The Seedling Method [Facebook Ads]

[Jsoh]

Hey Fastlaners,

Anybody out there in Facebook Ad Land?

I came up with an analogy that might help some people when it comes to running ads on Facebook...

Quick backstory on my experience:

Was directly involved with media buying for Tai Lopez (I worked for him), and so I saw on a massive scale how campaigns are ran that generate massive conversions (whether it's sales, leads, foot traffic, etc.)

And I have a passion for farming (just a hobby), so it led my mind to this method:

THE SEEDLING METHOD

Facebook ads work the same way as when Farmers plant their crops...

#1: The Farmer plants hundreds of seeds
- As far as Facebook ads go, you have to test variations of practically everything. I know that this approach isn't very friendly to the "barrier to entry", but it is the most effective. Planting more seeds (setting up more campaign objectives, ad-sets, and ads gives you the best potential for finding that special winning genetic (best converting ads)

#2: The Farmer monitors those seedlings and note their character traits

• Does it sprout fast?
• Is it resistant to pests?
• Does the upward growth continue or does it stall out?
• Does it yield a lot of fruit?
• Does the fruit taste good?

- In Facebook Ad land... Think of those characteristics of your ad metrics.

• Is the CPC low?
• What's your relevance score?
• Are you getting a positive ROAS (Return on ad spend)
• etc. etc.

#3: The Farmer culls (removes) the plants with undesirable character traits (the ones that don't make the cut)
- There's tons of ways to identify the losers, but I guess the best judgement of whether the ad is winning or not is the amount spent vs amount made. Some people make the mistake of stopping ads that are not very profitable, but even if you're breaking even you are still building your email list, branding your company, etc etc...

#4: The Farmer takes the winners -- duplicates them and scales his garden to max capacity with the best genetics (which gives them the best return on their yields)
- Taking your winning ads and focusing your spend on those is a great way to maximize your growth because the winning ads have the lowest CPA, CPC, and CPL along with highest relevance and other metric factors.

THINGS TO REMEMBER

1. The Farmer ALWAYS sets aside land to test new seedlings.
   - By setting aside some of your advertising budget to testing new ideas, concepts, and variations you give yourself room to stay ahead of the evolutionary curve. Facebook (and any online platform) evolve at record speed.
   
   They are always doing things internally that affects your campaigns. It's your job (or your marketing team's) to find the newest winners before your competition and that only happens with testing new seedlings far after your prior batches are cloned out and have been growing for generations...
   
   
2. Prized genetics can turn to junk as generations go on
   - Since we know Facebook is always evolving, what was profitable in the past could always start to slow it's growth curve or even go into the negative

Good luck!

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[itfactor]

This is method is very similar to what I usually use on my clients' campaigns. I usually start with 2 images, 4 text copies, and at least 20 audience interests for testing cold traffic. This creates 160 ad variations to test ou, then ramp up the budget on the winners.

The key is to be thoroughly systematically in measuring, labeling and recording your data, or you'll be lost by the numerous variations.

Some tips on this method:

- Keep the names of each at set simple and easy to identify. Instead of 'Car dealership campaign 2' (which doesn't say anything', use 'Image A Copy 3 Chevy Fans'.
You can easily see which ad creative was used, which copy variation and which interest you are testing at glance. Makes your life way more easier when you are scrolling through hundreds of ad variations.

- Leave the test ads to run for at least 4 days before you kill or upgrade it, to give FB's algorithm time to do its job

- I use qwaya when I'm setting up multiple adsets. It saves you hours of manual set up using the power manager. If you value your time, you need it for running bulk testing.

 

[Xeon]

The only requirement for the Seedling method is money.....

 

[itfactor]

Well.. technically you can test one adset variation every 4 days on a $10-$15 budget.
But it's very impractical because it takes a long time.

Testing 100 variations take over a year.

The only requirement for the Seedling method is money.....

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Jsoh]

This is method is very similar to what I usually use on my clients' campaigns. I usually start with 2 images, 4 text copies, and at least 20 audience interests for testing cold traffic. This creates 160 ad variations to test ou, then ramp up the budget on the winners.

The key is to be thoroughly systematically in measuring, labeling and recording your data, or you'll be lost by the numerous variations.

Some tips on this method:

- Keep the names of each at set simple and easy to identify. Instead of 'Car dealership campaign 2' (which doesn't say anything', use 'Image A Copy 3 Chevy Fans'.
You can easily see which ad creative was used, which copy variation and which interest you are testing at glance. Makes your life way more easier when you are scrolling through hundreds of ad variations.

- Leave the test ads to run for at least 4 days before you kill or upgrade it, to give FB's algorithm time to do its job

- I use qwaya when I'm setting up multiple adsets. It saves you hours of manual set up using the power manager. If you value your time, you need it for running bulk testing.

Thanks for the input. Have you ever ran your tests for 3 days? If so did you find any difference? I usually test for 3 days.

 

[Jsoh]

The only requirement for the Seedling method is money.....

Yeah I said it's not very friendly for a barrier to entry, but you really need to test a lot of variations to find success in a good time frame

 

[itfactor]

I find that testing anything less than 4 days creates mixed results, especially when the target audience size is above 10,000.

Thanks for the input. Have you ever ran your tests for 3 days? If so did you find any difference? I usually test for 3 days.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.