So many of the clients we work with tell us they’ve lost piles of money trying to scale their businesses using social ads. 

And that they’ve lost hope in agencies helping them, because agencies make big promises and never deliver. 

But it doesn’t have to be this way. 

It’s possible to scale your ecommerce business using social ads, and to achieve higher ROIs than you’ve ever seen before. 

It’s possible to grow revenues quickly too, it just takes a little bit of strategy.

In this strategy breakdown, you’ll learn: 

  • How we took a well-liked swimwear brand from $200,000 to well over $16 million, with an ROI of 6.6x. 

  • How to create a bespoke strategy that works for your business. 

  • Successful methods for generating revenue online.

  • How to increase ROI and ROAS

  • Advanced strategies for scaling an ecommerce business of any kind.


The case study. 


Our client is a swimwear designer and manufacturer, selling affordable swimwear. They were one of the fastest growing independent swimwear brands. 

But when they came to us, they were ready to give up on paid advertising. 

After spending up to $500 a day on social ads, and only making back the same amount (sometimes less) they couldn’t stand to lose another dollar on campaigns that weren’t converting. 

They had invested heavily in influencers and ads, but couldn’t find a way to make the strategies work. They couldn’t create a profitable model where they could put money into ads and generate more revenue than they spent on ads. 

After trying six agencies in the last two years that all promised the same results and never delivered, they were tired of huge agency bills that never amounted to any actual results. 

This burned through their cash-flow and put their business at risk. 

The client didn’t understand much about marketing, but they understood that marketing involves the entire user journey from end to end, not just running ads and hoping for the best.

They didn’t need media buyers or an agency that just executes campaigns for them. 

They needed a full-stack marketing agency that could handle every aspect of their marketing funnel. 

And they needed someone that could give results. Fast. 

You see, success with social marketing isn’t just about creating a campaign on Facebook and hoping for a miracle. It’s about evaluating and strategizing based on the entire user journey from first impression, to the sale or conversion, and all the way to understanding final lifetime value.  

The brand’s organic reach on Instagram was great, they had highly engaging content with strong social proof, a great repeat customer base, and great product. 

But they had but very poor ad performance.  

We knew that there was massive amounts of potential. We just had to convince the client that there was value in paid advertising still. 

 After two weeks of discussions and strategizing, we finally got to work.

So how did we do it? 


Step 1 – Discovery Mode 


Before we create strategies and execution plans, we have to start with a discovery phase. This includes a full analysis of everything that had been done in the past, and a full exploration of the brand. 

In just 30 short minutes with our client, we found out… 

More than 33% of ad spend was on boosted Facebook and Instagram posts. 

From a business owner’s perspective, this might seem like a good idea. Boost your posts more, and you’ll get more reach. The more reach you get, the more sales you’ll get. right? 

But it doesn’t actually work that way. 

Boosting posts will help more people see your content, you’ll end up with more likes and comments, but will this translate more sales? Probably not. 

Rather than reach, we needed conversions. 

Facebook tracks everything that you do on and off Facebook. Every purchase, every event, every click… 

Users are segmented based on actions they take, what they buy, what they click. 

Essentially, there are three kinds of users… 



  • Engagers are people who will engage with ads by clicking, sharing, or commenting. 
  • Clickers are people who are likely to click through to your website from an ad. 
  • Converters are people who buy or sign up. They’re the most valuable. 


You will notice that there is a small overlap between the groups. This means that someone can be in more than one group at a time. How is that possible? 

As a converter, users likely have specific times during the day or during their session where they are more likely to make a purchase or sign up, while other times they might just be clicking or liking posts.

By boosting your posts, you only put your ad in front of engagers that are not likely to buy. 

What’s worse? The client had a retargeting campaign running to retarget all people who engaged with their boost page posts. 

Why is it bad? Because they ended up paying twice to reach a user who is not relevant and is not likely to purchase.

They also brought a lot of users to their funnel, only to later spend money on them with email marketing, promotions, and other paid campaigns. But they didn’t see much success. 

The client had a failed remarketing system. 

The worst thing that can happen in a remarketing campaign is that you treat all customers alike.  In their existing campaigns, instead of segmenting users based on recent visits or actions taken on the website, (for example, segmenting people who added to cart or viewed a product page), they treated all users the same. 

Hot leads who visited the website yesterday were lumped into the same audience group as people who visited their website 9 months ago. 

Obviously a user who visited your website yesterday and added an item to their cart has a higher chance of converting than someone who visited your website 3 months ago.

Paying the same for both users, and prioritizing both as if they were the same is a costly mistake.

Another issue we found was that remarketing was set to all website visitors. One of the biggest mistakes when setting up remarketing campaigns is including all available audiences without adding actionable valuable conditions, like viewing a product page, adding an item to the cart, or initiating checkout.

When your remarketing strategy targets all users, you are forgetting that 40-60% of users bounce off the website and are not interested in your product. In addition, even if they didn’t bounce off, but didn’t get through the first step of your funnel, are they really worthy of being retargeted? Especially when it will cost you money?

Not identifying top products to use in campaigns. 

We all want to think of Facebook as a place that we can just dump our entire line of products and let Facebook deliver all of them and get some decent results from each. In reality, things work different. 99% of the items being advertising on Facebook perform poorly. Only 1% off products are defined as trending products and that users care about. Yes! Facebook is not a place to run a clearance sale or to get rid of stock. Facebook works only for trending products with a very high conversion rate. Traffic is expensive and only top performers will work in creating profitable campaigns. Our client wasn’t aware of the analytics insights that are available to evaluate products with most potential. Spoiler: Your top sellers are not always your most profitable items or the items with the most potential.  

$0 ad spend on existing repeat customers. 

Existing repeat customers are like gold to an ecommerce business. Existing customers have a staggering conversion rate, usually around 3-6X higher than prospecting visitors. 

If you have a strong brand with a customer base that has shown loyalty, why wouldn’t you spend money marketing to those people? You know they’ll buy. 

Many brands think that if someone is a repeat customer, you’re wasting money advertising to them because you can just use free methods like email marketing, but the reality is that most people spend a lot of time on facebook. You should advertise to them where they are. 

Not all customers are the same.  

One common mistake that we found across many different campaigns for this client was that all ads looked the same. This means that existing customers saw the same type of ads that prospects saw. 

Someone who has never heard about your brand and a loyal customer shouldn’t be targeted with the same ads. Why?

Your existing customers already know about your brand and product. Show them the things that they want to see and buy.

Prospective customers don’t know anything about your brand. Show them why they should be interested in your brand or product. 

Using a generic campaign structure.

Instead of building a strategy around their strong selling points and strongest channels, our client used some generic strategies that were offered for grabs by online marketing “gurus”. 

The problem with that is that instead of using their strongest creative and strongest online assets within their ads, they picked similar examples of high converting ads and tried to make their ads similar. 

Every brand and every product is unique. Similarly, every customer that visits your website or sees your ads is unique too. 

A generic campaign structure, and generic creative is just a waste of time and money. 

Ads that looked… like ads. 

It’s 2019 and we’re all bombarded with ads every second of our waking lives. As a result, we’ve learned to spot ads and ignore them quite well. 

The best ads on facebook don’t come off as ads, they look like organic content that grabs the users interest. 

In our client’s case, they featured generic ads that lacked their most important thing: their product.  The thing that made them money was nowhere to be seen in their ads. 

Long story short: people want to see your products, not your ads!

Using product shots rather than lifestyle photography.

Our client’s greatest asset was their killer lifestyle photos on Instagram. That’s what brought them the insane amount of followers and incredible organic traffic. But instead of using these images as ads, they used their product photos with studio lights and white background. 

Lifestyle photography almost always performs better on social media, because that’s what people are expecting to see. Not studio-shot ecommerce photos of your product. 

After we completed the discovery phase, it was time to move onto the next step. 


Step 2 – Strategy Planning 


After we finished discovery mode, it was time to create an effective marketing funnel that would utilize their strongest assets, highlight and promote their top products, and create a time-proof paid marketing funnel that would generate positive long lasting returns. 

A funnel that will not only drive massive revenues and profits, but one that will constantly evolve and get better and better as we moved forward 

Identifying products MVPs. 

If people only care about our 1% of trending products online, we don’t need to waste money promoting products that are not likely to bring positive returns. We want to promote the top 1-5% only. Our client’s website had 1200 product variations, so it was definitely a challenge to spot the MVPs, but we did it. 

In order to evaluate the best products, we evaluated the the following metrics: 


  • Top sellers on shopify analytics
  • Products with highest buy to detail rate and add to cart rate on google analytics 
  • Products with best performance based on Facebook product ID
  • Products with the highest organic reach and engagement from Instagram posts 

Let’s take a look…

Shopify top sellers 

Buy to detail rate on google analytics

Performance based on Facebook product ID

Instagram insights

After we cross checked all 4 analytics sources, we were able to gather a list of top 50 products that were likely to be that top 1-5%, and would be the best products to promote. 

In order to keep our strategy performing well, this was a dynamic process that changed weekly based on trends, supply, and ad performance. 

From there, we strategized further… 

Strategic placement. 

Because Facebook ads runs automatically both for Facebook and Instagram, many advertisers think they are the same. They think that whatever works for FB should work for Instagram. 

The data showed that this brand was doing well on Instagram, and pretty average on Facebook. But they were spending huge budgets on Facebook. 

When we analyzed the results, we found that the ROAS on ads on Instagram was 2.0X. Facebook ROAS was only 1.3X. 

We decided to run all ads exclusively on Instagram as this brand pretty much grew organically on Instagram and all creatives are native to instagram in terms of effects, colors and overall style. 

Creative style. 

Instead of using generic product photos from the client’s website, we decided to use only what made them known and was proven to work: their amazing Instagram photos. 

We used their untouched, unedited Instagram photos as ads, so viewers would feel like the ads were organic to the Instagram platform. 

Audience segmentation focus. 

Knowing that our client had more than 50+ purchases per week that were being tracked through pixels, we decided to avoid interests targeting and lookalikes, and segmented strictly based on the following audiences: 

Existing core base 

This included lifetime customers, repeat customers, and customers who made more than 5 orders in their lifetime. 

Followers & engagers

People who engaged with Instagram posts in the last 30, 60, and 365 days. 


We segmented remarketing by people who added to cart/viewed product in last 3,7,14, 30 & 180 days. 

These users were segmented into ad sets using the CBO method to let Facebook shift budgets between audience groups based on performance and goals.

Cold traffic 

We used pixel targeting which meant zero barriers and interference with pixel by using targeting or lookalikes. We only selected the appropriate geo-locations, age groups and gender. 

Once we finished the discovery and strategy phases, it was time to execute. 


Step 3 – Executing The Strategy 

When it came time to execute, we created the following funnel for our client: 

It might look a little complex but it’s actually quite simple. 

Assuming that the two strongest assets of our client’s business was their Instagram presence and customer base, we wanted to capitalize on that opportunity and use it to our advantage.

We didn’t worry about how many people saw the ads, or how many people we reached. We focused on how much money the campaigns generated. 

If it brings in more revenue, it’s more valuable to reach one person 10 times than it is to reach 10 people one time.

And this worked for us. The results were extraordinary. 

We focused on the number of times we reached a warm visitor, making sure our posts reached them. But not just any posts, it was the top 50 items that we carefully selected and that were updated monthly.  

Every post that was uploaded to the instagram page was blasted to all the audiences in the funnel and conversions started coming in at a shocking rate.

But that wasn’t it. We didn’t want to settle just for our Instagram funnel. We wanted the use Facebook machine learning to our advantage too. 

Creating the data snowball effect.

Every ad set on Facebook has a learning system. The more data an ad set has, the better it will be able to accurately predict possible opportunities and deliver steady results at scale. 

When we created new campaigns, our goal was always to find winners and scale them to a point where they get 100-200 conversions per day. 

Once we had a winning ad set with massive amounts of conversions, the results got better and better. We called this the snowball effect. 

Our goal with every new campaign launched was to find at least one or two ad sets that would have the snowball effect. 

With digital marketing, efficiency is key. We prioritized a single ad set with 100 conversions rather than 10 ad sets with 10 conversions. 

See here how our top campaigns really had the snowball effect…

Using Facebook machine learning to achieve over 19x ROI. 

We knew the machine learning power of the Facebook platform yields great results, so we used catalog ads to completely automate and dynamically show the right products to the relevant user.

This allowed us to use data driven product recommendations and show those relevant ads to the users who showed interest, but had not yet purchased.

And although our funnel was running only on instagram, we left the door open with remarketing campaigns for auto-placements, in case our users also browsed on Facebook. The algorithm would find cheaper opportunities for ad placements than on Instagram. 

The results? More than $74,117.27 spend with an ROI OF 19.22X on one campaign. 

By using our instagram funnel as our core ads  and the dynamic catalog ads to wrap up the deal, we were able to get the best of both worlds. The best of human creativity for selecting which products to feature, and machine learning to show target the user at the best time with personalized product recommendations based on AI. 

Saving time, maximizing opportunities with campaign budget optimization

Campaign budget optimization might be an overrated subject, but we were confident that by using it, we could leverage machine learning to find the best budget balancing between ad sets, and allow us to scale faster. 

Instead of wasting time on selecting and optimizing budgets for each ad set, we used budget optimization with manual bidding to give the system more freedom to work and shift budgets and spend across ad sets. 

Bid caps

We used bid caps on every single campaign we launched to ensure we can sleep at night while we have campaign budgets that are up to $10,000 per day and ensure that Facebook only spends the budget when it predicts it will be able to deliver the results that meet our goals. 

Generally speaking, our bids for existing customers were much higher as the audience size was limited and reach was limited as well. In this situation, high bids gave us better ROI. 

Never analyze by one day. 

One deadly mistake that advertisers make is that they analyze campaigns by one day of performance. But it’s impossible to know if a campaign will perform well in just one day. 

Facebook tracks purchases for 28 days from the time of click or view, and they do that for a reason. Most users don’t convert right away because they are not searching for your brand. 

Optimizing bids and budget by 7 days allows you to take advantage of gathering more data and time thus letting your campaigns perform effectively. 

And each day of the week differs too. We only analyzed and made changes to our bids and budgets based on how the ads performed in 7 days. 

Monthly re-creation and funnel improvements.

Obviously, not everyday was amazing and some days needed new creative and new content. 

We made changes monthly based on performance campaigns we created, number of posts per ad set, and budgets for each audience group. 

This was a working process that evolved from month to month and changed based on the algorithm learning, as well as our understanding of what worked best for this brand.

Results after the first 90 days… 

Results after 24 months… 

This strategy led to our client going from a small brand with potential to an amazing ecommerce success story. We were honored to work with this client, and to help them see such great results so fast. 

 The Final Word 

There’s no one-size-fits all when it comes to marketing ecommerce businesses. 

As you can see, it takes a lot of thought and planning to create a high-performing paid ad strategy. Every brand is unique, and every product is different. Every customer is also just as unique, and if you forget these things you will simply waste your time. 

If you found this article helpful, please share it via social media.