Why every ad worked. Or didn’t.
Every ad gets tagged across six strategic variables: persona, angle, hook, moment, framework, and format. Some live in the filename itself. The rest live in the asset’s metadata. Across tests, the system surfaces patterns: which combinations scale, which stall, and what to change next. A seven-stage methodology that turns paid social creative from intuition into intelligence.
Persona
The psychographic buyer, not the demo bucket.
Angle
Entry point into the persona’s desire.
Hook
The first three seconds.
Moment
The trigger experience brought to life on screen.
Framework
The persuasion structure of the ad.
Format
UGC, static, video, carousel, or creator brief.
· Sample tagged asset name
· And in the asset’s metadata, alongside the filename
- · Angle (entry into the persona’s desire)
- · Hook (the first three seconds)
- · Creator profile (who delivered the message)
- · Landing page (where the click landed)
Every variable tagged. Every result explainable.
Nothing here is improvised.
Stages 01 to 03 are the foundation: brand intelligence, audience segments, persona architecture. Set once, refreshed quarterly.
Stages 04 to 07 are the operating loop: build, produce, analyze, scale. Continuous, every campaign cycle. Loop learnings feed back into the foundation.
Brand Intelligence
Comprehensive market and audience research.
Audience Segment Identification
Psychographic segments from the research, not demographics.
Persona Architecture
Core desire, trigger experiences, language, objections.
Angle, Moment, Framework, Format
Systematic creative construction across the strategic variables, with hook stacks layered on top.
Production Pipeline
Strategist to creator to performance editor to account.
Performance Analysis and Strategic Iteration
The tagged-data layer surfaces patterns. Strategists set the hypothesis.
Scale, Whitelist, and Partnership Ads
Amplify proven concepts through authentic creator voices.
Before a single ad is written.
Augusta does not write a single ad concept until the research is complete. Creative built without research is guesswork, and guesswork is expensive at scale.
Brand Intelligence
A multi-source analysis of the brand, its buyers, its competitors, and the broader market, not just customer reviews. Six streams feed it: reviews, Reddit and forums, competitor analysis, adjacent market research, market sentiment, and brand-side intelligence.
Adjacent market research is one of Augusta’s key differentiators. We study brands solving a different problem for the same person, or solving the same problem through a different category. A framework that works for a premium coffee brand talking to ambitious professionals is likely to work for a focus supplement talking to the same person.
· Output: Brand Intelligence Brief, the source document for every stage that follows
Audience Segment Identification
Distinct buying audiences emerge from the research. These are not demographic buckets. “Women 25-34 interested in health and wellness” is a targeting parameter, not an audience insight. Augusta’s segments are psychographic clusters defined by a buyer’s specific relationship to the problem the product solves: their context, emotional state, trigger, and desired outcome.
Segments are discovered, not invented. We look for recurring patterns: the same language appearing in multiple reviews, the same frustration surfacing in multiple Reddit threads, the same trigger moment described by buyers who are otherwise very different people. Those patterns become segments.
· Output: Documented set of named segments, each defined by relationship to the problem
Persona Architecture
Each segment becomes a fully developed persona. Not a marketing exercise, not a user-story template. A production-ready creative brief that tells every member of the team exactly who they are talking to, what that person cares about, what they are afraid of, and what they sound like.
Each persona brief covers seven things: core desire (the life change, not the feature), success (specific and visceral), trigger moments (when the product matters), voice (verbatim from research), objections (what they don’t believe), visual world (what feels native), and hook angles (ways into the desire).
· Output: Core Avatar and Sub-Avatar documents, one per persona
Where research becomes creative.
Every ad is a structured combination of the strategic variables. Each one can be isolated, measured, and iterated on independently. This is what transforms ad testing from “let’s try a bunch of things and see what works” into a scientific process.
Angle, Moment, Framework, Format
Every ad is the execution of an underlying concept: the central creative thesis the variables test. The angle is the situational entry point into the persona’s core desire. The moment is the trigger experience brought to life on screen, Augusta’s primary scroll-stop mechanism. The framework is the persuasion structure of the ad. The format is the delivery vehicle: UGC video, mid-production, high-production, static, carousel, or creator brief. Each angle gets a stack of hooks, the first three seconds that decide whether the ad earns the rest of the watch.
Augusta maintains a documented framework library across video, static, and creator briefs. Combined with the variables above, the system produces thousands of testable combinations per persona. Frameworks are not templates. Each one has a different emotional trajectory and works better for different audience states and product categories.
· Sample frameworks from the library
Problem-Agitate-Solution
Identify the problem, intensify the pain, present the product as the resolution.
Us vs Them
Position the product as the superior alternative to what the viewer is currently using.
Founder Story
Personal narrative about why the product was created. Builds origin credibility.
Before and After
Show the transformation in concrete, specific terms from the persona's point of view.
Testimonial Stack
Multiple real customer voices validating the same core benefit.
De-influencing
Acknowledge what does not work before introducing what does. Builds credibility through honesty.
Problem-Solution-Proof
Identify the problem, present the solution, then validate with evidence.
Curiosity Hook to Reveal
Open with an unexpected or counterintuitive statement that demands resolution.
Production Pipeline
Four stages: creative strategist writes the script or designer brief, creator films or designer executes, performance editor cuts and refines specifically for paid social, then account entry with the naming convention applied.
The naming convention is non-negotiable. Every ad enters the account with two layers of tags: the filename carries brand, persona, moment, framework, format, and version. The asset metadata carries angle, hook, creator profile, and landing page. Without structured tagging, performance data is a flat list of spend and ROAS numbers. With it, the account becomes a multi-dimensional data set we can interrogate by any variable combination.
· Output: Tagged production scripts and designer briefs that make Stage 06 possible
We don’t report. We decide.
A report tells you what happened and leaves the next move to you. Augusta reads the account, finds what’s driving performance, and turns it into the next round of creative: what to build more of, what to refine, what to retire, what to test.
Performance Analysis and Strategic Iteration
Because every ad was tagged going into the account, we can see exactly which combinations are pulling their weight and which aren’t. The data surfaces the patterns; a strategist reads them and writes the plan for the next round. Nothing moves on a hunch.
Augusta does not automate strategy. The thinking stays human. The data can tell us the Mom persona is winning at half the cost per acquisition of the Guilt-Free persona. It can’t tell us why: maybe the moment lands harder this season, maybe a different creator would convert better, maybe the landing page is doing the work. Sorting that out, and deciding what to do next, takes a strategist who knows the account, the brand, and the market.
We’re strict about what counts as a win. We separate new customers from blended ROAS, so the concepts we build on are the ones bringing in buyers who’ve never purchased before, not the ones just re-capturing people who’d have bought anyway. Retargeting has its place. It’s just not what we shape creative around.
· Four iteration actions per concept
Scale
A winner scales on the media side, that’s not our lever. Ours is creative: we read why it won (the hook that stopped the scroll, where viewers dropped off, whether a high CTR means the offer landed) and build the next batch on those learnings.
Iterate
A concept with signal but not quite at target gets one specific, hypothesis-driven change, not a random reshuffle. Hook worked but the body lost them? We rebuild the body and run it as a clean test.
Cut
A concept that has had enough runway and still isn’t converting gets retired. We stop making that direction and put the effort elsewhere. A data call, not a gut one.
Test
Fresh concepts go into testing continuously, new angles, hooks, and personas, so there’s always something new in play to find the next winner.
· The balance that prevents plateau
The discipline isn’t doing any single action. It’s running all of them simultaneously, every cycle. Most agencies scale a winner, iterate variants until fatigue, and quietly stop testing once they have something working. The account looks stable for a quarter, then plateaus when the winner burns out with nothing tested to replace it.
Augusta runs scale, iterate, and test in parallel. Around 30% of testing capacity stays in net-new concepts at all times. Net-new means truly new angles, personas, or framework combinations, not variants of the current winner. That’s what prevents creative fatigue, expands TAM into new audience segments, and keeps winners compounding instead of running out.
Scale, Whitelist, and Partnership Ads
Once a persona-angle combination is proven in paid, with consistent CPA performance and a validated creative concept, Augusta identifies creators who authentically represent that persona and brings them in as whitelisted or partnership ads. The creative direction is already validated. The risk is already reduced.
Whitelisted content runs as a paid ad through the creator’s handle rather than the brand’s page. A viewer who sees an ad coming from a person rather than a brand is in a fundamentally different psychological state: lower skepticism, higher receptivity, stronger social proof effect.
Most agencies test ads. We test hypotheses.
A sample of the questions we work through when we read an account, never the full list, just the kind of thinking the tagging makes possible. Without it, performance is a flat list of spend and ROAS. With it, the account is a database we can interrogate, and every answer becomes a hypothesis for the next round of creative.
Is the Mom persona genuinely outperforming the Guilt-Free persona, or is the gap just where the spend happened to land?
Is the Problem-Agitate-Solution framework outperforming Us vs Them for the 3PM-Slump angle, or is the difference in the creator profile?
Is the static format generating a lower CPA than UGC video for the Older persona, or is the video being outperformed because of a weak hook rather than the format itself?
Which creator profile produces the lowest CPA on the Mom angle, and is that difference consistent across multiple ad concepts or specific to one creative?
Is thumbstop improving while CPA is flat? That would suggest a hook quality improvement that hasn't yet translated to conversion. Is the landing page the likely gap?
· A sample. The real list runs dozens deep, every cycle.
· Filename: brand, persona, moment, framework, format, version · Metadata: angle, hook, creator profile, landing page
The accumulated intelligence is the head start.
The Augusta Creative Production System is designed to compound. The output of each engagement period is not just better ads. It is better intelligence. A brand that has been working with Augusta for six months has a richer Brand Intelligence Brief, more refined personas, a larger database of validated and invalidated creative hypotheses, a proven roster of creator profiles matched to specific angles, and a performance record that tells us exactly which variables to change to move CPA in the desired direction.
This accumulated intelligence is the head start. An agency starting from scratch on a new account has no prior hypotheses, no validated angles, no creator matching data, and no performance history to draw on. Augusta’s system is designed to build that intelligence systematically and retain it in structured form, in the Brand Intelligence Brief, in the persona documents, in the performance database, and in the strategic hypothesis record, so that it compounds with every iteration rather than resetting with every new campaign.
“Most agencies test ads. Augusta tests hypotheses, documents the results, and builds on what it learns. That is the difference.”
See the system in your account.
15-minute call with Matthew. We’ll talk through your current creative and give you a read on where the leverage is. Whether you work with us or not.
Let’s chat