at once. I diagnose which one, with evidence, before touching a budget. The same evidence discipline runs through my market research and customer insights work, and the research audits below.
What the work below draws on, end to end — from diagnosing an account to writing the brief that fixes it.
Diagnosing why an ad isn't converting — market desire, awareness, and sophistication, weighed against real auction mechanics and policy risk — before touching budget or creative.
Business Portfolio structure, ad account and partner access, and Meta Pixel / Conversions API tracking with deduplication handled correctly from the start.
Turning customer reviews, comments, and forum threads into evidence-based research, with every claim traced to its source rather than assumed.
The full chain from insight to execution — persona, desire, concept, angle, hook, voice, format — written into a brief specific enough that someone else can produce it correctly.
Reading ROAS, CPA, and LTV:CAC correctly, and diagnosing a performance drop across funnel, attention, spend, and tracking health before recommending a fix.
Planning creative tests against a real budget, duration, and hypothesis, then scaling winners through campaign structure and bidding rather than guesswork.
Independent diagnostics on real, publicly available ad accounts — policy, offer, and pixel, tested end to end.
A four-part Meta ads diagnostic on Loop Earplugs: policy exposure, offer strength, and pixel integrity, tested end to end.
West Shore Home's Meta account, audited against its own landing pages rather than just its ad copy.
CXL Institute's pricing claim tested against its own stated policy and its live discount history.
Walking Noom's actual ad-driven quiz funnel, from goal selection to a gated results screen.
Solargain's Meta account, tested against a real quote-funnel submission and independent review data, not just ad copy.
More case studies in this vertical are coming soon.
Research into customer needs, objections, and buying motivations, with source references and a clear distinction between customer feedback, brand claims, and analysis.
A market research pass across seven source types and ~150 customers, Philippines (home market) vs. Singapore (expansion market).
Everyday-carry wallets tested against their own "built for life" warranty claim and a manual-pull card-access complaint a competitor already markets against.
Guest-side evidence on total price, listing accuracy, cleanliness, cancellation policy, and support — kept separate from host-side experience.
Price, credibility, course content, support, and results — paying-student evidence kept separate from audience opinion and promotional claims throughout.
A review of research sources, customer patterns, and key claims to identify unsupported conclusions, document corrections, and highlight evidence gaps.
Two untraceable citations and one overreaching pattern found and corrected directly in the pack, not just noted for later.
A durability claim narrowed after independently re-tracing it, a source count corrected, and one ad-spend quote flagged unverifiable rather than taken on faith.
Two exercises worked end to end on scenario data, to show the reasoning process — target-setting, root-cause diagnosis, and a defended recommendation — rather than repeat real client numbers.
Two funnel scenarios worked end to end: calculate the breakeven target, then diagnose three campaigns against it with a specific keep, scale, iterate, fix, or pause call for each.
Applying the FAST framework to a real before/after performance drop — ruling out creative fatigue, auction pressure, and tracking issues with data before defending one primary cause.
A structured test plan with the success threshold set before launch, an honest read of which hook actually wins once you look past raw attention, and a specific plan for scaling it without losing the result.
A full strategy-to-script exercise built end to end on an invented brand, to show the process rather than repeat real client research.
I moved into media buying and creative strategy from a decade-plus in recruitment and client operations — sourcing, screening, and managing stakeholder relationships end to end for Fortune 500 accounts and MSP vendors. The habit I carried over is the one this portfolio is built on: find the actual cause before proposing a fix, and back every claim with a traceable source rather than a hunch.
Right now that shows up as independent Meta ads diagnostics, market research, and research audits, all built on public evidence — the work above is the proof.
Have a role, a project, or an account you'd like a second opinion on? Send a message directly — it goes straight to my inbox — or email hello@mariannejaen.com instead.