Hey pookies,

Here are three things from the server, from the last 60 days. You can use all three this week.

1. Keep the part that worked. Change the rest.

18 Sep. George Rose

I wrote up a short overview of the creative strategy job. Two points from it are the ones to use this week.

If an ad is working, find the part that did it

Don't throw the whole ad out. Ask which part made it work. Keep that part. Change the rest.

  • The type of ad. A video or an image.

  • The idea. The main thing the ad is saying.

  • The person on camera.

  • The mix of those. Sometimes it only works because they are together.

Look for what a competitor is running that you are not

That missing thing might be a type of ad, a look, or a kind of customer. Sometimes that customer only shows up in the comments or the reviews.

Put the next test on an idea that already works

What you test first matters more than how many tests you run. Take an idea that is already working and change one of these:

  • What the ad is pretending to be

  • The type of ad

  • The situation the buyer is in

  • A smaller result they want, inside the big one

  • How the ad looks

These are some of the strongest moves you can make, and they usually win. Most people drop an idea before buyers are done with it. The simple bet is usually the right one.

❝

Try it this week

Pick one ad that is working.

  • Write down the part you think made it work.

  • Keep that part.

  • Change one other thing.

2. Copy a winner before you make something new.

2 Sep. George Rose

I posted the useful parts of a guide by Kamal Razzak. It is called Lessons From Training the Best Creative Strategists in the World.

Remake a winning ad before you invent one

New people at his agency have to remake the best ads before they are allowed to make something new. They remake those ads for:

  • Different brands

  • Buyers who already know the problem, and buyers who do not

  • Different kinds of customers

Changing a winner is harder than copying one. If you have never had a win with an idea, you cannot tell yet if your plan for that idea is any good.

Finish the ad within 3 days

If you wait 2 weeks to see the results, you forget the research and why you made the ad. He keeps the path from the plan to the finished ad inside 3 days. He got there by hiring more editors.

Read the results yourself

He does not want you to hand the ad off and come back later. Open the account and read these yourself:

  • Hook rate. How many people stay past the opening.

  • Hold rate. How many people keep watching.

  • CTR. How many people click.

  • CPA. Cost per result.

❝

Try it this week

Before you write a brand new idea:

  • Remake one ad you already know works.

  • Finish it within 3 days of writing the plan.

  • Read those four results yourself.

3. Decide when to turn an ad off.

24 Aug. George Rose

I posted notes from our 24 Aug strategy call. The notes do not say who these people are.

When you turn an ad off should match how many you make

There is not one right number for every account. The number has to fit how much you make.

  • One person makes 20 to 25 new ideas a week. He turns an ad off at 2x CPA.

  • On weeks he wants better work, he only makes 10 to 15 ideas. Those weeks, he lets each ad spend more before he turns it off.

  • Another person waits until spend hits 3x or 4x AOV. AOV is average order value.

Some losing ads turn around

One person had ads that lost money for 5 or 6 days, then became ads that spent a lot. Remember that before you lock a turn-off rule you cannot walk back.

❝

Try it this week

Write this down today:

  • How many new ideas you are making this week.

  • The rule you use right now for turning an ad off.

  • If you are at 20 to 25 ideas, his rule is 2x CPA.

  • If you drop to 10 to 15 so the work can be better, let each ad spend more.

  • Leave the rule open. One person watched ads lose for 5 or 6 days, then spend a lot.

- 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. I check that the facts are not made up and that the email is useful and relevant.

Some emails may have affiliate links. These are links that can earn a fee for the person who shares them.