What variables do YOU test for on Meta Ads?

Advertise on BHW
Duration: Run your tests for a sufficient period (usually at least a week) to gather enough data and account for day-to-day fluctuations.
 
Okay I appreciate they need a weeks duration but what variable would you test within that week please :)
 
When running ads without a strong custom audience, first A/B test different ad creatives (images, videos, and copy) to identify what best captures attention and drives engagement.
 
JDADAWGFG said:
When running ads for a company without a large/accurate custom audience what variable would you firstly test via A/B testing or two concurrent campaigns?


All responses appreciated!
Click to expand...
I always start by testing creatives (video vs. static image), because that's what has the biggest impact on CTR right away. After that, I test the hook in the first 3 seconds — it makes a huge difference even with the same audience. I leave audience for last when I don't have a very defined base.
 
JDADAWGFG said:
When running ads for a company without a large/accurate custom audience what variable would you firstly test via A/B testing or two concurrent campaigns?


All responses appreciated!
Click to expand...
You will need to run the test for a long enough period of time, usually at least a week.
 
To have enough data and notice a difference between days, try testing for at least a week.
 
Prioritize A/B testing (A/B Testing in Ads Manager) because:
Ensures objective results (Facebook distributes evenly for each variation).
Easy to measure effectiveness according to specific indicators such as CTR, CPC, ROAS.
Concurrent campaigns should only be used when:
You need to test quickly and have a large budget.
Absolute accuracy is not too important but want to feel the market response first.
Strategy suggestion: Start with A/B testing the target audience, then move to creative when you find out who is responding well. After a few rounds of testing, you can create a Custom Audience from the people who have interacted, then make a Lookalike Audience from there.
JDADAWGFG said:
When running ads for a company without a large/accurate custom audience what variable would you firstly test via A/B testing or two concurrent campaigns?


All responses appreciated!
Click to expand...
 
JDADAWGFG said:
When running ads for a company without a large/accurate custom audience what variable would you firstly test via A/B testing or two concurrent campaigns?


All responses appreciated!
Click to expand...

Audience targeting, creatives, landing page experience, interests, behaviors, and lookalikes, I always test for load speed and user flow.
 
I will test the wide file then narrow it down by breaking down the age by report
 
Start by testing different ad creatives (images, videos, headlines) and audience targeting (demographics, interests, behaviors). Focus on testing variations of call-to-action and landing pages as well. Keep testing budget allocation and bidding strategies for better optimization.
 
Start by testing ad creatives images, copy and placements. These have the biggest impact on performance.
 
EnEnTi said:
I will test the wide file then narrow it down by breaking down the age by report
Click to expand...
What do you mean by wide file sorry :)?

mori ekeir said:
Start by testing ad creatives images, copy and placements. These have the biggest impact on performance.
Click to expand...
Thank you that's really helpful, I'm surprised Copy makes a substantial difference - just to check by copy you mean whether the button says 'buy now' or 'find out more' for example?
 
I am actually curious as well, for running time. What's the minimum and maximum amount of Ad testing?
 
i play with different hooks, visuals, and offers first. targeting stays broad till I see what actually gets traction.
 
first of all, set your pixel. What is your end goal, people should buy your product? call you? Then, create custom audience , different age groups, different locations etc. Create engaging ads. Spend $15 per ad and check which is making you money. Delete those which are not working and scale those which are working.
 
Run the test for a long enough period of time . it makes a big difference even with the same subject.
 
Should use a/b test will bring more accurate and accurate results
 
Run both campaigns at the same time, keep the one with better results and remove the one with worse results.
 
Advertise on BHW
You must log in or register to reply here.
Sign up now!

Main Menu

Home Main Forum List Black Hat SEO White Hat SEO BHW Newbie Guide Blogging Black Hat Tools Social Media Downloads

Marketplace

Account Selling Content / Copywriting EBooks / Methods Hosting Misc Proxies For Sale SEO - Link building SEO - Packages Social Media Social Media - Panels Web Design More

Making Money

Affiliate Programs Hire a Freelancer IM Journeys Making Money Pay Per Click Site Flipping Want to Buy

BlackHat World

Dispute Resolution Forum Suggestions & Feedback Introductions Lounge
Back
Top