There are multiple options to build your following on Instagram with ads, but which one works best?

Well, I recently tested a slew of setups to find this out for myself.

Using five different organically posted reels on my Instagram profile, I tried a variety of setups to gain new followers and generate more views for my content.

Here’s how it played out.

The setup

To start, I ran five different campaigns with four unique targeting options.

This means two of the five campaigns had the same settings, but we’ll get to that.

Each campaign spent $1 per day over three days for a total spend of $3 per campaign, with roughly $15 spent overall.

Furthermore, each campaign used an existing reel on my profile to generate profile visits and, hopefully, followers.

And lastly, I staggered the start and end dates of these campaigns over the span of one week so I could get some semblance of separation in the results.

For the first campaign, I used the boost feature inside the Instagram desktop app and stuck with the settings the platform recommended.

This meant using the suggested audience feature that “targets this ad to people similar to your followers” without the ability to specify anything else, including location.

Given that most of my Instagram followers are from the US and are generally music makers and fans, this felt like an easy call.

For the second campaign, I once again used the boost feature but, instead, created an audience targeting fans of a collection of DSPs in the US, UK, Australia, and Canada.

For the third campaign, I boosted a reel to fans of music and recording in that same set of countries.

For the fourth, I created a traffic campaign inside the ads manager that optimized for Instagram profile visits, targeted the same set of four countries, and used an existing reel as the ad creative, effectively boosting a post once again, just in a different way.

The difference here though, is that I targeted a lookalike audience built from my Instagram engagement over the past 365 days.

And as for the fifth campaign, well, we’ll cover that one below.

​

The losers

Out of the gate, the first campaign did not go as expected.

When I checked in after the first 24 hours, I had a lot of new plays on my reel (and even a lot of likes and profile visits to match) but no new followers.

When I dug into the details, I came to understand the ad was only shown to users in Nepal, which I found to be… odd.

Until I remembered that a looooong time ago, I tested a Facebook page-like campaign (we’re talking like 10 years ago) and accidentally ended up with a bunch of followers from Nepal virtually overnight.

We’ve all learned the hard way at some point, am I right?

So, as it turns out, when Instagram says it “targets this ad to people similar to your followers” it’s talking about followers on Instagram and Facebook, not just the former.

And it seems to just turn toward the cheapest cost per result (no surprise there).

But I wanted to test that this wasn’t a fluke, so I decided to boost a second reel with the same settings just to be sure, except this time I boosted it from within Meta Business Suite instead of Instagram.

Hence the fifth campaign.

Unfortunately, the results were the same, so my assumption was confirmed—a ton of views and likes but no followers.

For context, you can see just how many more views these two videos got compared to the other three videos I used as ads over the same week (44.5k and 33.9k vs. 676, 518, and 642, respectively):

As much as this was a bummer to discover, it’s a good lesson learned that you really need to be thoughtful about who you target with every single ad campaign, otherwise, you risk nuking your data and will have to work hard to clean it up over time.

I already put in the work to do that on Spotify.

Now it looks like I’m going to have to do it on Facebook and Instagram too.

C’est la vie.

​

The winners

Fortunately, all was not lost, and I did discover a setup (or two, really) that worked.

First, the runner-up was the Instagram-native boosted post that targeted fans of music and recording in my primary four countries—the US, UK, Australia, and Canada.

While this campaign didn’t generate anything close to the number of views that the previous two campaigns did, the click-through rate was much higher (3.92% compared to 0.25% and 0.38% for the two losing campaigns) and it earned more new followers too.

And finally, first place went to the campaign I built inside the ads manager, my guess is due to the lookalike audience targeting.

This campaign had the highest CTR (5.25%) and the highest number of profile visits (12) in the primary four countries (as shown below):

It’s tough to attribute an exact amount here because I was overlapping campaigns, but I gained 24 new followers throughout the week with the highest daily numbers coming while both of these winning campaigns were active: 4, 4, 5, 4, and 4 per day over five days.

The results dipped considerably before and after that while the two losing campaigns were the only ones active.

And as for the last campaign—the one targeting DSPs in our primary countries—the results were mostly just forgettable.

High CPR, low number of profile visits, and a sort of “meh” CTR.

My best guess is the targeted audience and the content of the reel just didn’t match up.

Honestly, I wasn’t surprised about this one.

​

The takeaway

So what did we learn here?

First, the seed audience of followers on Instagram (and Facebook) makes a massive difference in the efficacy of this kind of thing, especially if you’re relying on Instagram’s suggested targeting features.

Second, assuming your audience is good, the in-app boost feature is quite effective for generating Instagram profile visits and, in turn, followers.

And it seems to be great for views too if that’s what you’re after.

Third, crafting a follower campaign inside the ads manager can be incredibly effective if you know what you’re doing.

But the true magic is most likely found in stacking features and objectives to deliver a multi-faceted result—boost a reel for views to one audience, leverage more defined custom and lookalike audiences for followers, and rinse and repeat.

But more than anything, as you and I both know, the content is what matters most.

It’s difficult to run a completely unbiased split test with this type of thing because I used five different reels.

Some were pretty good.

Some were just ok.

And that, no doubt, skewed the results.

The content matters.

And starting with higher-quality content just makes everything much, much easier.

​