Home > Paid Traffic Sources > Facebook & Instagram

---

[A-Z GUIDE] CRACK The FB ALGO + EXAMPLES (In-Depth) (51)

---

In order to be able to run campaigns profitably on Facebook you need to know how to work around their incredibly complex but stupid(not really, you will understand why later) algorithm. You probably heard before things like "dup the adset 3 times", "dup the ad 3 times", "run 1 ad per adset", "max bid 10x", etc. It can be very confusing sometimes, Facebook changes A LOT, it's a very dynamic environment.

There are so many aspects of the Facebook Algorithm, here are the main ones:

1. Auction System
2. Campagin Structure
3. Machine Learning
4. Quality Score

(click on the picture to see it properly)

In this post I will try to get right to the point and share the best practices we use. However, I still want you to understand the logics behind every single thing I'm showing you guys. Here's what I'm going to cover in this post:

1. How the FB Auction System Works
2. How to Structure Campaigns Properly + Examples
3. Machine Learning Practices
4. Facebook Quality Score System - WHY?? HOW??

Before We Start
Before I'm getting into how it actually works, I want you to understand that there's a reason for everything that happens! The algorithm is VERY smart, if you're not getting the results you want(not in terms of profits but in terms of spend/reach/day parting/etc) it doesn't mean the algo is stupid like many people say, it means you don't know how to make it work properly.

How the FB Auction System Works?
I highly suggest you to learn more about the VCG auction(Vickrey–Clarke–Groves auction)! From Wikipedia:
In auction theory, a Vickrey–Clarke–Groves (VCG) auction is a type of sealed-bid auction of multiple items. Bidders submit bids that report their valuations for the items, without knowing the bids of the other people in the auction. The auction system assigns the items in a socially optimal manner: it charges each individual the harm they cause to other bidders.[1] It also gives bidders an incentive to bid their true valuations, by ensuring that the optimal strategy for each bidder is to bid their true valuations of the items. It is a generalization of a Vickrey auction for multiple items.

Consider an auction where a set of identical products are being sold. Bidders can take part in the auction by announcing the maximum price they are willing to pay to receive N products. Each buyer is allowed to declare more than one bid, since its willingness-to-pay per unit might be different depending on the total number of units it receives. Bidders cannot see other people's bids at any moment since they are sealed (only visible to the auction system). Once all the bids are made, the auction is closed.
Read more about VCG Auction -
https://en.wikipedia.org/wiki/Vickre...Groves_auction

Bottom line, the whole FB advertising system is based on AUCTION! There are literally MILLIONS of auctions a day! In the real world, money matters, the highest bidder for a specific item is the winner, with Facebook advertising it's bit more complex. There are quite few more factors beside the bid such as relevance score, link ctr, engagement, etc - this means that an ad that’s high quality and very relevant CAN BEAT an ad that has a higher bid, but is lower quality and has less relevance. It's important to know that there is NO accurate formula for winning auctions, you need to understand that the advertiser that provides the HIGHEST TOTAL VALUE wins and gets shown!

Facebook Audiences & Pools
Things You Need To Know:

• When you you launch a brand new adset and you define a target audience, you actually reach a portion of this whole audience. I call those portions pools. Those pools can be segmented based on purchase behaviours, engagement(such as shares, likes and comments), interests, etc. The longer your campaign run, the pool is getting more and more relevant because of Facebook's Machine Learning(I will cover it later on).
• Your goal is to get FB to hit the highest quality pools / the most suitable pools for our product/service.
• The right way to tap into your relevant pools is by providing the highest total value as discussed in the previous section.
• Ads within an adset compete for the budget - DON'T RUN DIFFERENT ADS in the same adset!
• Usually only 1 ad sticks.

If you will duplicate a brand new campaign/adset 10 times on audience of 2+mill (to avoid overlap) you will see that you're getting different results on each campaign/adset - it happens because FB hits on different pools even if the targeting is the same. The image below demonstrate how the audience is segmented into pools, the sky-blue square represents the whole audience(WC, 40+, men, USA), and the purple circles represents the pools:


How to Structure Campaigns The Right Way?
The main reason that structuring campaigns properly is so important is because it will allow you to split test properly and analyse data better by getting more reliable and accurate data. If you don't structure your campaign the right way, you might lose a lot of valuable data about your campaign. There's no need to elaborate too much in this section, so let's dive right in.

There are 2 main ways to structure the campaign properly -

1. 1 Campaign - 1-3 Adsets / ad - 3 ads (same ad) - this is the best way to go with WC objective campaign
2. 1 campaign - 1-3 Adset / ad - 1 ad

For each unique ad, we will run a campaign with 1-3 duplicates of the same adset(same targeting & budget), inside each adset there are 3 duplications of the exact same ad(same image, headline, text, URL, etc).


EXAMPLES
Background Information
I want to share some background information about the offer I were running.

• Vertical: Finance
• Type: Lead Gen
• GEO: Sweden
• Payout: $55 for a lead

Below you can see 3 DIFFERENT campaigns, but they all act the same - only 1 ad sticks!



(click on the picture to see it properly)



(click on the picture to see it properly)



(click on the picture to see it properly)

I have tons of examples of this, I hope you get the idea.

Off-Topic Note: I know a lot of you guys run in the USA, and that's GREAT! However, try to explore other GEOs, don't limit yourself to one geo/vertical, get out of your comfort zone. If you are worried about the language barrier try to start with United Kingdom / Australia / New Zealand - those countries main language is English as well.

Machine Learning Practices
The golden rule here is that The Machine Learning Algorithm is driven by Relevance Score & Engagement. Facebook is trying to provide your content to the relevant people! If the algo sees that you have low engagement it will stop spending/burn your money because it doesn't know who to target.

Things You Need to Know About FB's Learning Machine

• There are 2 main phases of the FB algo - Phase #1 is Learning, and Phase #2 is Optimizing.
• The optimisation of the campaign is on the adset level - according to Facebook itself, these days it's enough to have 5 conversions in a 7 days window although I think you need to have at least 15-25 to start optimizing properly. If you think you can't reach 15-25 conversions a week with the current daily budget, you can raise your budget OR just change the optimisation goal (for example, from Purchases to ATC, or from Purchases to Leads, or from Leads to VC, etc).
• Give the algo time! Let it run for few days, take it easy - the algo is now working for you.
• The algo doesn't like changes! Once you launch, don't touch the settings of the campaigns unless you have to - you can change your bids and budgets, besides that nothing!

Facebook Quality Score System
Facebook has an internal measurement system that ranks your account based on your activity. The higher rank you have the better treatment you will get from the algo - winning more auctions, getting better traffic, getting campaigns approved faster, lower CPCs, higher thresholds, etc. We call this internal measurement Account Quality Score, or in short QS. It's really important to understand this measurement because you can have killer campaigns, but if your account QS is low than you will not get the expected results. Imagine squeezing 10%, 20% or more for just having a good QS. QS MATTERS! Bottom line, like in life, when you treat others well, they will treat you well too - what i'm trying to say is that you need to work with the algo, satisfy it!

How to Get a GOOD QS?

• Have a consistent high Relevance Score on your ads
• Launch daily, even on small budgets.
• Engage with Facebook Help Center - ask questions, read!
• Build your fan-page, post content, comment on people's posts, reply to DMs, etc.
• Test your domain weekly - make sure you're not getting blacklisted on - http://multirbl.valli.org/ (https://prnt.sc/h3u1ka)
  
• Don't use low quality domain extensions such as .xyz!
• Don't create more than 1 Business Manager if you don't have to.
• Don't create more ad accounts if you don't need to - (however I do recommend to have at least 2 ad accounts, in case one gets flagged).
• STAY COMPLY! - don't get disapprovals, check your ad creatives, domains, landers, content twice before uploading it to FB. Also make sure your website is FULLY complianced with FB - have a privacy policy, terms of use, disclaimer, contact form.
• Have Facebook on your website - place the pixel, if you can also add a like box to your website, allow visitors to engage with Facebook through your website (allow them to like and share, and if possible to comment as well).
• Add few CCs to your Ad Manger account - pay on time, reach thresholds, make sure your CCs not maxed out.

Be GOOD with FB, and FB will be GOOD to you!


Final Words...
This post is a bit longer than I expected it to be, but it's worth reading. I'm sure that it will help for a lot of you guys.
Please let me know your thoughts about this.
Feel free to share your experience with the FB algo - practices, ideas or just general things.



Mation

---

Great post.

So often - in AM and in other disciplines - people focus on the tricks that gain immediate benefit, not understanding the systems that make the tricks work.

But if you understand the systems, you can generate the tricks - and come up with new ones that your competition won't anticipate.

Valuable stuff.

---

Awesome breakdown Mation! I follow a very similar structure. I've also been testing with lifetime budgets vs daily. I'm getting better results with lifetime, the reason being I believe is that Facebook don't have a daily cap to spend our budgets when running lifetime. So, say your offer converts best on a Tuesday and Thursday, Facebook seem to make sure that they allocate a good portion of budget to those high converting days and pull back on the poor converting days. Something I've noticed anyway and continuing to test with good results.

---

Solid information!

Love the simple yet effective graphics as well.

Thanks so much for taking the time to share such great insight!



Amy

---

Great share!

---

Exactly!
I know quite few affiliates that just know tricks on FB, and when they can't use those tricks anymore they are lost.

---

This helps so much if you are doing Facebook for the long term.

Understanding their algorithm is absolutely key - you have to learn how to feed it the right information.

---

More of this! These are the kind of threads that give real deep insights for affiliates of all levels. Nice!

---

Grat info, thanks for the detailed post!

I think studying the algo of any new traffic source is a must these days if you want to do volume.

And looks like high ctr, relevant content is where every source is going.

---

- Don't create more than 1 Business Manager if you don't have to.
- Don't create more ad accounts if you don't need to - (however I do recommend to have at least 2 ad accounts, in case one gets flagged).

Can't really agree on this to be honest. Spreading your camps over multiple accounts is better for several reasons;
- you don't need to put budgets very high (better daily 10x50 euro than 1x500 euro), this way you stay off the FB radar
- playing with smaller budgets on more accounts keeps your cpc's low & steady, for smaller geo's/audiences especially very important
also saves a lot of time, not have to check, optimitize and refresh your ads all the time
- you are building on a group of accounts with (I assume) good reputation, very important for the long run
- like you said yourself, if one accounts gets banned you are not left empty-handed

Decent guide furthermore, thanks!

---

You're having good points here, and I agree with you on some of them.
Please pay close attention to what I said -
- Don't create more than 1 Business Manager if you don't have to.
- Don't create more ad accounts if you don't need to - (however I do recommend to have at least 2 ad accounts, in case one gets flagged).

If there's no need, don't just create ad accounts and BMs.
FB likes activity, it's directly related to the other point I mentioned which is "Launch daily, even on small budgets."


Cheers!

---

I'm totally obsessed with this, I literally NEED to understand why and how things happen in AM, I can't sleep until I reach at least solid level of understanding

Great post mation, I hope you have more like this one stored for us, I love it

---

I think it really matters what are the situations in which you think you have to and which you don't. I guess the first obvious one is that if you run some offers FB doesn't like and you are likely to get your account banned, you want more but I can see situations in which even white hat beginners would want more than one ad account.

Just my personal situation, an ad account got suspended and had to wait about 3 days to have my ID checked and account re-enabled - if I had only 1 ad account or only 2, and this would not have been re-enabled since FB tends to be slow and sometimes random with their reinstatement, wouldn't I have risked losing all the Business Manager?

Where's the right balance? And how much does the QS impact your very early reach if you are still using the ad account to test out waters on offers that are not 100% proven, in order words, you don't use more accounts for scaling.

---

GREAT QUESTIONS!
1) Theoretically, if one ad account gets flagged it not affects the other ad accounts/BM(besides lowering the QS), however, FB might get your whole BM flagged because of this one ad account - there can be many reasons for getting flagged on an ad account such as - suspicious payment method, unusual activity and for policy violations. Policy violation is the most risky flag you can get.
2) This is a tricky question - unfortunately, there's no any accurate measurement or a tool that can tell you your ad account QS - however, you can maintain a good QS while testing campaigns, there shouldn't be a problem! Facebook doesn't rank your account QS based on how much you profit, FB ranks your account QS based on the engagement your ads get(check out the "How the FB Auction System Works?" part of the guide).

Few general things you NEED to know:

• If your personal ad account is flagged, you can create BM, make sure you do it ahead of time.
• If you have only ONE ad account under your BM, and it gets flagged, than you can't create another ad account. So make sure you create 2 ad accounts ahead of time.
• Facebook allows you up to 5 ad accounts per BM and only 2 BMs. However, in case you need more they can raise the limit - do it only if necessary!
• Have a different CC on each ad account. Never use same CC more than one time!

Mation

---

Well pointed, sir...well pointed!

---

Great post mation. I don't wish to cloak but want to make decent profits.

What do you suggest is a good cadence or time frame for scaling and raising the bids/budget on a campaign? For example: 2 months if more aggressive type of camp/offer vs 2 weeks is more on the whitehat side of things? I keep hearing conflicting advice.

---

Great share and explanation about the different pools and how to increase the QS! This will help a lot of people for sure.

---

I paid close attention to what your said. "Having to" and "need to" are subjective terms. Nobody HAS to do anything ofcourse, if you want to run on one account. But If you want to scale to xxx/xxxx a day I highly suggest to launch camps on multiple accounts. If you create 5 Business accounts, why not use them all to create good reputation. After that you can request for more BA's, if you have been a good boy you will be rewarded by FB with loads of accounts....

---

Thanks for this great guide and making it so easy to understand for a newbie like me also.

---

Hi @Mation,

Excellent Post. Thanks for shifting the way I look at FB Ads....

Can you clarify the ideal Campaign Setup? Below are some questions....

Campaign -> What Conversion Setting Would you recommend?
-> Adset - How Many Adsets? Is the targeting the Same across each Adset?
-> Ads - Are these 3 Duplicate Ads?

---

Hey @kidbrando,
First of all, thanks for your feedback.
About adsets it's up to you. About the ads just dup 3 times the same ad (you can also use the same post id).

Let me know if you have more questions or something is not clear.

---

Great post, thanks for your share!

---

Helpful informations, thanks for this man.

---

Thank you for this! How did you get so much knowledge about Facebook? It's hard to know what's real and what's not when researching online.

---

Personal experience and masterminds.

---

Brilliant!
Thanks a lot of this.

---

Wow, what an amazing Guide! Thanks a lot Mation, lots of golden nuggets in this short guide.

---

Hello Mation!

That is the great work you did here, thanks!

I have a question about structuring the campaigns and the SS of Ads Manager that you provided.
From what I understood 1 campaign should contain 1-3 Ad Sets (which are the same) and each Ad Set should have 1 or 3 ads which are also the same.
Please correct me if I make a mistake here but your three screen shots shows the results of 3 Ads that are in 1 Ad Set in 1 Campaign.
In all of them one Ad is performing better than two others so when talking about pools you mean that each Ad hits different pool or each Ad Set hits different pool?

---

In general each ad-set hits different pool.
In each ad-set, usually only 1 ad performs better than the others. This why I dup 3 times, to make sure that one of the ads sticks.

Makes sesne?

---

So when you check the results you pick the best performing Ad (as on your SS) and the best performing Ad Set as well.
Now it makes sense, thanks!

---

Hey @mation , thanks for the guide.

Did you see any hardcore changes to your campaign performance for last few weeks? I noticed that my campaigns become crazy - stable CPM and CTR for ads, but CVR living its own life - every day it`s different.
I`m confused - could it be some fundamental changes to the algo, or it`s just end of Q4 affect so much, or anything else...

---

Sorry to interfer here, but possible reasons for that might be;
- check with network/advertiser if the offer still works well - test similar offers and see if shows same differences in CR
- you might saturated the audience, find new audiences or expand a bit with age only for example
- if you scale hard, you will get at the end some low quality traffic too, which will hurt conversion rates

---

Yo, thanks for attending to my problem.

We already discussed this at iStack chat)
No of the decisions we discussed worked - I`ve tried dupe adsets, dupe campaign, make fresh ads with same post id, changed creatives.
No positive result at all...
It`s still converting in the very strange way - different results every day.
I also noticed a strange behave for some adsets - awesome CVR% for 1st day after creation and sh*tty results for the rest runtime.

Now will try the same funnel at the new acc, and waiting for April...

---

I'm confused. I understand duping the ads so that 1 adset contains 3 duped ads, but I'm confused about the adsets themselves since it sounds like what you're saying is that the "different pools" applies at the adset level and the ad level. Let's say we just have 1 audience we're targeting. Instead of having just 1 adset with 3 duped ads, it sounds like you're saying we should have 3 duped adsets AND each adset contains 3 duped ads. So for targeting 1 audience you'd have 1 campaign, 3 duped adsets, and 9 total duped ads. Am I understanding that correctly?

I can understand most people would want to avoid this due to 3x your test budget, but now I'm curious how effective it could be assuming that's what you were actually saying? Was just chatting with someone and they said they're comfortable running the 1 adset with duped ads and hopefully running it over a period of 2-3 days will give it a good chance. I guess you could spend the same in 24 hours but then dupe the single adset for a total of 3 adsets.

Curious to see how you guys approach this, especially in terms of testing brand new offers/products that you don't have pixel data for.

---

I thought I wasn't confused, but this last few posts did confuse me again.

I get that dup'ing adset level picks a new pool, but why have the exact same ad duped 3 times if it is shown to the same pool?
Logic would dictate to have 3 different ads and dup the adset, and test which ad works best with which pool, and kill the other ads.

Otherwise, to me it is just skewed machine learning. Say in a certain GEO, at a certain time - dunno - people get home from work or a certain niche does something online collectively

Ex: 3 same ads in same adset

There is a sudden influx on clicks on a certain ad, the algo will favor that ad and demote the others. So that ad will have better QS, have better reach, quality and naturally CTR. If they are exactly the same ad targetting exactly the same pool of people, then statistically there should be absolutely no variance if you let it run long enough.

Unless, each ad is sent to a subset of that pool, and not truly randomized. Then all of this takes on a different logic and makes perfect sense. If it works this way, then FB just became 1000x simpler to work with - for me at least - running no AM stuff but rather brand building.

---

thanks!

---

thanks very good stuff !

I have one questions for those following this optimisation method, do you desactivate manually the 2 less performing ads or justkeep them alive expecting the algo will send them less traffic?

---

Depends on the EPC ofcourse. Let's say your best ad give 100% ROI and the two "bad" performers do a 50%, why not let them run right? But if they are not profitable kill it yeah.

---

Hello,first thanks for this post. i've started about a month ago and i've had few sales, not very succesful. Now i'm following your advice here, i've made 1 campaign with 3 adsets and 3 ads for each adset, i did have a sale in 24 hours but my reach seems to low, about 100 per adset per day on audience that's fairly broad 27M. The budget for adset is 5$, i've had campaigns for the same video and same targeting with lot more impressions and reach. Any thoughts on that?

---

I'm new to this forum, so i hope someone responds to my questions, thank's. Just another newbie question, when you say "duplicate adset" do you mean adset or ad within the adset that is profitable, because i have 3 ads per adset and some profitable some not? Also i have 3 adsets with 3 ads each, and the budget for each adset is 5$, and for each ad is 5$, so it might be a stupid question, but i have a feeling it only spends adset budget. Is there a difference? Thanks

---

Thanks for your questions creator111!

Unfortunately, mation is extremely busy nowadays with his business, and no longer has time to grace the forums.

I would suggest starting a new thread to ask those questions, so that other FB experts may see them and give you an answer.

Thanks for being a part of this community!



Amy

---

What would you say is a good minimum daily budget for this type of FB testing?

---

Just to clarify the campaign structure should be one of the following:

1. 1 Campaign ---> 1-3 Adsets --> 3 ads (same ad)
   Can the adsets have different detailed targeting/interests? and why have 3 of the EXACT same ad in each adset??
   Before it was stated that "Ads within an adset compete for the budget - DON'T RUN DIFFERENT ADS in the same adset!" So is it okay to have multiple ads in 1 adset as long as they are carbon copies of each other
2. 1 campaign - 1-3 Adset / ad - 1 ad

   ---

   Regarding the example images they aren't loading for some reason

---

Why can't I see the picture?

---

If you click on the "broken" icon of the image it will open in a new tab (at least here). On mobile I have more troubles seeing images too agree!

---

Very helpful, especially about the quality score. I wish I had read this sooner because just got a business manager disabled but luckily I had already set up a second business manager and my personal account is still active.

---

great info.

thanks a lot

---

Thanks for the post! I have a question about this.. been hearing a lot of gurus saying this but it still doesn't make sense because of .. well.. let me ask you this..

say I want to obtain purchases from the facebook audiences and I have 0 as I'm starting off.. if I optimise for view contents, my adset will be optimised for view contents not for purchases. Do you propose to change the optimisation of the adset by editing it after getting about 25 view contents? And then editing the adset again to optimise for purchase after it got 25 adds to cart?

I tried that many times.. and when I edited it didn't really work out.. or are you suggesting something different?

Thanks!

---

Newbie questions.

1- I have like 10 FB fan pages I opened in the last 10 years but don't really use, should I delete them?

2- I just found out that I have 6 business managers ( one of these business managers has one account ) should I delete them?

3- I'm running ads under my personal Facebook profile ( they lead to a Clickbank product ) Is this OK, profitable or should I run these ads under one of the business managers?

Please help!

Luis

---

1-3-3 right?

Thanks a lot!


Sent from my iPhone using STM Forums mobile app

---

Such a great post.

---

Home > Paid Traffic Sources > Facebook & Instagram