DTC Daily DTC DAILY FUNNELS
Atlas Coffee Club
← Breakdowns Funnel breakdown · Atlas Coffee Club · captured 2026-09-21 draft · awaiting review

Atlas Coffee Club runs nine ads under one line about quitting big-name grocery coffee, then lands the reader on a first-person article that blames machine harvesting and over-roasting for the bad cup and closes on a free first bag.

The Article That Sells9 ads · 0 livelongest running 79 doffer: For a limited time - Get Your First Bag Free open in library ↗

1 · The argument, in one line

Top ad opens
“I was buying coffee from the big name grocery stores until I learned this…”
→
Page headline
“Before You Buy Grocery Store Coffee, Read This”
→
Offer
“For a limited time - Get Your First Bag Free”

Message match: Continuous. Every ad's primary text names big name grocery stores and the page opens on the same buyer and the same two villains, machine harvesting and over-roasting, though two ads promise 50% off where the page gives the first bag free.

2 · The ad set

creatives9 ads, 8 distinct creatives: rank 9 is the same image as rank 5 (Atlas vs Store Bought checklist)
formatsUGC talking head 4, Us vs them 3, Lifestyle 1, UGC static 1 (5 videos, 2 statics, 2 carousels)
hooksCuriosity 5, Question 2, Contrarian 1, Social proof 1
angles5 distinct: Unique mechanism 4, Convenience 2, Simplicity 1, Novelty 1, Desired outcome 1
proofStatistics 1, Testimonials 1; 7 ads carry no labelled proof
Diversity read

A portfolio on one persona, the grocery-store coffee buyer, with all nine ads under the same primary text and a "Big Coffee" headline. The spoken openings are what vary: "Coffee is a fruit. It's literally called a cherry.", "Ranking coffees by where you buy them.", "Why does grocery store coffee always need so much creamer?" and "Europeans would laugh at America's morning coffee because big coffee cuts every corner possible." Four of the five videos then run the same harvest-and-roast mechanism, so the variety sits in the hook more than in the argument.

3 · The top ad, beat by beat

0:00Surprising open“Which camera? This one? Coffee is a fruit. It's literally called a cherry. It's supposed to be”
0:07Problem“harvested ripe like literally anything else you'd eat. But mass market brands, don't do that,”
0:11Problem“they harvest everything at once. Ripe, unripe, doesn't matter, and then they over roast it to”
0:16Agitation“cover up how bad it actually tastes. And they call it bold. I call it burnt and lying to me.”
0:21Product introduction“So now I drink Atlas Coffee Club. They work directly with farmers. They only take coffee”
0:25Mechanism“that's in season and they roast it after you order, not months before, after. I think you can”
0:30Demonstration“actually taste where it comes from. Ethiopia tastes different from Costa Rica. Revolutionary, I know.”
0:34Call to action“You want to try it?”
UGC talking head · Curiosity hook · Unique mechanism ("they roast it after you order, not months before, after.") · 0:36

4 · The page, in order

01 Hero: first bag free, headline, the morning routine 02 Taste, how it works, pause or cancel, first bag free, As Seen On

The offer sits above the headline and again at the close, with the whole case in between: the routine, then The Truth About Big Coffee (machine harvesting, over-roasting), then Atlas and the top 1% of farms. The middle repeats the top ad's pattern almost beat for beat (unripe harvest, over-roast, called bold, really burnt), and the only proof is an As Seen On row after the last offer.

5 · Why it works

  • The ad's opening line becomes the page's first sentence.. All nine ads say the reader used to buy from big name grocery stores; the headline and first line of the page speak to that same buyer and promise the facts the ad teased. The click feels like reading the rest of the post. · “I was buying coffee from the big name grocery stores until I learned this…” · “Before You Buy Grocery Store Coffee, Read This” · “If you buy coffee from a big name grocery store, then you may not be aware of some pretty scary facts..”
  • One mechanism, told twice in the same order.. The top ad and the article both blame the bad cup on a harvest that picks everything at once, then on roasting to hide it. A reader who believed the video meets the same two steps in longer form. · “they harvest everything at once. Ripe, unripe, doesn't matter, and then they over roast it to” · “So they use what’s called “machine harvesting” to pick their coffee.” · “After harvest, and in an effort to mask all of the shortcuts taken, most coffee chains over-roast their beans.”
  • It flips the label on the bag against the brand.. The ads and the page both take the word big coffee uses to sell dark roast and rename it burnt, which turns the reader's own grocery bag into evidence. · “And they call it bold. I call it burnt and lying to me.” · “They call it bold or full bodied.” · “Big coffee companies will try to spin this in marketing as BOLD or FULL-BODIED, but in reality it’s just really really burnt”
  • The pain is the milk and sugar the reader already adds.. The page's routine ends in a wince and a sugar fix, and two ads name the same fix as proof the coffee is bad. That makes the problem something the reader did this morning. · “Drink. Wince. Add sugar and milk.” · “Then they train you to fix it with creamers, syrup, sugars, anything to cover it up.” · “not burnt diner coffee that needs three sugars to be drinkable.”

6 · What to take

  • Write the page's first sentence as a reply to your ad's primary text, using the same noun for the buyer (here: big name grocery store).
  • Name two concrete steps the big brands take (harvest everything at once, over-roast) and tell them in the same order in the ad and in the article.
  • Take the word your competitor prints on the bag (bold, full-bodied) and give the reader a blunter name for it that they can check in their own cup.

7 · Watch-outs

  • The offer changes at the handoff: ranks 2 and 4 say 50% off the first order or right now, while the page offers the first bag free.
  • The sourcing number changes too: rank 4 says the top 5% of farms, the page says the top 1% of farms.
FOR THE REVIEWER
  • The rank-1 ad ran 11 days, while ranks 3 to 9 ran 79 days; rank is taken from the input as given.
  • ad_opens uses rank 1's primary text, since its spoken open starts with off-camera chatter ("Which camera? This one?").
  • Page sections cover only the opening (cut at "The Truth Ab") and the close; the machine-harvest and over-roast argument and the top 1% claim are in page_text only, so the page strip has two moves and should be checked against the screenshot.
  • The CTA button text and a five-star row are in page_text or the screenshot but not tagged in any section; no reviews were labelled.
  • Page type is The Article That Sells with no byline or masthead; The Reason-It-Works Page was the labeller's runner-up.
  • The pull date is not in the input; only the page capture date is. On-image text on ranks 5, 7, 8 and 9 is not in the corpus and is not quoted here.
Ad pull: 9 ads (all 9 in input, none live, longest 79 days); page captured 2026-09-21; transcripts for all 5 videos; labels verified 139 of 145 quotes (the 6 misses are on-image text from statics and carousels). Nothing here is estimated. Every quote is copied from the ad or the page. Meta publishes no spend or results; nothing here is a performance claim.