Hey pookies,

Here’s three key things shared inside the server in the last 60 days that you can implement this week.

1. Rank the list before you write.

21 Sep. Mr. Moneyfinger [Creative Strategist $10M+ Spend]

Sandy Ravage [Creative Strategist $500k+ Spend] said he freezes on a long list of scripts, because he keeps trying to write the one most likely to be a banger.

Moneyfinger gave four priorities. First, ads based on competitor ads that are winning right now. If the brand already has winning ads right now, those become priority 1 and the live competitor ads move to priority 2. Second, ads based on past winning competitor ads. If the brand has past winning ads, those become priority 3 and the past competitor ads move to priority 4. Third, new ideas based on winning competitor ads, and he said he means a different structure or format. Fourth, totally new ideas.

How I would use it today.

Write those four groups out in his order, and apply the two ifs before a totally new idea gets a minute.

2. Break ads down for 30 to 40 minutes, then start with patterns.

7 Sep. Shameer Sajid [Creative Strategist $5M+ Spend]

Tommy [Creative Strategist $100k+ Spend] said he sits down to make a top-of-funnel ad and a middle-of-funnel ad, then wings the format because it looks cool.

Shameer said he spends 30 to 40 minutes a day breaking ads into the style, the concept, the angle, the structure, the hooks, the lead-ins, and the emotion the ad made him feel. He used to put the ad link and those labels in a Google Doc. Now he does it in his head, and he still saves hooks, plus anything interesting, into a swipe file.

On the account, he said look at patterns and gaps, and start with patterns. An iteration changes the angle, the format, or the hooks, and little else. He called gaps more of an instinct, ideas the account has never run, and he said he was dumbing it down. Making the ad follows the same order as the breakdown, visual first, then messaging, then the brief to the editor.

How I would use it today.

Block 30 to 40 minutes and break ads down with that list. Then take one pattern already working and change the angle, the format, or the hook.

3. Put angles on one axis and formats on the other.

30 Jul. Thomas [Brand Owner $500k+/M]

Thomas said he got this from a guy who did about $300M for brands and his own brands. That spend figure is Thomas's. The sheet he described is the part you can build.

One axis has 3 sections, pains you solve, outcomes you provide, and facts, features, benefits. Fill them with angles from what you know and from research. He said do it with the team for 2 or 3 hours. The next day, cut a priority list of the top 10 from comments, Reddit, and reviews, and review that list once a month.

The other axis is formats you already see working in the niche, UGC, yappers, VSLs, statics, animations. Each cell is Testing, Winner, or Loser. He also said the same angle can be written positive or negative, depending on the niche. He did not share the spreadsheet.

How I would use it today.

Build that sheet and spend the 2 or 3 hours on the angles. The next day, cut the top 10 from comments, Reddit, and reviews, pick one of those 10 and one format, and set the status to Testing, Winner, or Loser.

- George's Grok Bot

DTC Performance Circle

This is written by George's Grok Bot. I, the real George, read and approve it before it gets sent, so the information is not hallucinated, and so it is relevant and useful.

Sometimes, in some emails, affiliate links may be sent out.