Hi, Facebook like Google has a desire to serve ads that are relevant and useful to its audience.

Otherwise, people don’t come back as much.

Google rates ads by its quality score which in a simple format is the price per click you have bid times the click-through rates of people seeing the ad.

Facebook recently changed the way it ranks ads relevance to weight more user value ads. Less clickbait sizzle more steak. Must be a machine learning thing as I don’t know how it does this. Know it does consider objective chosen.

Message Quality

But for you trying to get cheap ads I’d say to make your ads “remarkable”.

Make them good enough that people like, comment and share them.

Use both live and pre-recorded video.

Don’t wing it. Script it first and make it as good as your gear will let you. Pay attention to the audio quality and if need be edit in Audacity.

I’ll often do these steps:

  1. Noise removal
  2. Compressor (-15 to -18)
  3. Equlizer (Curve to Bass Boost)
  4. Repeat Step 3 with Treble Boost
  5. Hard Limiter (-3.0)
  6. Normalize
  7. Then “generate > silence” to remove breath noise.

When you are in the ads manager have a close look at the performance and clicks report. Find your relevance score out of 10. Higher it is the cheaper your ads.

Audience Quality

Make your audiences as well defined as you can at scale. By this I mean you might have a small audience of 5000 odd customer email addresses.

Export those from your CRM with the lifetime value from lowest to highest.

Create an LTV lookalike that is large using 1–3% match. Hopefully, you will end up with 2.3 Million odd people Facebook thinks are similar.

Then when you run your ads to this tightly defined audience but an expansive audience, the Facebook machine learning can optimise the serving to people that it is most relevant to.

That means a higher click through , like, comment, share rate and the less you will pay.

Testing Quality

Dennis Yu is the guy to follow, watch, learn about the $1/day for 7 days boost technique. Make a video boost it for $1/day for 7 days. See if it has high engagement by your audience and if so put more budget behind it.

If not engaging then don’t.

Not sure who coined the term starve the donkeys and feed the stallions. Metaphorically OK as that is just horrible otherwise.

But don’t stop there. Test everything. Sometimes images are better than videos especially if going to a landing page. Long copy versus short copy. Different hooks like social proof, scarcity, like-ability etc

Quality Program

Don’t think cheap ads rather an effective campaign. I like to think of a score out 10 for each program.

Does the program deliver a 10/10 ROAS – return on ad spend? (7 day post click, 1 day view attribution).

It’s like a vending machine – you put in a dollar you get $4 change out. That is ok, but at $6 you are laughing and $10 you are speechless. (Yes depends on margin & someone else’s analogy I love.)

Anyway, hope this helps you and if so please upvote, share or comment.



eCommerce Growth Idea Checklist


Leave a Reply

Your email address will not be published.