Posts

Facebook mobile.

[Case Study] Facebook and Instagram Ads: How to go from $0 to $43,000/day Ad Spend on Facebook in 60 Days

Facebook ads for mobile apps.

Crunch time. An App partner (iOS & Android) needed to get a massive volume of installs ASAP to support a critical holiday important for their annual revenue and new customer acquisition. 

Over 60 days we went from $0 to $43,000 per day ad spend on Facebook & Instagram. Along the way we helped the app achieve its best revenue day ever, tripled their Instagram following (a positive side effect of the massive spend, and engaging ad creative), and brought in millions in lifetime value (LTV), all while keeping cost per install (CPI) on target. 

We’ll be using “CPI” a lot in this article, so take a moment to sear its meaning into your brain before you read on. Say it with me: “CPI = Cost per Install” 

Here’s what’s included:

  • Overview
  • Campaign Setup and Results Summary
  • Days 1-30: Testing to Find Facebook Ads, Audiences, and Settings that Scale
    • Creative that Scale 
    • Audiences that Scale
    • Ads Settings that Scale
    • Tracking Troubleshooting
  • Days 31-60 The Ramp Up and Final Push
    • More testing; audiences, creative, and settings 
    • Push Budgets to Winners  
    • Event Specific Creative Push 
    • Emergency! Account Spend Limit Hit & Workaround 
    • Emergency! App Stability 
    • Final Push and Rising CPI
  • Summary 
    • Top Ads
    • Performance 
    • Final Assessment and Key Takeaways

Campaign Setup and Results Summary: 

App install campaign data from Facebook Ads.

  • The Client: iOS and Android App with in-App Purchases 
  • Primary Channel: Facebook and Instagram App Install Ads
  • Supporting Channels:
    • iOS App Store Ads 
    • Google App Install Ads 
    • Pinterest App Install Ads 
    • Email (To drive in-app purchases and user-adoption post install) 
  • The Geo: 🇺🇸
  • The Results:
    • 60 Days
    • $263,000 Spend 
    • 97,975 Installs 
    • $2.68 Cost per Install (CPI) 
    • 9.2m Facebook Reach 
    • 26.8m Facebook Impressions 
  • Best revenue day in company history 
    • Bonus Performance Metrics
      • 2.1m video views 
      • 54,000 post reactions 
      • 11,000 post shares 
      • 6,000 New Instagram followers
      • Multiple ad variations went viral 
      • Featured in the App Store

Days 1-30: Identify Facebook Audiences, Creative, and Settings that Will Scale 

After an initial planning, strategy, and goal alignment phase we jumped in. In order to protect costs and efficiency, we spent the first 30 days testing creative concepts including layout variations, and ad copy; we tested 68 audiences, and different bid optimization strategies. This section will get into what and how we tested in the first 30 days. 

Finding Creative the Scales

We went through years of creative in the Facebook Ads Manager and identified ad variations and their attributes of top past performers. We looked at metrics like click through rate (CTR), cost per install (CPI), ad quality metrics and more to identify top ads. We looked at the ads copy, creative, CTAs, and ad formats (single image, carousel, and video). 

This research revealed: 

  • Top formats: Carousel, Single Image, Video 
  • Top creative: was explanatory or had humor 
  • Top Audiences were list and event based lookalikes 

The research resulted in 12 ad concepts to be paired with the different ad formats and creative. We used existing client creative, and worked with their design team on new creative options. 

Creative testing leaned on Facebook’s algorithm to serve the ad variation that would work best. Ad sets typically had 2-6 creative variations, and the Facebook algorithm would move budget to top performers, but the Facebook Algorithm didn’t work alone. 

The Tuff team would monitor ads daily, and turn off any ad variations or ad sets that had high CPI. This would force Facebook to spend more on the other variations, which sometimes would go on to become winners.  

Audience Research

The audience investigation was similar. What audiences had the best performance metrics? Were they segmented by age, gender, or geography? Were there exclusions (users specified not to receive an ad, e.g., existing app users?) 

Though we had initial ideas on what audiences would perform well, seeing historical audience performance, gave us a running start on audience development. 

Audience targeting results on Facebook.

For early learnings we looked at Facebook Demographic reporting in the Ad Account Overview. By Looking at Mobile App Installs compared to Amount Spent, we were able to identify which demographics were likely to have the best performance. In this instance we’re looking at the ratio of mobile app installs to amount spent by age group. To note, this trend didn’t hold through the 60 day push, so it is important to continue to target all ages so opportunities for installs aren’t missed. 

Note: Audiences are selected at the Ad Set level in the Facebook Ads Manager. At times in this Facebook case study, you may hear audiences and ad sets used interchangeably. 

Generally, the Facebook campaign structure is as follows: 

Facebook ads campaign structure diagram.

In the first 30 days we tested 63 audiences, paired with a mix of creative. For our tests, we’d typically have 4-8 ad sets per campaign, and 2-6 ad variations per ad set. Over the course of the test period we would turn off underperforming ads and ad sets to push more spend to winning mixes of audiences and creative variations. 

Facebook campaign structure.

As Facebook Ads performance data came in ads and ad sets we’re turned off, signified in the chart above by “🙅‍♀️”. Only about ¼ of our ad sets survived month 1 testing, and 3 creative variations (of 12) rose as top performers. 

Days 1-30 Summary

TESTED

  • 21 Campaigns 
  • 63 Ad Sets
  • 12 Ad Variations 

RESULTS 

  • $9,821 Spend
  • 7044 Installs 
  • $1.39 CPI 
  • Three top creative variations identified 
  • 15 core audiences identified 
  • Cost-cap bid strategy identified as effective at this scale…
    …but would it last? 

Days 31-60 The Ramp Up

With top audiences, creative, and bid-strategies identified we were prepared to ramp up spend. On day 31 we had 3 campaigns live, 5 audiences, and the top three ad variations running. We spent $950 that day. From there we inched up budget at the ad set level daily, and by day 45 we hit the $2,000/day mark. 

More testing; audiences, creative, and settings

Although some audiences didn’t succeed in the 30 days, there was a chance it was due to the creative and ad settings mix. We ensured these previously tested audiences were given a chance to be paired with top creative and ad settings. 

During this period the client continued to develop new creative, and that was tested as well using the assistance of the Facebook algorithm. We also isolated some creative to guarantee Facebook spend. 

A big driving force behind this rapid Facebook Ads ad spend ramp up was a holiday important to our client’s business, and we wanted to ensure that creative specific to the holiday was served. We tested 10 holiday ad variations, and only one version took off. It would become one of our best ads for the entire 60 period, though it underperformed two non-holiday specific top ads. 

An interesting take away from the top three ads is that the event specific creative didn’t perform as well as more general app functionality ads. This may be due the holiday not being applicable to everyone who sees it. 

Facebook ad example.

Here is the top event-specific ad. It was a carousel and showcased the product offering. More on how they ad variation performed in the results and summary section. 

With top audiences and creative identified we were ready push spend way up leading up to the holiday, however, the the path to spend ramp up we encountered two speed bumps.

$5,000 Ad Account Spend Cap Hit & Workaround

Did you know Facebook has a default $5,000 ad spend cap per ad account, even on established ad accounts? That’s right! This spend cap is different from the ‘spend limit’, and can only be increased by Facebook Support. Unfortunately, support can have a several day turn around, which would be after our peak push. 

In this instance, we were able to find an immediate workaround that allowed us to keep advertising. We created new ad accounts. With each ad account having a $5,000 cap, we created 6 additional ad accounts that would take us up to the $35,000 per day in spend.

We did finally hear back from Facebook Support and get the spend limit increased, but this was a good stop gap measure to keep ads rolling. 

App Stability Issues

With this massive ramp up underway the app encountered some stability issues due to the influx of new users. This led to a temporary pause on campaigns while app capacity was addressed. 

The takeaway from this experience is that if you are undertaking a massive app install or web traffic push, have the app and website stability on your radar. How many users can your app support? Who and how will stability be addressed if there’s a crash?  

Thanks to the app’s solid team of developers, app stability issues were resolved quickly and the ramp up would continue. 

Final Push and Rising CPI

The final push up to $43,000 was made possible by attention to detail and willingness to sacrifice CPI for more spend around the important holiday.

Facebook CPI results for app installs.

As we spent more in the final days of the campaign, CPI rose as well. We followed our earlier approach of pushing budget to top ad sets and creative variations, and some of these groupings had literally hit their limit (See $5,000 account cap section above). 

The cost cap bid optimized campaigns also weren’t spending their full daily amount as Facebook struggled to find users at the cost and volume we had set forth, so most campaigns were switched to a lowest cost bid strategy, which contributed to rising CPI. Low cost bid optimization, however, increased volume and ensured campaigns would spend their set amount daily. 

Fortunately the increased CPI in this final push was acceptable because the volume of installs, and subsequent LTV supported the costs. 

Summary 

Top Three Ads:

Facebook ad results from installs.

Our top ad was a carousel. Carousels are known to do well on mobile devices, because they can be used to convey information without the bandwidth of video. This particular ad was also highly engaging and received thousands of reactions and 100s of shares over the course of the campaign. 

Facebook ad results from installs.

Our second best ad was an informal video shot on an iPhone, that shows the end product of the app (a physical card and gift). This video was paired with concise text and a headline on what the app does. The informal nature of the video may have been what made it resonate with people. 

Facebook ad results from installs.

The third best ad was event-specific, and also in the carousel format. It showcased end-products of the app, and also had the clear concise message about what the app does. Because it was event-specific, costs may have been higher, as it wasn’t relevant to everyone who saw it. All this said it was our third best ad of dozens of ad variations run over the 60 day campaign period. 

Results 

Facebook ad results for install campaigns.

Getting to $45,000/day didn’t happen right away. There was little visible activity in the first 30 days while the hard work of research, audience, and creative testing was executed. Increasing spend was only possible because of this critical foundation established early on. 

Spend peaked several days before the event this campaign was based around, after which we settled in at $1,000 – $2,000 / day spend. 

Facebook ad results for install campaigns.

Installs tracked closely with spend. This consistency was the key to our confidence as we pushed budgets up. 

Facebook ad results for install campaigns.

CPI was initially very high. To note the budgets on these learning days was very small. And CPI dropped considerably and stayed low once initial learning was complete. 

Key Takeaways to Ramp Up Spend Quickly on Facebook 

  • Test to identify top audiences, creative and settings 
  • When an audience fails, try new creative 
  • Be considerate of technical issues 
    • Is your site or app capable of the increased traffic from this kind of ramp up? 
    • Have you requested that the default $5,000/day ad spend cost cap be lifted on your Facebook ad account? 

Finally, with these principles in place, don’t be scared to push up ad spend.

Facebook has a propensity to spend what you give it especially with low cost bidding, but this spend isn’t always paired with the desired results. In this case, however, through the approach of finding audiences, creative, and ad settings that would scale, we were able to achieve a massive spend and results ramp up in a 60 day period. 

Advertising brand on Instagram.

A Crash Course in Instagram Audiences for Your Social Media Ads

Advertising brand on Instagram.

With a global following of over one billion active monthly accounts (as of June 2018), it’s no wonder Instagram (IG) advertising is such a popular way for brands to build awareness. But where does one even begin on this beast of a platform?

Like with any platform, finding your ideal audiences will be a good start, and segmentation is a very effective way to do this (and it can be easily done on Instagram!). Once you’ve narrowed these down, you can start to test and segment your audiences to determine the very best ways to target them with social ads. Finally, you’ll want to keep them engaged over the long term so that your brand can continue to thrive, grow, and prosper.

In this post, we’ll talk you through ways to build Instagram audiences for social media ads, plus share ideas of how you can put these audiences to the best use.

How can I narrow down my Instagram audiences for ads?

Most likely you’ve already defined your brand, but if you haven’t already, it’s the first thing you’ll want to do. A solid understanding of your brand or brands will give you key insights into who your demographic is so that you can start narrowing down your Instagram audiences. It’s not rocket science, but you do want to be strategic. Doing so will save you so much time when crafting your content.

By the way, your Instagram audiences ARE your target consumers. On Instagram. That being said, not all of your target consumers are on Instagram, and not all Instagram users are your target audience, so a little IG research is going to go a long way.

Assuming you know the who, the what, and the why behind your product or service, you can start searching Instagram for the brands most similar to yours. For example, if your company makes simple graphic design tools for users, then maybe you’ll want to search for brands like Canva, PicMonkey, and Crella, for example. Dig deep here and figure out how their followers are engaging with them.

During your research, see if you can gauge which users are actual customers of your competitors. It’s really easy for people to like, follow, and even comment on Instagram, but if you can determine which ones spend (or intend to spend) money based on how they’re engaging, this will be really useful data.

Speaking of Canva, here’s an IG post for a current campaign of theirs where they’re donating all proceeds from paid images to Australia emergency bushfire relief and recovery programs.

Instagram ad example.

Check out the user comment on the top right and compare her comment with the user’s comment at the bottom with the clapping emojis. While there’s nothing wrong with claps because they still signify engagement, the user at the top is indicating more interest. Of course, there’s no concrete proof that she’s a Canva user, but it appears she has an affinity for the app. The more intel you can gather on users like her, the better. Take note, as these will be the types of followers you’ll want to go after (in your respective niche, of course).

What are some ways that I can test my audiences on Instagram?

Once you’ve narrowed down your market and gotten that part out of the way, the most daunting part of building up a following is done. The fire is lit, so what’s next?

For starters, you’ll want to determine how your followers are engaging with your Instagram content, and analytics tools are great for this. There are quite a few effective IG analytics tools that are totally free and there are some great ones that you can pay for as well. It really just depends on the types of features you want. There’s nothing wrong with reinvesting a little cold hard cash back into your brand, but it’s also nice to save those dollars from time to time. Try mixing it up.

IG analytics tools can give you actionable insights about your followers’ behavior –– information like when they’re online, what they want to see in their feeds, and what types of products they’re buying. This is going to help you to really target your audience which, in turn, will help you allocate your marketing budget more effectively.

Here are three of our favorite free and paid Instagram analytics tools:

  • Sprout Social: Sprout Social’s data-driven IG analytics tools is one of the most powerful analytics tools out there. It lets you do things like identify top posts during specific times, schedule and streamline your posts, and monitor hashtags. Plus, the reports themselves are very clean and easy to read.
  • Instagram Insights: IG Insights is, of course, a freebie and you definitely don’t want to underestimate its robustness. After all, it’s got the insider’s inside scoop on what’s happening with your audience, and it’s a very user-friendly tool as well. You can analyze your followers and access your precise demographics as well. Like Sprout, the reports are also easy to read, and the metrics it provides are super valuable.
  • Union Metrics: Union Metrics offers a full suite of tools for a fee, but they also have a free Instagram Account Checkup Tool. With the checkup tool, you’ll have access to detailed reports that analyze your most dedicated followers, average post engagement, and show you what your top hashtags are. It allows you to easily understand your metrics so that you can generate more effective campaigns.

Companies like Socialbakers, Iconosquare, and Squarelovin also make great analytics tools. The best way to test your audience is to switch up your ad content as often as possible and to try as many different tools as you can.

Another fantastic way to collect valuable data on your audiences is by taking advantage of the Instagram stories and polls feature to learn more about their likes and interests. The more data you’re gathering in different ways, the more you’ll be able to spot trends with accuracy so that you can really finetune your targeting tactics.

How should I segment my Instagram audiences?

Once you’ve uncovered some key trends within your audience, you can then start showing relevant ads to different segments of your IG followers. The reason why marketing segmentation is so important, in general, is that you can maximize your marketing budget and resources by targeting them more effectively. So, let’s get to segmenting on Instagram, specifically, and show you how to do this effectively.

First, if you’re using IG’s hyperlinked hashtags feature, this can be a great place to start. Presumably, by now, you’ve been doing your homework and testing your audiences, so you know who’s clicking on what. Your hyperlinked hashtags can give you some of the most valuable information you’ll collect on your followers because it tells you precisely which products they’re browsing. It’s also something that brings them one touchpoint closer to a purchase. If certain audience members are clicking repeatedly on specific hashtags, you now have some valuable information you can use to segment.

Generally, segmentation can be broken down into four main categories:

Instagram audience segmentation.

  • Geographic: Geographic segmentation lets you target your customers based on where they live. It’s effective because audience interests, values, trends, and preferences are going to vary in different cities, states, countries, and regions.
  • Behavioral: Through behavioral segmentation, you’re targeting your audience based on things like their purchasing habits, how they go through their decision-making process, and their attitude toward the brand.
  • Demographic: Demographic segmentation is the most popular segmenting strategy among marketers. It involves the process of dividing your followers up based on variables like age, gender, family size, and income, among other things.
  • Psychographic: Psychographic segmentation can be a little more challenging because it’s highly subjective, but it’s also very effective because of this. You’re targeting based on intrinsic traits that have to do with your respective audiences’ values, personalities, interests and opinions, motivators, and lifestyle choices.

If you’ve been in business for a while, but you’re simply trying to get more Instagram engagement (or you’re new to the platform altogether) the good news is that you probably know most of this stuff. Once you’ve figured out how to navigate Instagram and do the things we’ve already covered, like narrow your market and test your audience, you can segment based on what you already know about market segmentation.

After you’ve narrowed down your segments, Instagram lets you easily create custom audiences for your business by going to the “Audience” tab on your insights dashboard.

Custom audiences on Instagram.

Once you select “Audience”, you’ll be given the option to “Create a Custom Audience” from the dropdown menu. Follow the prompts from here to start creating your segments. Once you’ve done this, you can start using your marketing prowess to create ads that appeal to your different segments. Easy-peasy.

When is it time for me to test new audiences?

Finally, if your I.G. marketing strategy just isn’t working anymore, then it might be time to test out new audiences. Something to keep in mind is that demographics are always changing. New trends are constantly emerging, audiences get older, and populations become more racially and ethnically diverse. Remaining nimble and ahead of the curve will be a better tactic than simply waiting until your engagement is dropping, but if it does get to that point, try to act fast. Stay on top of what your followers are doing as well. All of this will help you to regroup and re-segment if and when necessary.

Other good times to test new audiences are in situations where you’re introducing a new product or service, or if you’re going through a company rebrand. If you’re already in the habit of constantly retooling your marketing message and adapting your product or service to suit a wide array of demographics, then an adaptation marketing strategy can also be very effective. When you have a business that experiences steady growth, then it’s a near guarantee that your audience is going to evolve and change, so just be prepared.
_____

Understanding your Instagram audience is the best way to create growth strategies that work. The platform has the highest average engagement rate of all social media platforms, so it really is one powerful tool. Use these tips and tricks to make the most out of it so that you can start segmenting effectively and continue watching your brand soar to new heights.

Got more questions about finding your perfect audience using segmentation? We have answers. Give us a shout at hello@tuffgrowth.com or schedule a 30-minute strategy call with our team here. We’re looking forward to chatting with you and helping you plot your growth.