If you use Facebook as a marketing platform, you must know about campaign budget optimization.
If you don’t, then it’s best to familiarize yourself with it now.
Campaign budget optimization is the future of Facebook advertising; what you know now about running ads on Facebook will be irrelevant in the coming months.
You’re going to start setting budgets at the campaign level as opposed to the ad set level. Facebook’s machine learning algorithm will then distribute that budget to the ad sets.
A semi-hands off approach to running ads which will also optimize budget allocation and in turn, reach.
So to set you up for success, I’m going to teach you how to dominate campaign budget optimization (CBO).
If you don’t know how to run a successful CBO strategy by now, you will by the end of this article!
We’re continuously testing CBO Facebook ads, here are our results for the Month of December 1-31, 2019:
As you can see, our total amount spent was $188,076.00, and our return was $425,338.99.
However, our breakeven point was 1.4 return on advertising. We also experienced a 25% loss in sales on our Ads Manager dashboard due to tracking errors.
Our gross sales were approximately $531,673.74, and after the cost of goods and ad costs, we profited approximately $268,367.34 while testing campaign budget optimization.
So there are no doubts that this change will be a good one.
The only caveat is the new structure of your campaigns.
Because with CBO campaigns, you relinquish the majority of control over the ads and pass it along to Facebook’s machine learning algorithm.
Additionally, CBO requires that you start with larger budgets.
But don’t worry! I’ve devised a few plans for tackling CBO at any budget!
In the following paragraphs, I’m going to walk you through:
If you are brand new to eCommerce, as well as Facebook ads, I recommend checking out our Ultimate Guide on Facebook Ads for eCommerce.
In that article, you will find everything you need to know about what to include in your ads and how to set them up.
For those of you who are already pretty familiar, let’s go over the structure of the campaign.
The best way to show you how campaign budget optimization works, I need to show you a breakdown of the campaign structure.
The campaigns are the umbrella collection of ad sets that contain the ads. The campaign level is where you’ll set the budget.
The ad set is the group of ads that share the same targeting, bidding, and scheduling. Before CBO, the advertisers set the budget at the ad set level.
The ads are the creatives that are shown directly to your audience and drive your sales.
Putting It All Together
As shown in the visual above, the ads reside in the various ad sets within a single campaign.
It will help to reference back to this image when I go over what campaign budget optimization is.
Campaign budget optimization (CBO) is an ad strategy that establishes the budget at the campaign level. The budget then gets dispersed to the different ad sets by Facebook’s machine learning algorithm.
The campaign actively monitors each ad set’s targeting and performance.
The campaign attempts to allocate budget to the best-performing ad sets based on the objective or pixel even you’ve applied.
It uses real-time data, every day, at the ad and ad set level to distribute the budget most optimally.
In short, Facebook’s machine learning allocates the campaign budget to the best performing ad sets.
According to Facebook,
“We optimize your campaign budget in real-time on an opportunity-by-opportunity basis. Our goal is to get you the best results possible, and for the cost of those results to align with your bid strategy.”
Additionally, instead of manually shifting budgets between ad sets, Facebook’s machine learning assesses all data points and adjusts the budgets for you.
Once the machine learning shifts the budgets, it passes the information off to an automated system that will place your ads.
In short, CBO optimizes the budget as well as the reach.
Last year, strategists at Facebook noticed that advertisers were hitting a wall when it came to making profits on their platform.
They had also noticed that machine learning could outpace what marketers could do manually.
So they devised a solution to optimize ad delivery through CBO, as well as to eliminate the workload that comes along with manually adjusting budgets.
The machine learning achieves this by looking at multiple data points throughout each campaign and adjusts the budget accordingly.
So overall, this will shave off a significant portion of your ad monitoring while yielding profitable results.
I know it sounds scary to let an algorithm take the reins on a task dependant on nurturing. But, if you keep reading, you’ll see how simple this change will be.
The shift to CBO is supposed to finalize within the first few months of 2020 officially.
Since the effects will be widespread, we’re going to show you the best ways you guys can adjust to the mandate.
1. Set broad parameters
Facebook’s machine learning algorithm needs time and space to learn. By setting broad parameters for audience sizes and budget allocation, you’re allowing it to optimize to its fullest potential.
These parameters give the algorithm an adequate range to discover proper placements and access as many opportunities to meet your campaign goals.
2. Align goals with realistic outcomes
You should continue to optimize your ads for conversions and purchases. However, be realistic with the time allowance and amount of money you budget to yield the best results.
The system can navigate towards the goal entirely on its own; it just needs the proper foundations to do so.
3. Each ad set requires at least 50 conversions for stable delivery
Variables such as the time of day and the day of the week can affect your delivery.
So don’t lose hope when you launch your ads, and the metrics aren’t promising.
It can take up to a week to see accurate reporting.
So sit back and let the machine learning do its thang.
4. Set minimums and maximums
Setting budget minimums and maximums is relatively new.
Sometimes the machine learning can over-allocate the budget to a single ad set while not leaving much money for the rest.
So to make sure this doesn’t happen to your ad sets, take advantage of this feature.
Remember our fun little photo at the beginning of this article?
The one that showed our 2.5X return during our trial phase.
Well, I’m going to give you the exact blueprints that we use to dominate our CBO campaigns so you can see the same results.
The first outline will be for your test trials, and the second outline is for your CBO launch!
Just keep in mind, the algorithm takes time to learn, so don’t get discouraged if your results don’t meet your expectations at first.
Before we get into how you should structure your CBO test trials, let’s cover the key takeaways that will set you up for success.
Step 1: Optimize for the purchase or conversion objective goal
Step 2:Create your CBO campaign with grouped audiences you want to test
Step 3: Create the ad sets
Step 4: Set a minimum budget for each ad set
Step 5: Select placements for newsfeed only
Step 6: Use the same ads in each ad set
Once you collect the proper amount of data, you can then take your best performing variables and plug them into your CBO campaign.
Step 1: Start with a $150 Daily Budget
Step 2: You should have 4-5 ad sets per campaign
Step 3: Launch your campaigns at 3 a.m. EST and give them at least one to two days to stabilize their delivery
Step 4: When your campaigns have a consistent 2X return on ad spend, scale up by 20%
If you don’t have $150 at your disposal daily, don’t worry, many of us don’t.
So we recommend starting with one campaign that contains 2 ad sets, with one ad in each ad set.
Start with a daily budget of $25, and don’t worry about setting a minimum.
I do want to warn that the data tends to look pretty depressing when the campaigns start.
So don’t compulsively check your metrics each hour.
We recommend checking once in the morning and once in the evening. Remember, give the system time.
The point of CBO is to learn more about Facebook’s machine learning than human interference.
It’s going to take a lot of patience to let Facebook’s system do the heavy lifting but trust the system.
With our experience, CBO has been extremely effective and has accomplished what Facebook has intended to do.
We’ve profited on the front end, which has allowed us to retarget more and bring in a profit on the back end.
CBO becomes more consistent as it scales.
So increasing the budget on a CBO campaign doesn’t affect the performance as much as it would on the ad set level.
Below is a checklist to guide you through constructing and launching your own CBO campaigns.
All ad accounts will begin getting switched over to CBO permanently.
The most significant changes to note are:
The change will take over all ad accounts within the next six months.
You need to begin testing out CBO’s of your own.
Reference our steps in this post and use the checklist we’ve made to guide you.
Now is your chance to get a head start and learn what works best for you before having no other options.
And if you want to succeed with campaign budget optimization, make sure to download our free Copywriting Bible.
The Copywriting Bible is a breakdown of the best copywriting techniques used across multiple 7-figure eCommerce stores such as Urban Outfitters, Zappos, and Apple.
You can use these same techniques in your ads to hook browsers and drive sales.
If you have any special CBO techniques of your own, share them in the comments below!
Do you recommend group 1% of different lookalikes into a single CBO
I.E 1% Pageview, 1% content view, 1 % initiate checkout, 1% atc
I don’t have a LLA yet. Should I still create a CBO with 2 add sets for $25 to test?
And once I have enough conversions, can I add a LLA with a broad audience to the existing CBO?
Hi Patricia! Great post 🙂
I’d love to have your insights about 2 doubts I’ve.
The first quesiton is about expand detailed targeting. What’s your advice?
1. go to winnings ad sets and tick on the “expand detailed targetin”
2. duplicate the winning ad sets and tick the “expand detailed targetin” on that new ad sets?
Another question is – A lot of people are saying that we shouldn’t pause or add ad sets within a CBO. Cause it’ll mess the algorithm optimization. Do you have the same experience?
Is the strategy applicable for any country like PHILIPPINES?
Thank you so much
Hi – I could really use some advice. I’ve heard that when targeting a very broad audience, say 6 million people, that you should duplicate ad sets exactly the same with the same creative and test the algorithm as apparently this can vary ad set to ad set? And slowly turn off the ‘losers’. Is this good practice? Is there a better way to improve results? I’ve had experiences where ad sets just simply don’t want to kick it up a gear so to speak and duplicating them seems to help. Any advice is appreciated!
If I have a wall post image/ad that I want to (boost) as an ad on facebook or instagram,
and I create a campaign, should I run the exact same post in ads towards multiple audiences at the same time? To see which audience reacts the best to it? OR is there some sort of overlap that happens, say likes were coming on said post from one audience does fb or ig calculate and charge for engagement coming from all audiences at once because they are all funneling to the same post? Or do I need to run seperate posts of the same ad to seperate audiences to prevent that cross over?
You can do that, you don’t really need to start worrying about overlap until you’re spending $1,000+ on ads
What if I test a CBO campaign at $100 and it’s doing really well? I really am trying to scale fast so can I just double the budget by duplicating in a new campaign? I’m paranoid of scaling the original budget and having performance tank.
You should always scale the original budget by 20% as mentioned in the post, you can always dial the budget back if it is too much
Thank you for your post. You have done a good job
I’m glad that you liked it Sea!
Great article!! Appreciate the tips and tricks. I was trying to access the Copywriting Bible you referenced but the link isn’t working. Is there another place to access?
Thank you again!
Great article Patricia many thanks! If you have a successful cbo (that has stabilised after 50 purchases) and you duplicate this cbo will the duplicate cbo re-enter the learning phase or will it replicate the learning and stability of the original cbo? Thanks
Technically they should, but on rare occasion they won’t
Hi Patricia, you keep mentioning that is best to turn off ad rather than ad set. I don’t get it, if the ad set targetting is not working, why still bother to keep it?
If you have an ad set with 4-5 ads in them, then you just turn off the ad that isn’t delivering. If you are going the route with 1 ad per ad set then turn the entire ad set off
Hi Patricia, thank you for the very helpful article! I have a pretty basic question that I haven’t been able to find the answer for, and hoping you might know.
If I have multiple ads within an ad set, is Facebook also automatically budgeting within the ad set itself on what ads should get the most spend based on performance? I would assume so but I couldn’t find anything to confirm or refute it.
Yes, Facebook automatically favors the ad sets that perform the best 🙂
Hi Patrica! Great article thank you. What happens if you end up with just one ad set within a campaign which is profitable- ie you’ve turned the others off gradually. Is this ok to do? And can you then scale this campaign so it again just has that one ad set?! Or is there some sort of science to retaining several ad sets, irrespective of how they may be doing? I’ve ended up with 8 campaigns, all with just 1 ad set!
Yeah, that’s totally ok to do! as long as you’re making that money 🙂 you don’t want your winning ad set competing for the budget with one that’s underperforming
Superb article. How do we get in touch to discuss more please?
Glad you liked it!
You can follow us on Facebook: https://www.facebook.com/SkupEcommerce
Or join our Facebook group: https://www.facebook.com/groups/useskup/
Hello! Great advice, thank you. I did well with manual budgets but am still struggling with CBOs so I’m trying to stick with the £25 budget/ 2 ad sets for now. My Q is this:
1. do I need different interests in each ad set? If I’ve already found my ideal interest groups, can I just use the same for both ad sets..?
2. When scaling, should you always use LLA? What happened to the winning ad set audience?! Is that just to help Facebook during the testing phase?
Thanks so much- apologies if I’ve misread/become muddled somehow 🙂
So you don’t necessarily need new interests but the point of having the different ad sets is to test new audiences, placements, bid strategy, etc – so it would make sense to see which audiences + interests work best for you
And you can scale by increasing the budget (vertical) or scaling with LAA’s (horizontal) the winning ad set audience is going to provide you with the data to use for your LAA
I have a question about the minimum and maximum. Do you guys remove the minimum and maximum of the ad sets at a certain point? Does removing a minimum or maximum make the ad set have to reoptimize and screw things up? I’m currently running 7 ad sets at $100 a day with each having a minimum ad spend of $7 so half of the budget.
That’s a great question! However, I don’t recommend removing them once you set them… it will shift the way the ad sets behave and you don’t want to cause big changes for something that’s already working
Could we have 3-4 ad sets with a $25/day budget or just 2
Sticking with just 2 ad sets will be more profitable for you. CBO works best with a good portion of the budget being spent on each ad set, when you add 4 ad sets with a $25 budget, you’re giving your campaign little room to spend.
Hi Patricia, thanks for this information.
I just have few questions:
Do you have any rule inside your CBO campaign to stop adsets? For example. If my adset spent x2 my Break even point i stop that adset..
On the other hand, you recommend having between 4-5 adsets/cbo but in the case we pause any of them, can we add new adsets? I have read that this change adding new adsets “feeds” our CBO and keep it “alive”
Thank you for your help
You can set rules and add in new ad sets, but I don’t recommend doing so. Adding new ad sets to a pre-existing campaign can cause it to re-enter the learning phase. Don’t fix what ain’t broken 🙂
Thanks to you guys for this! I have a question as regards, adding or removing interests from an active adset?
In the past, I have noticed making changes to an active adset can disturb its delivery? What are your thoughts about this? Can I make a frequent change (change like adding more interest, excluding ages ranges, or excluding custom audiences) to active adsets without it affecting its delivery?
I don’t recommend tinkering with active ad sets. You can pause any ads within them that aren’t performing well and creating new campaigns with new ad sets and updated targeting + interests
Should we select purchase everytime even if our pixel doesn’t have a ton of activity for it or should I go with a lower conversion like add to cart that has a lot of activity as someone on a smaller budget?
Hey Derek, you always want to choose the purchase objective – that’s what you’re chasing after 🙂
Great Article. After studying the comments as well, I just have one remaining question:
Why do you recommend starting the Campaign at 3am?
That’s a great question 🙂 It just gives Facebook ample time to start spending the budget for the ad to be profitable and get reach
Excellent article. Thank you so much for this. My question is when you have found the winning audience after testing, do you then create a new campaign with that winning audience? Or do you duplicate and/or edit the original testing campaign?
When it comes to winning audiences and ads, you never want to change up anything other than the budget. So scale your budget 🙂 You can create additional campaigns with the proven audience as well
Hello, I currently have a problem with my CBO campaign. After running a campaign with broad interest to get data, I created a Look alike audience for 75% video viewers and started a CBO campaign for 5 segments that I previously created. But when the campaign was activated, it didn’t spend any amount of the budget ($100/day for 5 ad sets) and still no reach so far. Can you please help me with the problem? Thanks a lot
unfortunately, this sounds like an issue you’d have to contact Facebook about
Question on Step 5: “Select placements for newsfeed only”.
Does this mean to deselect all placement options except “Facebook News Feed”? The course says to only deselect “Facebook right column”, so I’m a little confused. Thanks in advance for the clarification!
For CBO we test only in the newsfeed since there’s already more testing being done by Facebook on the ad sets 🙂
Hi, thanks for the value
I have a question about testing, as you know in E-commerce we should test much interest to find winners, for example, let say I have 30 interest and i want to test all of them, should I put every 10 interests in a single campaign?
Hi Ismail 🙂
You can test as many interests as you want to get yourself to that 1-2 million sized audience
Hi Patricia, can you contact us please? We would like to ask you if you have any tutorials or assessments.
We want your help. Is that possible?
Hey Rob, you can always reach me at email@example.com 🙂
I’m not in E-Commerce but am a Fitness Trainer self taught and still learning about social media marketing – Thank You Digital Marketer. This was a great informative article. Just what I needed as I experienced similar out the gate recently using CBO.
Prior to CBO, I would normally get up to 8 Leads per day but in a recent campaign I ran this month (December 2019), it took 3 days to just get 6 leads. I was disappointed and shut it down after day 3. After reading this article, perhaps I acted prematurely. But then in the only other CBO campaign I tried in June 2019, I was able to acquire 6-8 leads per day from the outset on day 1. Go figure!
In the past I’ve used something similar to the Michigan Method (multiple (usually 3) ad sets that had identical setting (audiences (cold), targeting and budgets) within a campaign. I’d run all three and the first one that would get traction would be allowed to run and I’d shut off the other two. Wasn’t the best approach as it was expensive out the gate.
With CBO, FB has pretty much automated what i would do manually.
1) Is it ok or ill-advised to create identical ad sets and identical audiences in a campaign with CBO? The intent is to see which ad set with similar targeting gets traction first.
– Or does FB consider this as siblings within a campaign fighting each other?
2) Should i try to differentiate the audience a bit? Perhaps different geo-targeting per ad set? — With regard to gender target settings, I usually include both within an ad set.
–This is because I’m concerned that with CBO, while the Algorithm may somehow give more
preference to an ad set at the expense of the other, that ad set may contain only one gender
(i.e., female). Their could still be some prospects of the other gender (male) in the other ad
sets. Basically I don’t want just one gender in my fitness classes, so I include both within ad
What say you?
As far as your first question is concerned, we don’t advise creating identical ad sets or audiences within the campaign and pertaining your second question – you should be changing up your targeting to prevent any overlap. Your final comment addresses the benefits that CBO accomplishes, you are letting the algorithm work and figure out who the best person is from a day to day basis 🙂
1. What conversion window do you choose for campaigns?
Under the “test phase blueprint” above, it says to use “one day click”. But I read in a comment below, Julia mentioned using “7-day view, one day click”, but I do not even see that option in my ad sets. The only similar option I see is “7 days click click, one day view”
Could you please clarify 🙂
2. Do you always recommend audience sizes of 1-2 million for each ad set in CBO’s?
I’ve found that in the past, Facebook always allocates the most traffic to the ad sets that have the largest audience size. For example if one ad set has potential reach of 2 million and one ad set has 4 million. The larger ad set gets the majority of the traffic.
For the conversion window – you should be using 7-day view, 1-day click, it’s what we’re currently using so it should still be an option 🙂
As far as audience size goes, you want to stick with that 1-2 million, it provides better quality traffic that will be more likely to buy than just bounce off your store
If I haven’t found a winning/profitable campaign yet should I still use this strategy? If not is there an article or link you guys might have to something that might help me find a winning campaign?
Sorry to hear that your campaigns haven’t been working out – I recommend checking out the following page to find the campaigns weak points: https://skup.net/fix-failing-ads/
Once you go through and diagnose the issues in your campaign, continue to use the strategy laid out in this article for a new one 🙂
Thanks for all of the great information! Could you please advise on the questions below:
Context: My product is priced at $39.95, and performs best on IG story and IG feed. I am following your $150 outline above!
1. Do you always recommend to optimize for purchases using “one day click” as the conversion window?
2. Can you give a little guidance on how you might set you max and min restraints for each ad set
3. For excluding previous people who have engaged with my ads, would you exclude past purchases or past page views?
4. Side note, do you ever optimize for other conversion objectives when testing new products? For example when a pixel has no data and you are testing new products
Thanks for the feedback 🙂
1. We recommend the 7-day view one day click. This allows for the best optimization during low budgets.
2. To set the min and max, go to the Budget & Schedule section. There’s a link that says, “Ad Spend Limits Within This Ad Set” and you can set both the min and max there!
3. This is something to really worry about when you are scaled up. Once you are running at a thousand dollars in ad spend or above, you can start excluding past buyers. After that, you can do past visitors as well. But initially, it’s not something to focus on.
4. Andddddd finallly, no, always optimize for purchase objective.
hope this helps!
What percentage do you use for that lookalike?
Also – for a new e-commerce business with one product, is it anything i should change from what you describe above? I.e. I don’t have any lookalike and sales yet…
For what we described it would be 1%.
And if you don’t have any sales or LLA’s yet, just start out with interests 🙂
You’re talking about target audience consisting of 1-2 million people. Is that size based on percentage of the total population in the US? Or would you say it’s also relevant for me, trying to reach swedes, or should I build audiences of 1-2 percentage of the total population of swedes on Facebook and Instagram? There’s about 7 million swedes on Facebook and Instagram.
The size is typically based off lookalike audiences and that uses a % of the population so yes you are correct
It’s impossible for us to say where you should target since we aren’t familiar with your business or target demographic… That being said you can try this:
1. Create a CBO Campaign
2. Create a few lookalike audiences in different countries
3. Use those audiences in different adsets within that campaign
This is the PURPOSE of CBO Ads and how it should be used, it’s a great way to let an AI define what audiences are most profitable and let Facebook do all that heavy lifting for you 🙂
Thanks so much for your article, it is wonderful!
May I ask you if I test 10 different interest, each interest has the audience size of 200k to 500k per interest, can I make 10 CBO campaigns with 10 ad sets within those campaigns and then when I find the potential audience of the above 10, I will make a CBO campaigns contains of say 3 to 5 ad sets containing the audience I find before and scale?
And for testing, how long should I continue to test since for the first 2 days, as you said in the articel, the matrics do not look promising.
I don’t suggest doing this. Group similar interests together into 1-2m person pools. Run a CBO of 4-5 ad sets (each get 1 pool). If testing doesn’t look good after 2 days, then you should watch the third and kill it if it isn’t profitable
recently I created 3 campaign and tested CBO…and for each campaign, I created 1 adset only. after 24 hour, facebook did not utilize my budget and only spent 20% of it. can you advise me facebook did not utliise my entire budget?
So it can take anywhere between 3 and 7 days for CBO campaigns to stabilize, so it may just need some more time 🙂
Also on Patricia’s note. It could be that the audience is too small. Typically we recommend having no less than 2 ad sets preferably 3 when using CBO. The good news about that is you’ll be split testing and finding an even better ad overall 🙂
Many thanks for your prompt reply.
I understand what you mean. I will move on to another product of higher prices and use your suggested CBO testing strategy.
The only aim is to find a blockbuster and scale it, not to try turning a product that can sell but without profit to become a winner.
I have to bear in mind not to be emotional when giving up a losing product.
Have an awesome day!
It’s what we’re here for 🙂
Come back when you need help on the next product and let’s get that bad boy rocking
Hi Devin, I can’t even begin to explain how much your help meant to me!
Please see below for more information in relation to my CBO (scaling) campaigns & potential winner:
Kindly let me know if the below data is enough. I am more than happy to provide you with more data with excel (which is clearer) through email if you don’t mind. I doubt if this product is my winner or not.
1. Initial CBO Testing
Campaign Budget: $15/day (ran for 6 days)
Product Selling Price: $14.95 (incl. tax & shipping fee)
Average Product Selling Price (per unit): $16.45 (incl. tax & shipping fee)
Average Product Cost (per unit): $7.38 (incl. tax & shipping fee)
No. of Ad Sets: 8
Audience Size: 3,100,000 – 21,000,000
Gender: Female only
Ads version: 3/ad set (same for each ad set)
Ad Copy: all the same
Results (no. of sales per day)
Selling Price: $14.95
1st Day: 3 (LOSS)
2nd Day: 0 (LOSS)
3rd Day: 2 (LOSS)
Selling Price: $17.95
4th Day: 0 (LOSS)
5th Day: 2 (LOSS)
6th Day: 2 (1 repeated customer) (LOSS)
Total number of Unique Outbound Clicks over the period: 174
Total number of Checkout Initiated over the period: 21
**My FB pixel has stopped firing Add To Cart activity since last week. Hence I use IC rate for data studying.**
2. 2nd CBO Testing Campaign (after analysis the age data of last campaign)
Campaign Budget: $15/day (ran for 7 days)
Product Selling Price: $24.95 (incl. tax & shipping fee)
Average Product Selling Price (per unit): $24.95 (incl. tax & shipping fee)
Average Product Cost (per unit): $7.37 (incl. tax & shipping fee)
No. of Ad Sets: 7
Audience Size: 830,000 – 1,700,000
Gender: Female only
Ads version: 1/ad set (same ad set)
Ad Copy: all the same
Results (no. of sales per day)
Selling Price: $24.95
1st Day: 1 (GAIN)
2nd Day: 1 (GAIN)
3rd Day: 2 (GAIN)
4th Day: 1 (GAIN)
5th Day: 0 (GAIN)
6th Day: 0 (LOSS)
7th Day: 0 (LOSS)
Total number of Unique Outbound Clicks over the period: 193
Total number of Checkout Initiated over the period: 9
**My FB pixel has stopped firing Add To Cart activity since last week. Hence I use IC rate for data studying.**
3. Current CBO Scaling Campaign (after analysis the age data of previous 2 campaigns)
Campaign Budget: $180/day (7 days as of today)
Product Selling Price: $27.95 (incl. tax & shipping fee)
Average Product Selling Price: $27.35 (incl. tax & shipping fee)
Average Product Cost: $7.50 (incl. tax & shipping fee)
No. of Ad Sets: 4
Audience Size: 1,500,000 – 13,000,000
Gender: Female only
Ads version: 1/ad set
Ad Copy: all the same
Results (no. of sales per day)
Selling Price: $27.95
1st Day: 5 (LOSS)
2nd Day: 8 (LOSS)
3rd Day: 6 (LOSS)
4th Day: 3 (LOSS)
5th Day: 7 (LOSS)
6th Day: 5 (LOSS) – turned of 2 unprofitable ads from 2 different ad sets
P.S. Audience Size of the remaining 2 ad sets: 1,500,000 & 13,000,000 respectively
7th Day: TODAY
Total number of Unique Outbound Clicks over the period: 1.273
Total number of Checkout Initiated over the period: 65
**My FB pixel has stopped firing Add To Cart activity since last week. Hence I use IC rate for data studying.**
Million thanks for your time!
Have an awesome day.
1. The $15 per day CBO is wayyyyy too low on spending for 7 ad sets – you are basically sprinkling just a few dollars on each ad set. I’d recommend going through our suggested CBO strategy and reduce down to 4-5 ad sets max. Also, for CBOs to work you need to give it some spending… in my recommendation I’d give it $50 per day to test (with the 4-5 ad sets).
2. Try to sell something higher priced. $25 is the minimum i would test a product but would suggest $40-50. $15 is way too low.
3. Don’t jump to a higher budget CBO like that if the testing failed. More money will not save a losing product.
I am so glad to find your blog!! You’ve answered so many of my queries!
As you mentioned, “So each ad set needs about 50 conversions for stable delivery, meaning it could take up to a week to see some results. But if it’s not performing after about 3 days make sure you just pause the ad, you don’t want to kill it.”
My question: What exactly does “ad set not performing after about 3 days” mean? Do you mean not in a profit after running the whole CBO campaign (with 4-5 ad sets) after about 3 days?
I am running a CBO campaign with 4 ad sets, FB is focusing to spend money on 2 of the ad sets. I am in a loss even after 3 days, the 2nd day was the best (even in a loss) performing day.
I have tested and proved few potential winners (including the one just mentioned above) with a small budget of US $15/day with CBO campaign, and found that the products were still profitable after 3-day testing. Hence, I tried to scale them with a larger budget (over $150) and each time the campaigns started to perform badly day by day so that I paused the campaign.
Could you kindly give me some valuable advice?
Many thanks in advance.
Hey Jose, glad you enjoy the content!
If a single ad set within a campaign is not performing after 3 days, i would pause the ad (not the adset) within that ad set. “Not performing” for us means spending more than the revenue of 1 sale with no purchase within 1 ad set. Understand that a CBO campaign with multiple ad sets may take more time to begin working vs an ABO campaign.
$15 to $150 is quite the large jump. Depending on your audience size, it might be to big…. are you targeting millions of people?
We would suggest starting with a smaller CBO if you are not targeting a broad set of people.
More info could help us suggest more steps.
Hello, I wanted to ask, for how long an ad should run, until I can see if it is good or not/ when I should kill it.
And is it a good idea using Facebook`s “Expand detail targeting” when matching an audience?
Bc the audience is (in my opinion) much too large with 170.000.000 people. But on the other side, these poeple have the same interests, so it would be cheaper linkclicks…
And how big should the audience size be with CBO? (At the moment 2 ad sets, 1 ad each ad set)
Thank you for your help 🙂
So each ad set needs about 50 conversions for stable delivery, meaning it could take up to a week to see some results. But if it’s not performing after about 3 days make sure you just pause the ad, you don’t want to kill it.
As far as the Expand Detail Targeting, we tend to flip it on when we’re scaling. FB is pretty good at matching now… so when we have a proven winner, we will target some good keywords and flip that on in a 4-5 ad set CBO.
you want to have about 1-2 Million audience size 🙂
Hi Patricia! Thanks for the timely information. It actually made sense that FB would make this change. Much appreciated!!
I’m glad you enjoyed it, Mel!
Hello, I have a few questions: So I should create a few ad sets (4-5) with 4-5 ads in them, where I always change one parameter (for example another targeting).
And is there something the ads in one ad set should have in common ( ex. that in one ad set I only change the targeting, in one ad set only something other….) or how is it?
What is the lowest budget possible to start with? Because I read in this article that you can also start with 25$. I know, the more the better, but is it even possible to run profitable ads with only 25$ and 2 ad sets?
So for the test phase, the ad sets should be split by audience type and use the same creatives for each ad set – that way you can see what audience type responds best to the ads.
The lowest budget possible is the $25 a day with 2 ad sets and one ad within each ad set
does the setting of min and max spend per adset still exist? I can’t find it. I tried with different CBO optimization types and adsets but didn’t manage to find it.
The min and max do still exist 🙂 They’re located in the Budget & Schedule section. There’s a link that says, “Ad Spend Limits Within This Ad Set” and you can set both the min and max there!
first, great article, thank you!
second, what are you doing with having a profitable CBO – you increase the budget on every 2-3 days with 20%, but until when? Or do you duplicate the same CBO and start with a higher budget from the bat to find more pockets?
So we recommend just increasing the budge 🙂
Hi, thanks for the valuable information.
After a week with CBO campaign with 5 active adsets, please, I have some doubts:
– Should you pause the worst adsets? or pause only your ads?
– Should I create another CBO campaign with the best adsets and increase the budget? Did the original CBO campaign pause?
– Can I add other adsetds in an active CBO campaign?
1. It’s better to turn off ads instead of adsets
2. Yes, you can scale it up by 20%
3. We don’t recommend adding new adsets to an active CBO campaign
Hope this helps 🙂
Hi team, GREAT ARTICLE! ..curious to hear some insight as I’ve toe-dipped into CBO. Before, we would use a lower budget to test different audiences and simply up the budget on the ones which perform well.
NOW.. Would you recommend taking the top 5 ad sets into a CBO campaign with say a $100 budget?
THEN with retargeting we didn’t plan to use a CBO, rather just a $50 / day into one ad set for those who visit our website with 4-5 ads.
Thank you in advance!
Since Facebook will be mandating CBO you should really start to familiarize yourself with the ad strategy even for retargeting ads 🙂
With a large campaign of over 30 Ad Sets, what’s the best way to manage the ad sets that we don’t want to spend as much money on. For example our Instagram Story Ads vs the Instagram Feed ad.
WE really don’t recommend that many ad sets unless you are spending hundreds, if not thousands a day. That’s too much for Facebook to have to spread out to on a low budget. The best way to optimize these campaigns is to turn off the ads… at least that’s what we are finding
Question: I understand CBO optimizes for ad sets, should I have many ads for each ad set? as in 1-2 ads per ad set? or 3-5? why or why not? Does CBO optimize for the ads as well then or just the ad sets?
I am really confused about this part, otherwise I feel I understand everything.
We recommend 4 for up to $100/day and 5 if above. CBO works best when it can choose between a big pool of people. Yes, CBO will start shifting traffic to the best ads within each ad set, and the best ad set within each campaign
Hi there! Great article! May I ask do we duplicate the same adsets (with the same targettings) onto the CBO campaign to test which pockets of audiences works best ?
or is it purely putting in non-duplicates adsets?
also the structure for CBO campaign would be purely putting in adsets that previously worked well , and not organized by generic category?
Eg: all lookalike purchase adsets should be one CBO campaign or it can be mixed with other interest targetings under the same CBO campaign?
Hey Charlene! So happy to hear you liked the article! We haven’t seen any great results coming from adding additional adsets into a working campaign because it throws off the learning. And as far as the structure of audience separation, we haven’t focused too much on that, we’re making sure we use proven audiences and hit 5 or so audiences within an adset. However, as you progress with the product you should always move towards look alike audiences over interests.
Hi, is there any chance to do a CBO in a smaller budget?
Yes 🙂 we’ve been working on the strategy to allow companies with lower budgets to also utilize CBO advertising
We should have it posted in a month or so
Great article, easy to understand.
I wonder if even with the CBO coming, it will still be possible to set a budget for the ad sets?
Thank you again
You can set minimum and maximum budgets 🙂 I’m glad you enjoyed the article!
This was a wonderful article and it resulted in me having the guts to do CBO.
I’m on my third day in and $300 in this hole. However I figured collecting the data comes with a cost and the algorithm will get smarter, so I will patiently wait. Moreover my store developer didn’t set up my payment on the store, not did he tell me, so there were 18 checkouts that couldn’t proceed further.
Nevertheless, my question to you is in regard to the look a like audience.
What did you create it based on? Conversion, initiate checkouts, certain percentage of video views? I figured I can use the data from the first three days and create a look a like audience that I will run another CBO campaign.
Hey Sam! I’m glad I was able to give you the little nudge to get it started! First things first, make sure you get your payments set up because you’re wasting money running ads to a store that can’t make sales. As for the LLA, any of those options work 🙂 Testing it out will be your best bet.
So, should we watch HCOMM’s course over the ad section with matt Schmitt on how facebook advertising is now? Or not
Yes, familiarize yourself with the basics of Facebook advertising, we’re currently in the process of updating that section of the course to introduce CBO 🙂
Such a great article! How many ads creative or ads should you have in each ad set in a CBO campaign? Did you the same ad for all the ad sets?
I’m glad you found it useful Mika! We recommend 4-5 ads within each adset and different variations of elements between the ads 🙂
Amazing stuff here..
So I’m testing CBO right now and right off the bat it did AMAZING. Day #3 hits and it did horrible.. However, I decided to increase the budget by 20% on Day #3 because Day #1 – #2 I was very profitable.. Now i sit and wait.. is it pretty normal for CBOs to just have completely bad days to recharge itself in a sense?
Hey Raden! It’s super exciting to hear that you’re already starting to test CBO 🙂 From what we’ve seen with our own account, there are ups and downs, however, increasing the budget will give Facebook some extra room for placements and optimizations!
Here again, question in regards to increasing and decreasing the CBO budget. I know its mentioned that it does not mess around with the algorithm and optimization of the ad sets within. Is that true? and if not, i heard from the grapevine that it does. The campaign I have that literally became profitable day one is slowing down on me and metrics have gotten expensive.. I plan to just leave it a lone and let it run even on break even days or under break even days. So far still above lifetime ROAS..
Hey Raden, welcome back!
The point of Facebook shifting budget to the campaign level is to help ad sets better perform as you increase budget since they can evenly distribute more budget throughout the various ad sets, instead of forcing just one ad set. This decreases the likelihood of budget increases “messing with optimization” so to speak. That being said, there are numerous other factors in play on any ad/offer… like audience sizes, product appeal, and so on which can still effect your overall performance…. and may cause what you are mentioning above. CBO is no magic bullet but it’s showing to be much more reliable on a macro scale 🙂
I’m actually finding more success incrementally vertical scaling..
Go with whatever works!
Good afternoon, Patricia. Are you optimizing for purchase or add to cart with $150 a day? Thank you in advance. —Daniel
Hey Daniel! You optimize for purchases 🙂
What do you recommend for a budget for testing new product or if you’re not a an established store yet?
Hey Marcy! I recommend starting off between $50-$100 depending on what your budget allows 🙂
My business requires creating short term campaigns that run for just 4-5 days. Creating long-term campaigns do not fit into my business model.
Now, by the time CBO has enough data about my campaign and ad sets, it would already be time for me to stop my campaigns. So what would you recommend I do to make full use of CBO?
Since all campaigns will be forced to use CBO soon I recommend testing and seeing how you can make it fit your business
When you say set your min/max budget, where exactly? So if the max budget is $150 what would the min budget be?
So go to the “Budget & Schedule” section while setting up the ad sets, and then click on the link that says “ad spend limits to this ad set”, and then the min and max portion will appear. However, don’t mark the ‘max’ to the recommended $150 budget, I would portion the lowest to $10 and highest to cap at $75 🙂
What would you recommend for a new store that has not really had a successful campaign yet for this new cbo. Do I just start with it now or try to find winning campaigns with the old way while it is still available?
Now is the best time to start experimenting with CBO so I recommend diving right in, soon enough, you won’t have any other options. Start familiarizing yourself now and don’t start off with the $150 budget mentioned in the post – that’s for seasoned store owners, start smaller 🙂
do you know if HCOM will be updated with this and have additonal content due to that change?
Hey Andrej, HCOM will be completely updated with CBO information soon 🙂
Do you recommend this strategy for testing new product?
Hey, Bet! So, it would be best to familiarize yourself with the new system by using previously winning products and audiences, however, if you’re brand new to eCommerce you totally can because you will have to switch over no matter what starting September. If you are brand new, you don’t have to use such a hefty budget either 🙂
Do you use this method for testing new products as well? or is it just for the winning product?
Very soon this will be the only way you can run Facebook Ads. So we’ll be switching all of our campaigns to CBO soon
Starting in September Facebook will force accounts to switch
Got it. That makes sense. Thanks Devin. ????
When you say $150 budget do you mean $150/Day or for the month?
That is $150.00 Per day, keep in mind this is a tutorial for more advanced stores that already have winning campaigns
Thanks for the information. Because of this article I have gotten the guts ton push CBO. Though I’m still in the hole and not profitable yet, I figured I can use the data that Im obtaining, and make a new campaign based on that. ‘
I do have a question, for your campaign you stated you made a Look A Like audience.
Was this audience based on video views, conversions, add to carts etc. Thanks
So happy to hear that! The algorithm performs better after a couple of days. As far as LLA – any of those work 🙂