Facebook ads are hard. But there’s a process behind creating successful campaigns.
In this post, I’m going to outline the 8 step process that we used to take a marketing subscription/SaaS from about 100 paying monthly users, to more than 3000 paying monthly users using paid acquisition on Facebook.
Short backstory: The product was a monthly Instagram growth service. (Examples: graminator, simplygram); The same advertising dynamics will work for any subscription services/SaaS with a lower price point.
In the end, the campaign generated more than $1,100,000.00 in lifetime revenue from only $92,707 in ad spend:
For the 6-7 months that this campaign lasted for, we were generating about 500 new subscriptions (paying users) a month.
The customer lifetime value for the product was around $350 ($25 for the first month +$50/100 month after that) and the cost per new user acquisition was under $30.
Before the campaign, although the client had a product/service that the market wanted, the exposure that he was getting organically was too slow and limited for any significant growth so he reached out to us ( https://infoscaling.com/ ) for help.
After a few weeks of adjusting the funnel and testing new campaigns, we built a campaign that was able to acquire new customers at $25.08 per sale. (Which meant break-even in the first month and profit after that)
Here’s the process that you can use to replicate this:1) Clearly define the main target avatar and best segments of the market
One of the most important steps that define chances of success in paid advertising is defining who you’re speaking to (and who you’re going to target) in your creatives.
From analyzing our clients current user base, we found that the top-paying (they stayed the longest) customers were fashion and beauty influencers (and wannabe influencers) and adjusted the copy&creatives to speak to them. We also adjusted the targeting accordingly (you’ll see it in #7).
Targeting the audience (both with actual ad targeting and adjusting the creative) that is most likely to pay the most has an effect on LTV and subsequently ROAS (return on ad spend).2) Find the best converting offer/free trial model (crucial)
Although this business was founded back in 2015, it took almost 3 years to get it to actually produce any substantial revenue. One of the major reasons why it took that long was having an offer that the market didn’t respond to. The initial offer (free trial) was 25 free followers.
The conversion rate (from visitor to claim offer) was high – but the conversion to paying customers from that offer was terrible.
We first needed to find an offer that was actually converting to paid users at a high enough rate.
We got the client to switch the offer to a 2-day Free Trial .
However, using an Instagram growth service for just 2 days was too little time for the vast majority of users to see any growth of their accounts so they didn’t see any value in keeping the service.
Next, we tried with a 50% OFF for the first month – with a promise to get 250 targeted followers and reach 15,000 Instagram accounts. The visitor to conversion rate was much lower than in the first 2 cases, but people who bought kept paying. This was exactly what we needed. ) Invest in ad creatives production 3 years ago, you could use stock photos on Facebook and your ads would still work. This is no longer the case. You have to invest time and money in quality creatives (especially video) if you want your ads to work.
Here are the main types of ads that we test when we start working on a new project:
a) Animated videos — they’re expensive but work really well for a lot of products (eg. SaaS). We tested a few of those but the results were not satisfying in this case.
b) Videos recorded by a client or his audience – testimonials/presenting the product. These are easy and fast to create and work really well in a lot of industries.
c) Image ads – there’s an unlimited number of variations that you can try here (the more your budget allows you, the better. Split-testing is key)
We instructed the client to get 20 different testimonial videos from his customers and created 15 animated videos for testing.
The top-performing ad was a simple testimonial video recorded in a car. The testimonial was from a fashion influencer with a big following (implying that they can too get a big following) so it perfectly hit the target avatar, built trust and presented the offer. (You can see the video (and the other ones) in 5).4) Analyze data & adjust accordingly
After the first 30 days of the campaign, the campaign generated first 400 users and grew the MRR from $5,000 to $25,000.
We found that Instagram Ads generated 95% of total sales in the first month. Although the CPA was slightly higher than on Facebook, we couldn’t have enough traffic volume on Facebook in order to scale like we wanted to.
Also, most of the buyers were women between 25–45 years old ( that also influenced our decision on what testimonials to use in ads).
You need to constantly be monitoring the data and adjust your campaigns accordingly & make decisions based on the data.5) Optimize ad creatives
After testing the initial creatives and finding the best ones, you need to optimize those further (with the goal of reducing the CPA).
The hypothesis was, if we took the best performing video ad and add subtitles and the right text on the top of the video (classic Gary V style) — we’ll increase our CTRs, decrease CPCs, and ultimately reduce our CPA which started going up a bit.
We came up with 10 different headlines for the video and tested them all. Few of them worked really well.
We continued testing different variations throughout the campaign, both to decrease the CPA and have fresh creatives to counter banner blindness.
Here’s a lot of different video variations that we used: https://drive.google.com/open?id=13xv8WWTkUjiksYCuk28CB3cnRMW04SeQ6) Utilize conversion events, custom conversions, and the pixel.
Setting up the Facebook pixel properly on the campaign start is crucial to every campaign’s success.
The funnel looked like this:
Ad -> Sales Page -> Checkout Page 1 -> Thank You Page 1
Ad -> Sales page -> Checkout Page 2 -> Thank You Page 2
In order to be able to segment the audiences for retargeting and to track the success of the ads, we set up the basic FB pixel (page view) on all pages, and set up the purchase pixel on the thank you page to track purchases.
All the campaigns on the account were optimized by using conversion events and custom conversions. Generally, the event you want to optimize for is the event you want (eg. Purchase/Lead/Click).
The more data your pixel gathers, the better it can optimize in the future, which helps with long term campaign success.7) Scale with lookalike audiences
From a technical perspective, scaling your winning campaigns mostly depends on finding new audiences you can show your ads to.
When we start working on a new account, the first step that we take is creating custom audiences of people depending on what stage of the funnel they’re in (or visited; tracked by FB pixel).
The more custom audiences you create — the more relevant you can be to different audience segments within a funnel(by showing different types of content/offers depending on what stage in the funnel they’re in).
From custom audiences, then, we create lookalike audiences. The best lookalike audience you can have is buyers (purchases) lookalike.
Here are some of the lookalikes we’ve used in this campaign.
a) Purchases (7 days) -> Lookalike 1%, Lookalike 2%, Lookalike 5%, Lookalike 10%
b) Purchases (30 Days) -> Lookalike 1%, Lookalike 2%, Lookalike 5%, Lookalike 10%
c) Purchases (90 Days) -> Lookalike 1%, Lookalike 2%, Lookalike 5%, Lookalike 10%
d) Purchases (180 Days) -> Lookalike 1%, Lookalike 2%, Lookalike 5%, Lookalike 10%
As you can see, just by having ONE custom audience broken down into smaller audiences based on a specific time frame (7–180 days) — you’re able to create 20–30 different lookalike audiences for the scaling stage.
Also, since we knew from the data that the best-paying customers were fashion influencers, we extracted the emails from those buyers, created a custom and then lookalike audiences from that. This was one of the best performing cold audiences.
Later on, when you have a few thousands of actions (ideally purchases) recorded on your pixel — you can throw out the targeting and let the FB pixel to do the heavy lifting for you. This works best when you’re scaling worldwide and have a potential market of 10–20MM customers.8) Leverage warm & hot retargeting audiences (CPA Reduction)
Retargeting people and moving them from one step of the funnel is crucial to any campaign’s success.
In retargeting campaigns, we were getting better CPA on Facebook than Instagram (because we could retarget them on the desktop for very little $) and our retargeting ads were super relevant.
When we started scaling the overall CPA on cold traffic went up to $30-$35 per sale, so having a good approach in the RTG campaigns became necessary. (It always helps to do this right and we focus on retargeting a lot in all of our clients’ campaigns.)
The results for retargeting were 575 sales total at $17.51. This significantly impacted the overall CPA and kept it where it needed to be.
We found that 2 out of 10 custom audiences we were using were most profitable & brought in the most sales so we focused on those. (Instagram page engagers (30 Days) and checkout abandoners (Last 7 Days & Last 15 Days)) https://imgur.com/a/wV2Ttza9) Stay Compliant With FB TOS
And the bonus lesson from this campaign is… you should never build a business in a way that it’s totally dependent on other platforms. Although, having an extra $1,000,000 is not too bad IMO 🙂
The business was destined to be shut down sooner or later. I knew that one day Facebook / Instagram will crackdown on Instagram growth services – which happened in the end and the business was prohibited from advertising. The client also started experiencing problems with service delivery due to crackdown on bots.
We work with a lot of clients and being compliant with TOS is absolutely necessary – Facebook’s really sensitive and you don’t want to lose the ability to advertise. (Although FB advertisers will know that you’re never really safe)
I hope you found this useful and let me know if you have any questions in the comments.
F(AWXM) 