Should I still be targeting my ads?

Niche products, ads examples, some good advice, and event invites.

Hello everyone, how are you doing?

This is the 5th episode of Snapshot, and it’s great to have you here.

Today I’m going to be sharing with you something that keeps coming up again and again.

If there are any topics you’d like to find out about, please leave a comment, send me a message, or some kind of smoke signal.

I’m writing all of this for you.

Let’s go.

---

Should you use interest targeting when your product is niche?

I get this question a lot, especially from founders building niche products.

Things like:
→ A virtual tabletop for tabletop roleplaying games
→ An app to help parents get their newborn baby to sleep
→ An app for caregivers of people with Alzheimer’s

Fairly niche, right?

And we’ve tested this across nearly 100 products.
And the answer is: don’t bother with interests.

Skip the targeting. Fix the creative.

---

When we’ve gone broad with the right message, performance has always improved.

One example:
We tested adding “Parents” to the targeting for an app to help parents build secure attachments with their children. Result? CPA increased by 23%. Uuuuum. We don’t want that.

Okay - there are some things we do cut on:
→ Subscription apps with account controls on 25+ where it makes sense for the audience (in 90%+ of cases), because people who are 25+ convert much higher from trial → paying than those younger
→ Gender, if the product is clearly for one (e.g. a financial education app for women)

But what we don’t do (not for conversion campaigns) is, say:

→ “Parents aged 25–40 who’ve engaged with baby sleep content”
→ “D&D players who like fantasy and maps and read books”
→ “Women 45+ with health brand interactions in the last 7 days”

---

Why not?

A few reasons:

1. The algorithm knows our audience better than we do.

In 2014, Stanford and the University of Cambridge ran a study that found that a computer could predict someone’s personality & preferences more accurately than:

→ A co-worker with just 10 Facebook Likes
→ A roommate with 70 Likes
→ A family member with 150
→ Their spouse with 300

And back then, the average user had 227 Likes. NOT MANY.

Now? Meta has data on 3.4 billion people, and it knows everything: what we watch, say, look at, like, share, click, buy.

So yeah. It probably knows who to target better than you. No offence.

---

2. Cutting your audience increases your CPM.

We see this again and again.

If you cut your audience down from everyone in the UK to just “interested in XYZ”, your CPM is going to go up. Sometimes 2–3x.

The same thing happens when you optimise deeper in the funnel.
Want to move from install to purchase? Your CPM will increase. Will your ads convert 2–3x better? Not always.

Same tradeoff here.

Will your niche targeting make up for the increased cost? Maybe. But probably not.

---

3. Fully automated targeting has been around for a while.

This has been a slow rollout over the years.
Google removed targeting from App Campaigns (then “UAC” for the real ones) about a decade ago when they plugged DeepMind into their algo. That same tech now powers smart bidding, asset generation, and full campaign automation across Google Ads (think PMax).

Meta started pushing this around 2020 with AAA (Automated App Ads). Now they plaster Advantage+ on everything, when it’s more like DisAdvantage+.

And if you’ve read the tea leaves, you’ll know where this is going. Here’s what Mr Mark said the other week:

“We’re going to get to a point where you’re a business, you come to us, you tell us what your objective is, you connect to your bank account, you don’t need any creative, you don’t need any targeting demographic, you don’t need any measurement, except to be able to read the results that we spit out”

Do I think they’ll actually get there in the next couple of years?
No. Not yet.

Right now, turning on A+ for copy gives you uncompliant messaging.
A+ for sound? Weird and icky.
A+ 2D/3D animation? Text cropped off the side.

The one thing that is working is A+ for audience targeting. That’s battle-tested now for over a decade, and for once, it’s nice not to be the guinea pigs.

---

4. Interests are now “suggestions”.

Literally.

If you open Meta and build a new ad set, you’ll see that interests, behaviours, and other targeting criteria are now labelled as “suggestions”.

That means the algorithm will take a look, and if it doesn’t think those suggestions are relevant, it will ignore them. You’re not cutting your audience. You’re just giving the algo a nudge. And it might knock your hand away.

Does it hurt to put them in? No.
Will it meaningfully change delivery? Also no.

---

So what does matter?

The creative.

That’s our biggest lever now. That’s how we signal who the ad is for.

→ Want to target moms? Say the word “moms” in the copy.
→ Want to reach people who play tabletop roleplaying games? Use language only a GM would recognise.
→ Want to find salespeople who drive to meetings? Show them in the car.

If your creative is specific, it will find the right people.

We recently had to pause work with a client because we were driving too many leads through their product. The ads worked too well. They weren’t set up to onboard at that scale.

They didn’t change their targeting. They changed their message.

---

APERTURE SESSIONS

We had our London Summer Party last night, so thank you to everyone who came down. It was great to see you 💕

London: We’re having our next one in September, so if you’d like to come, RSVP here.

Berlin: We’re celebrating our 3rd Birthday in Berlin in November. RSVP here.

---

Thank you, as always.

Have a great week.

Hannah

---

Thanks for reading. If you liked this post, you can get more writing from me by upgrading your access.