THE $1M AD FRAMEWORK

What kind of agency is behind the testing patterns in this episode?

TNT operates as a performance creative agency inside a direct response ad agency model — Kira Hug and Christy Ingkavet both hold the title of creative strategist and associate copy chief, and Christy specifically describes the company as “TNT Performance Agency” rather than a traditional paid advertising agency that only handles media buying. The agency’s client base is direct to consumer brands, and the testing patterns in this episode came from reviewing 16 of those clients’ ads that each crossed $1M in revenue.

How much should I spend before judging if an ad is worth scaling?
$1,000–$2,000 in spend over the first 1–2 weeks is the minimum threshold for a reliable read, according to the data behind this episode. Below that, there isn’t enough data to separate a real signal from noise — and an ROAS at or above a client’s target at that spend level is the clearest early sign an ad is worth scaling further.

Is this the video sales letter funnel format, and does the format matter?
Both case studies in this episode run as long-form video ads that build a case for the product before the sale — the structure most marketers would recognize as a video sales letter funnel, even though TNT doesn’t use that specific label internally. What actually mattered more than the format itself was proof: four formats (Hidden Anatomy, Myth-Busting Expert, Emotional Stories, Podcasts) have collectively driven over $30M in spend across TNT’s client accounts, including likely supplement offers given the mechanism-driven, DTC nature of both case studies. The takeaway from the episode is to use a format with a track record rather than inventing a new one.

How does YouTube compliance testing work for these DR YouTube ads?
YouTube’s compliance tolerance shifts month to month, so a video sales letter YouTube ad approved a year ago can get flagged later with zero changes made to it — which is what happened to the “Artery Cement” case study. TNT’s fix for these VSL YouTube ads is to strip out specific timeframe claims (turning something like “7-second artery flush” into “at-home artery flush”) and launch 2–3 ad body variants simultaneously, each with a progressively heavier compliance pass, rather than guessing which single version will clear.

What does DR funnel optimization look like once an ad starts scaling?
The episode lays out a specific order: pick a proven format first, put significant budget behind clickbait testing, test hooks second (expecting smaller gains per test), and don’t stop testing once an ad is already winning. Once an ad clears its target ROAS, the recommendation is to recycle proven elements from past ads — hooks, clickbaits, toppers, talent — rather than reinventing everything from scratch, and to double down on the one or two elements actually driving the result.

Does spokesperson testing work differently on YouTube versus direct to consumer Meta ads?
Yes — in the “Discredit Fiber” case study, seven different spokesperson personas were tested on the same offer (a gut/digestive health, likely supplement, product), and certain personas performed better on YouTube while entirely different ones won on Meta. A custom, in-house-built avatar also outperformed generic AI avatars once those became overused across other advertisers’ ads — the same logic applies to voice talent, since one popular AI voice (“Adam”) stopped converting once it became too recognizable from overuse.

Why does a scaling ad plateau, or stop growing after it was already performing well?
Two patterns from the episode explain most of this. First, “hook starvation”: the bigger an ad’s spend gets, the smaller the share of revenue that tends to come from hook testing, unless a fixed percentage of the testing budget (TNT targets around 15%) is protected for it. Second, the “proven elements gap”: nearly half of the $1M+ ads reviewed never reused a previously-proven clickbait, hook, or post-text from the account — leaving an estimated $20K–$200K in additional spend unclaimed per ad, usually due to bias toward testing what a writer believes will win rather than what’s already proven.

How do I know which tested element is actually driving my ad’s revenue?
Look for the 80/20 pattern described in the episode: across the 16 ads studied, just two tested elements typically drove 50–80% of total revenue. In “Discredit Fiber,” for example, $3M of the ad’s $5M in total revenue came from clickbait testing alone. Once you identify your ad’s needle-mover, the move is to double down on it while keeping secondary tests running.