fbpx

Don’t Trust Your Facebook Rep, Always Trust the Testing Data

DimNiko Marketing | Always Trust Your Testing Data

Facebook Reps are beneficial to the success of our client’s accounts. 

They warn us on specific changes, latest policy updates, and the structure of our campaigns so they can perform better. 

Sometimes, they even provide the ‘perfect’ testing budgets or how long the campaigns should run for (longer the better). 

Read More: Best Practices When Ads Manager Is Recording The Wrong Data.

Throughout my career as a Facebook Media Buyer, I’ve heard this tip more than 20x from different reps: You need to run your campaigns for 7 days and not touch it! 

For me, this never made sense since all the money that should be profitable was lost in the first 3-4 days if I didn’t touch it. 

Even if the campaign ‘magically’ would turn on the good side, I still wouldn’t be able to break even because of the money lost on day 1-4. 

Of course, there are campaign budgets and the CPA needs to be considered since not all campaigns and accounts will perform the same. 

But, generally speaking, if on day 1-4 I don’t see good results (especially on day 2-3), I will kill the testing campaign regardless of the Facebook Rep. 

I trust my data and I know my process of killing, tweaking, and re-launching test campaigns. This has shown me better results.

Much better than setting the budget and allowing automatic performance to optimize in the next 7 days.

I believe I can compete with any Facebook rep with my testing process since I probably have more ‘hand-on’ experience when it comes to ads. 

But, I’m always open to testing the ‘latest’ suggestions from our new and old reps!

I believe you can’t be a nerdy-greedy human who thinks he knows all about FB ads. 🙂 

As I don’t. 

I’m always open to learn and test new things and recommendations.

However, I know my data and the KPI my clients need to be profitable. 

So I’d prefer to kill the test before it reaches the point of no return. 

Last week, I did a new test. 

I calculated the best testing campaign budget for a specific account, set the structure exactly like the FB rep told me, and put it on auto for 7 days.

After 5 days, I was so curious about what was going on with my ‘big test’, that I checked it and killed the whole campaign immediately.

The budget was $650 per day and my KPI is $25 CPA in this account. 

After around $3k spend, the overall CPA was $45.

So don’t tell me now that FB will optimize the whole campaign magically in the next 2 days and the CPA will drop by $20.

Even if it decreases 2x, it won’t remove the damage that was already done. 

So I killed ‘the big test’ again and said a few bad words in my mind to this rep.

Although the structure she suggested was very interesting and the budget calculation was on point.

She recommended reducing the number of campaigns and combining budgets to get out of the learning phase and use the budget more efficiently. 

This was a recommendation for Prospecting Campaign – (maximum 4 ad sets) 

Ad Set 1 – Broad Audiences 

Ad Set 2 – Lookalike audiences 

Ad Set 3 – Detailed Targeting audiences with Audience expansion on 

Ad Set 4 – Custom Audiences 

To get out of the learning phase, each ad set needs to get at least 50 purchases every 7 days. 

If your average cost per purchase is $40, the budget she recommends should be at least $286 per ad set ($40 x 50 (purchases) = $2000 (weekly budget) => daily budget $3250/7 = $286 (per ad set, just to get out of the learning phase). 

If you were going to set up all 4 ad sets, the minimum daily budget to get out of the learning phase will be $286 * 4 = $1144 daily budget.

She recommended allocating enough daily budget at the campaign level to get out of the learning phase and to scale. 

For example, if you move forward with the recommendation above, she recommended a daily budget of $1716 for a campaign with 4 ad sets (approximately daily budget of $429 = ($286 x 1.5).

The same recommendations apply to lower-funnel campaigns as well.

Any time you’re setting up a campaign, keep in mind the average cost per conversion (purchase) in the past 30 days and use that cost per conversion to do the math above.

What’s my point of this article?  

Test on your own!

I’m not saying this strategy might not work. I believe that in some accounts where higher CPA or margins are possible, or where branding is more important than conversion, it can work for sure.

We tested this strategy in 4 of our client’s accounts and only one account had decent results.

All other 3 were a big failure.

But maybe it was just the time of the year. 

Or an algo hick-up that ruined the performance.

All a 3rd party FB issue that is hidden from us.

If you did some similar test, please let us know how it goes!

We want to figure out for what type of accounts and/or products this strategy might work well. 

If you want us to run your ad account for the best ROAS, you can book a call with us here https://dimniko.com/msp-apply

Have a successful and awesome day!

Matej 

ARE YOU READY TO SCALE YOUR ECOMMERCE BRAND?

Want In On The Juiciest Brand Building Strategies? Subscribe to Our Daily Newsletter

 

Hong Kong:
Unit 1102
11/F 29 Austin Rd
Tsim Sha Thui KL
Hong Kong

Petoria:
60 Glenwood Road
The Aviary, First Floor, Office 2
Glenwood
Gauteng - South Africa
0081