I’ve spent over $30M of my own money on Facebook ads. That spend is behind over $15M in sales.
Most people who tell you ads don’t work are making the same few mistakes. They test badly, they show the same ad to everyone, and they let Facebook decide where the money goes.
Here is how I structure testing, funnel stages and spend so the money goes where people actually buy.
Stop wasting money on bad testing
The biggest leak I see is testing too many things at once. New ads, new audiences, new budgets, all in the same week. When something works, nobody knows why.
You need a control. Change one variable at a time.
I run testing in three phases.
Phase one: test ads
Keep the audience fixed. Pick one interest, for example frequent travelers, and run 10 to 20 creatives against it.
Because the audience never changes, the winner is obvious. It’s the ad.
Phase two: test audiences
Take your winning ads, say the top five, and put them into several ad sets with different interests. Frequent travelers, entrepreneurship, a Gary Vee style audience.
Now the ads are the constant and the audience is the variable. Whichever ad set wins is telling you who your buyer really is, and that’s information you can use in every future campaign.
Phase three: scale
Combine your best ads with your best ad sets in a scaling campaign using campaign budget optimization. This is where the most money goes.
Why go to this much trouble? In my experience the top 5% of creatives produce about 80% of the profit. Your whole job in testing is to find that 5%.
Set kill rules so losers die fast
Use automated rules to shut off underperforming ads. Examples I’ve used:
- Turn off an ad after $15 spent with fewer than one lead
- Turn off an ad after $30 spent with fewer than two leads
Adjust these for each campaign. My cutoffs have sometimes killed ads too early, so the thresholds need tuning based on your numbers.
If an ad that used to work stops performing, duplicate it and try again.
When my average cost per conversation sat around $14, higher than I wanted, the fix wasn’t new audiences. It was better ads.
Build the funnel in three stages
Not everyone should see the same ad. I split campaigns into top, middle and bottom of funnel.
Top of funnel. Complete strangers. Exclude anyone who has already engaged with your Facebook or Instagram pages, so you’re actually reaching new people.
Middle of funnel. People who have engaged with your social profiles, sales pages or websites. They know you a little.
Bottom of funnel. People who gave you an email or phone number but haven’t booked a call or bought. Exclude anyone who has booked or become a client.
The bottom of funnel job is to push people to book. I use client testimonials here, because these people need proof, not an introduction.
Don’t underestimate this stage. Retargeting often takes about 20% of the spend and produces about 80% of the results. The same pattern shows up everywhere in ads: a small slice of what you do drives most of the return.
In one account my bottom of funnel audience was about 500,000 people. That’s broader than I’d usually recommend, but the smaller audience wasn’t spending efficiently.
For top of funnel I ran two campaigns: one pushing people into my group and one pushing to a VSL. Every stage has one goal, moving the person one step closer to the action you want.
Spend where people buy, not where they click
Facebook likes to spread budget evenly across ad sets. I never let it.
It also tends to push traffic toward whatever gets the highest click through rate. Clicks don’t pay you. Results do. A cheap click from someone who never books is worth nothing.
I want the money going to whatever has the lowest cost per lead or cost per sale. Here’s how I control that:
- Use rules at the ad and ad set level.
- Check the account once or twice a day.
- Turn off inefficient ads by hand.
- If no ads in an ad set are performing, turn off the whole ad set.
In one example nearly every ad was off except two. One was producing results around $11 each, so that’s where I wanted the spend.
I’ve watched a client blow a lot of cash by spending everywhere and never tightening efficiency. Don’t be that account.
Ads always work at the right return
Here’s the mindset shift. Ads always work if you get a good enough return, meaning you make a sale for a limited amount of ad spend.
When someone says ads don’t work, what they usually mean is they’re spending too much per sale. That’s a testing problem, a funnel problem or a budget control problem. All three are fixable.
The takeaway
Test one variable at a time so you find the few creatives that make most of the profit. Split your audience into top, middle and bottom of funnel. Then force your budget toward the lowest cost results every single day.
That’s the system behind over $30M in spend.
Want to know what’s capping your business right now? Take my free 90 Day Ceiling Plan. Answer 18 questions and get a written plan, built on your own numbers, that names the one thing holding you back and what to fix first.
