Content as Targeting

PLUS: Social Engineering, and Phishing/Vishing Using AI

Author

Startup Spells 🪄
February 07, 2025

Content as Targeting

The digital ad landscape changed drastically in 2020. Apple’s ATT (App Tracking Transparency) restrictions limited advertisers’ ability to track users.

At the same time, TikTok’s popularity surged, changing how people consumed content. Platforms shifted from relying on user actions like likes and shares (active signals) to monitoring watch time and engagement patterns (passive signals).

The Past - Active Signals

The Present - Passive Signals

How Content Became the Best Targeting Tool

“TikTok doesn’t need you to like or share a specific piece of content for them to know that you like it. All you need to do is just watch it all the way through.”

~ Hamza Alsamraee

These passive signals now provide platforms with thousands of data points per user session, shaping the future of ad targeting.

Active vs Passive Signals

Pre-ATT:

  • Broad Content

  • Narrowly Targeted

Post-ATT:

  • Narrow Content

  • Broadly Targeted

From Social Graphs to Interest Graphs

Older social media platforms like Facebook and Instagram prioritized social connections. Users saw content based on who they followed.

The Social Graph Algorithm

The Social Graph

TikTok introduced the interest graph, which ignores social connections and instead groups users by content consumption habits.

The Interest Graph Algorithm

The Interest Graph

With the rise of passive signals, these platforms now have thousands of data points on you with each user session.

Data Expansion via Passive Signals

This allows algorithms to serve relevant content efficiently, making traditional targeting less necessary.

Collaborative Filtering

The depth of content with short form interest graphs + The number of actions in products = content becomes self targeting ("my For You Page knows me")

Ads Function Like a Stock Market

Ad networks operate like financial markets. Advertisers bid for placement, and AI-driven optimization determines the best-performing ads.

“You have $1 billion of spend on Meta every year, but with ATT, you can’t target that narrowly anymore. It’s like trying to beat the stock market when you try to target narrowly.”

~ Hamza Alsamraee

Since traditional targeting is less effective, advertisers must rely on content to reach the right users. Instead of using audience lookalikes or retargeting, advertisers must create content that attracts the right viewers organically.

Self-Targeting Content in Action

Good content naturally attracts the right audience. A financial app, Rocket Money ran ads like “Watch me cut my monthly expenses in half.” instead of promoting general budgeting tips. This level of specificity ensures the ad appeals only to people interested in financial management.

Rocket Money Specific Content

It filters kids who won't engage with such content for more than 5 seconds and with enough kids swiping quickly through it, the algorithm will stop serving it to kids. Replace "kids" with any other gender, age, target persona.

Content as Targeting

Another example comes from 18 Birdies, a golf app. Their ad used insider golf terminology, ensuring only actual golfers understood it. This type of targeting works better than setting demographic filters.

18 Birdies Golf App

The golf app's target audience is primarily 35 and older, but they set their campaign targeting to 18+. Despite this broad range, the algorithm naturally attracted the right audience of 35+ due to content.

Perfect Targeting

Forced targeting backfires. One client targeted 25+ and ended up attracting teenagers because their content felt trendy and gimmicky.

Wrong Targeting

The New Marketing Model: Volume and Experimentation

Traditional media buyers tested audiences to find the best match. Today, advertisers must test content to identify what resonates.

1 in 30 ads drives half the total ad spend.

~ Hamza Alsamraee

Instead of randomly testing different creators and ideas, structured experimentation is necessary.

The best approach involves testing both hooks (the first few seconds) and concepts (the video’s main idea).

  • Old approach: Testing 10 different hooks for the same concept or 10 concepts with the same hook.

  • New approach: Testing a combination of different hooks and concepts to fully explore content variations.

Retaining Creative Learning

Brands often cycle through influencers without documenting what works. Instead find a creator that performs well and work with them consistently. Testing multiple creators at first is essential, but retaining top-performing ones improves efficiency.

Creative Learning Retention

"Most of the time, app advertisers will just go through and churn through dozens of influencers. What you’re going to realize is the creator themselves is the biggest independent variable."

~ Hamza Alsamraee

Creative Iteration

A systematic approach helps advertisers refine content strategies over time. Keeping a record of successful hooks, formats, and styles prevents repetitive mistakes and accelerates campaign performance.

Volumetric Testing

"You need to be more structured in your experimentation. Instead of just hiring random creators or testing random concepts, you should have hypotheses around what content works the best."

~ Hamza Alsamraee

The Future of Paid Acquisition

The landscape has changed, but advertising remains as effective as ever. Traditional targeting methods have lost precision, but self-targeting content fills the gap.

Post ATT

Algorithms no longer need detailed user data to match ads with the right audience. Instead, content itself determines who sees an ad.

Content As Targeting Works

“You’re not beating the algorithm at scale. You’re just buying an index fund.”

~ Hamza Alsamraee

Paid acquisition success now depends on volume, structured experimentation, and creative learning retention.

Advertisers who embrace this approach can achieve performance levels similar to pre-2020 campaigns, despite tracking limitations.

Hat Tip to Hamza Alsamraee for the insights.

Top Tweets of the day

1/

Google doesn't know how to spot AI. If it did, it wouldn't promote UGC-sites like Medium, Reddit, LinkedIn Pulse, & Quora.

So you need to make it human enough that people don't realize its an AI. It is very easy to do this with prompts now so you can actually produce a human-readable AI article in <$5 per article.

2/

There's a B2B business in any niche.

3/

Interesting insight. I tend to click more on "I" videos too.

Rabbit Holes

What’d ya think of today’s newsletter? Hit ‘reply’ and let me know.

Do me a favor and share it in your company's Slack #marketing channel.

First time? Subscribe.

Follow me on X.

More Startup Spells 🪄

  1. How Billie Eilish Used Instagram’s 'Close Friends' Feature to Boost Her Album Sales (LINK)

  2. Secret behind Airbnb's Billion-Dollar Empire? Spamming (LINK)

  3. Get Targeted Leads from Reddit (LINK)

  4. How Alex Hormozi Got Tons of Reviews for $100M Offers (LINK)

Reply

or to participate.