Jump to content

Project Verum: The Foundations is now available! projectverum.org

Andy Shwetzer

$500K Selling A Posture Brace With Facebook ™ Ads (Verum Video Summary)

Recommended Posts

$500K Selling A Posture Brace With Facebook ™ Ads [Drop-shipping Case Study]


This is a case study on a recent store of mine that focused on selling a posture brace.


This will focus on the ads strategy side. I'll post another with the product research side for those who are interested.


Our numbers:


$528,000 in sales after accounting for returns

$300,000 in ad spend between the two ad accounts

$79,000 in COGS


This left us with roughly $149,000 in profit or 28% profit margins. Not bad for two months on one product.




This strategy is broken down into 5 easy phases that will work great for anyone trying to run a 1-product store. As a bonus, I’ll also be going over how to get around Facebook disabling ad accounts because that’s what happened multiple times while running this store.


But before we get started with Phase 1, let’s go over the creatives or video ads we used.


What we did with this ad was first come up with the basic format. Starting from a spin around, going into an animation, and finally ending with the benefits & features.


Along with the standard sequence, we came up with a caption that we pretty much used for all of our tests. The ad focused on convincing people that the brace is effective for good posture, it doesn’t waste time telling them why good posture is important because people who need a posture brace already know the importance of good posture.


Once we had our standard sequence, we came up with a few different intros and used them to start the sequence.


To find the clips for this ad, we took videos from Aliexpress listings and also used a few clips from dropshipping videos. It’s pretty easy to find old dropshipping ads on youtube if you search for your product.


You should be careful taking videos from people though, and use common sense to avoid getting in trouble. Don’t take videos from popular youtube videos and make sure to come up with your own content once you know something is working.


Another great option we use for original video content is Fiverr. Not only can you get people to demonstrate your product, but you can also get 3D animations done for pretty cheap. If you’re looking for models, look up “product video” or “product model”, and if you’re looking for product animations look up “3D product animation”.


Now, before you start Phase 1, make sure you have 1 basic video sequence with three different intros. When you’re picking out your intro, make sure it’s something eye catchy that will stop someone from scrolling. If you need inspiration, just look at the ads on your newsfeed and pay attention to what gets the most views.



The campaign and adset structure for this phase and all of the phases that follow are very similar. Refer back to this point in the video when you’re setting up ads for each phase and use these settings.


All of the phases use campaigns with buying type set to auction and objective set to conversions. We turn campaign budget optimization on and set bid strategy to lowest cost. The daily budget depends on the phase that you’re in.


For adsets, we optimize for purchase, and set new adsets to start the next day.


For location, we target our special ePacket list which can be found in the facebook group.


We let facebook decide the age and gender for the most part, but we will sometimes end up narrowing that down if we see performance for certain demographics underperforming.


When we’re starting ads though, we almost always just let facebook decide. Even if we’re selling a men’s product, we’ve noticed that Facebook can still find female buyers for us - especially as campaigns optimize. Language should always be set to English(all) unless you’re running ads in a foreign language.


For detailed targeting, or interest targeting as we like to call it, we make sure to exclude dropshipping as an interest. This is going to help hide your product and store from other dropshippers looking to sell the same product.


The interests we include change for each phase, and we don’t include any interests when we’re targeting Custom or Lookalike audiences.


We leave placements on auto, optimization on conversions, and conversion window on 7 day view 1 day click.


Using those settings in Phase 1, we set a daily campaign budget of $200. If that’s too expensive, you can use a smaller budget than we did, but in that case, just test less adsets.


We test 10 adsets with our $200 budget, so if you have a $100 budget, you should test 5 adsets.


For our 10 adsets in this campaign, we targeted interests related to posture.


The interests I used included Vertebral column, shoulder, neck, spinal cord, poor posture, cervical vertebrae, lumbar vertebrae, aging, massage, and old age.


Note: You shouldn’t think too hard about what to target, just pick something broad with at least a few million people that you think might work. Then use the suggestions feature to find more interests once you have a few in mind that you want to target.


For our initial tests, I started with poor posture which is an obvious fit for the product, and then used suggestions to find the rest.


When you’re going through these suggestions, make sure the majority of your audiences have at least 5m people. Some of my interests actually have over a hundred million.


The benefit of starting with broad interests is that they’ll give you much cheaper impressions and allow you to scale anything that’s profitable without worry about burning out your audience.


For each adset, we also set a minimum spend of $10 for each adset using our rule of distributing half the CBO budget evenly amongst the adsets.


Then we tested 2 different ads for each adset, but we actually recommend you to test 3 ads when you’re trying this. Since running this store, we’ve noticed that three has been performing equally as well while allowing you to test more creatives with the same budget.


If a campaign wasn’t profitable as a whole we’d normally kill it after 2 days.


When you’re analyzing your campaign after two days, you’re looking for any adsets that are profitable.


If your adsets haven’t even spent your target cost per purchase yet though, you should be looking for metrics like cost per add to cart that indicate it will be profitable later down the line. As far as your different video ads go, you want to look for the video ad with the best performance on Average Video Watch Time and Cost Per Link Click.


If you run this campaign for two days and find that none of your adsets are even close to being profitable, then I would scrap all of your creatives and start again from there. If all of your adsets are performing equally bad, it’s a good indicator that your creative is just not good enough.


If you test new creatives and they do bad again, it could be your product as well. The last hope in this case, would be to skip to phase 3 once you have enough data, but you could also save your money and start over again on a new product.


Anyways, once you have an adset and ad that you think are working, move onto phase 2. Don’t forget to kill your Phase 1 campaign if it’s not profitable as a whole.




In phase 2, we took the best performing interests and creatives to make another $200 CBO.


Since we had two winning interests, we duplicated each one five times to get a total of 10 adsets. We continued to set a minimum spend of $10 according to the rule from earlier.


Depending on the number of adsets you’re taking from Phase 1, you should duplicate each of them enough times to have roughly the same total number of adsets as Phase 1.


For example, if you found three winning adsets in Phase 1, you should duplicate them 3 times each for a total of 9 adsets in phase 2. It’s not exactly 10, but the goal here is to just get close.


You should also duplicate your winning video creative 3 times for each adset.


The idea behind this has to do with Facebook Artificial Intelligence and the differences in starting optimization for each adset, but that’s a topic for another video.


When we ran this campaign, our results were disappointing to say the least. In fact, it did worse than the original campaign, but that’s just how Facebook is sometimes. Luckily, the point of Phase 2 is just to gather enough data to move onto LookAlike Audiences in Phase 3.


Of course, it would be nice to make money at this point, but even if you don’t, you can just move right on to phase 3 as soon as there’s enough data. That’s exactly what we did here - just waited until we could move onto phase 3.


We recommend you wait until there’s at least 50,000 views, until you move onto phase 3. Having enough data is important for making a reliable look alike audience.


The general rule you want to follow in these first two phases is to collect as much data as you can while losing the least amount of money possible. Of course you can be profitable in these first two phases as well, but it’s also okay if you’re losing a bit of money.


Look Alikes will perform much better anyways so you don’t want to linger on these first two phases trying to make money through interest targeting.




Phase 3 is when we start testing look alike audiences and also start a retargeting campaign.


The first lookalike audience to test is 75% video viewers. You want to make 10 audiences for each percentile inclusive from 0%. That would be 0-1%, 0-2%, 0-3% and so on. You also want to make sure you’re selecting worldwide when you’re making those lookalikes.


While you’re making those lookalikes, make custom audiences to use for your retargeting campaign.


Make audiences from the last 7 days for Page View, Top 25% time on site, FB engagement, 75% Video View, Add to Cart, and Initiate Checkout.


Setup a new CBO campaign with a daily budget of $50 to start, and use the custom audience to make 7 adsets - one for each custom audience.


As far as ads go for retargeting, you should make some new ads. We like to use photo ads like this that have a really strong call to action like sale ends today.


After two days, you should start adjusting the retarget budget as needed. If it’s super super profitable, you can try to squeeze more out of it by increasing the budget. If it’s not doing so hot, you should lower the budget.


Try to find a good ratio between your retargeting budget and the budget for your other campaigns. Use that ratio to scale your retargeting campaign as you scale up your other budgets.


Also keep in mind that cheaper products will need less retargeting, while more expensive products may benefit from heavier retargeting. If your product is really expensive, you might even want to expand the time frame included in your custom audience.


After setting up the retarget, we setup a campaign with 10 adsets. Each adset was targeting one of the 10 look alikes we created before, and they all had 3 duplicates of our video. Unfortunately, the lookalike did not perform as well as we hoped.


Instead of giving up, we tested a few more videos until eventually we landed on our winner. If we weren’t so lazy, we would have tested more videos from the beginning instead of just hoping that the lookalikes would perform.


If you’re testing 75% Video Views and it doesn’t work well you can try testing other look alikes such as 95% Video View or Page View. You will need more data for these Look alikes though.


Eventually you should find a few good look alike audiences that are working well. At that point you’re ready for phase 4.




In phase 4 we try to scale the look alikes that are working well. This is when your results should start becoming a little more consistent.


So soon as we noticed that our 1% Look alike for 75% video view was performing well, we made a new CBO campaign with a $200 budget and duped the adset 5 times.


After a 2 days it was working super well, so we started doubling the budget every two days until it eventually reached $1500.


When you get a campaign with this kind of budget, results really start to even out as your campaign optimizes.


If you find that you’re not able to scale a campaign without performance tanking, you should try duping the campaign a few times instead. You can even dupe a campaign like 10 times if it’s working really well, but it doesn’t respond well to scaling the original budget.


We stopped scaling this campaign at $1500 though and just let it run as we moved onto our favorite phase. You should also move on from this phase once you’ve spent $5000. You can leave your profitable campaigns on from this phase while you test out the last phase.




In our fifth and final phase, we’re looking to scale sky high with broad targeting. This is what I like to call the “end game. “


Broad targeting is where you let Facebook decide basically everything besides the CBO’s budget and the ad’s creatives. As far as targeting goes, everything is left on auto pilot.


I don’t give Facebook any audience whatsoever - no custom audience, no look alike, and no interests besides excluding dropshipping.


For this type of campaign, I split it up into three adsets. 1 adset targeting the USA, 1 adset targeting Canada+UK+Australia+New Zealand, and 1 adset targeting worldwide with a few third world countries removed. 


I set a budget of $500 and just let Facebook do it’s thing. At this point, the only thing you have to do is adjust the budget.


After two days, I’ll make one of two decisions. I’ll either double the budget because it’s profitable, or I’ll kill it and go back to phase 4 to collect some more data before trying again. If it’s kind of in the middle, i’ll just let it coast another day or two until the decision is clear.


The reason why broad targeting is superior to all other forms of scaling is because:


It requires hardly any work to maintain. You don’t have to constantly go out there find more working interests. The size is basically the entire world.


It’s super easy to scale because audience size is massive, meaning that you can literally scale as high as you want.


You don’t have to refresh creatives because the audience size is so big. Eventually we did hit about 50M reach on one creative and had to refresh it, but up until that point everything was pretty smooth.


It optimizes over time to become better and better without you having to do anything. Meaning that once it’s optimized at a certain budget you can just crank it up, let it optimize at the next budget, and then rinse and repeat the process until you hit a point where your cash flow just simply doesn’t allow it.


Luckily in the case of Mr-Posture, the broad target worked super well. I ended up scaling it to $10,000 a day and letting it coast. It ended up spending $150,000 before Facebook decided to start giving us problems.


While we were coasting in Phase 5 letting the money flow in, of course we ran into some Facebook trouble.


This was during a period when Facebook was randomly flagging accounts like crazy. This is still happening to some extent, but during the summer when we were running this, it was like an everyday thing. Even before we got disabled for good, we were getting disabled like every day and having to submit appeals.


When we got disabled for the last time, we got around it with a new ad account under a new business manager.


If you ever find yourself in a similar situation, you should simply make a new business manager and ad account. Most of the time, you’ll be able to use your original page and website without issue.


Using our new ad account, we started right away with broad targeting using the original pixel. Since Facebook let’s you share pixels across business managers, it’s really easy to use your old data.


We just set the budget on day 1 and let it run until this ad account was disabled too. It was probably the easiest money we’ve ever made considering we just let it run on autopilot from the start. 


Again these were our numbers:


$528,000 in sales after accounting for returns

$300,000 in ad spend between the two ad accounts

$79,0000 in COGS


This left us with roughly $149,000 in profit or 28% profit margins. Not bad for two months on one product.


If we were going to do this again though, we probably could have done better. This is what we would have done differently.


After seeing the broad campaign do so well, we should have asked Facebook to increase our spend limit on the new account


We could have made another ad account after the second one was disabled


We could have incorporated some other products into the store that are related to posture correction

We also tested a bunch of random ad strategies like manual bidding that ended up lowering our profit in the end


Overall though, we were happy with the final performance of the store. I learned a ton about how saturation works in drop shipping, and it gave me the confidence to pursue other products with even more competitors and a longer history of being sold.


If you have any questions , drop them below. 

  • Like 1
  • Thanks 1

Share this post

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Create New...