When it comes to targeting audiences with TikTok ads, the most effective approach combines the platform’s unique behavioral signals with a structured, test-driven methodology. TikTok is not a traditional search or social network-it’s a discovery engine. That means your targeting strategy must account for how users interact with content, not just who they are demographically. Here’s how to do it right.
Start with the Right Foundation: Understanding TikTok’s Targeting Layers
TikTok offers several targeting options, and the key is to use them in a layered, strategic way rather than relying on any single one. The platform’s strength lies in its ability to surface content based on user behavior and intent, not just static profiles.
Core Targeting Options
- Demographic Targeting: Age, gender, location, and device type. This is your starting point-broad but essential for narrowing the pool. Avoid over-segmenting here, as TikTok’s algorithm benefits from scale.
- Interest & Behavior Targeting: TikTok categorizes users by the content they engage with (e.g., “fitness enthusiasts” or “DIY lovers”). These are powerful because they’re based on real-time actions, not self-reported data.
- Custom Audiences: Upload your own customer lists (email, phone numbers) or retarget users who visited your website via the TikTok pixel. This is where you move from cold to warm audiences.
- Lookalike Audiences: Build an audience that mirrors your best customers. Start with a source audience of at least 1,000 users (from your custom audience) for optimal results. A 1% lookalike is the most precise; higher percentages (2-5%) broaden reach.
- Hashtag & Creator Targeting: Target users who have engaged with specific hashtags or followed certain creators. This is especially useful for niche or community-driven campaigns.
The Best Way: Combine Broad Targeting with Creative Optimization
In our experience managing over $2 million in TikTok ad spend at Sagum, the single most effective strategy is to use broader targeting (such as wide age ranges or large interest categories) and let TikTok’s algorithm optimize based on creative performance. Why? Because TikTok’s AI is exceptionally good at finding the right users when given enough freedom and quality signals.
- Set your targeting to “broad” or use large interest categories (e.g., “Food & Drink” or “Fashion”) instead of hyper-narrow options like a 1% lookalike from the start.
- Prioritize creative that hooks within the first 2 seconds. TikTok’s algorithm rewards videos with high completion rates and rewatches. If your creative is strong, the algorithm will automatically show it to users most likely to engage.
- Use Spark Ads to turn organic posts into ads, leveraging existing engagement signals. This can significantly lower your cost per action because the content already has a proof of concept.
Structuring Your Campaigns for Success
A well-structured campaign is critical. Without it, even the best targeting will fail. We recommend a lean, test-driven approach.
Campaign Setup Framework
- Start with one campaign per objective (e.g., “Traffic” or “Conversions”). Avoid mixing objectives in a single campaign.
- Create 3-5 ad groups per campaign, each testing a different targeting approach (e.g., one broad demographics, one interest-based, one lookalike).
- In each ad group, use 2-3 different creatives (videos, not static images). Test variations in hook, format, and call-to-action.
- Budget wisely: Allocate at least $50-$100 per ad group per day during the learning phase. TikTok needs enough data to exit the “learning” status and optimize delivery.
- Monitor for 7-10 days before making major changes. Let the algorithm stabilize-don’t kill a campaign after two days of poor results.
What Actually Works: Real-World Tactics
Based on our experience, these specific tactics consistently deliver better audience targeting outcomes:
- Use “Interest & Behavior” targeting as a primary lever, not demographics. For example, target “Entrepreneurship” and “Business Coaching” interests for a B2B product, rather than age 25-45 alone.
- Leverage TikTok’s “Comment Targeting” (available in advanced settings) to reach users who have commented on videos within specific topics. This is a goldmine for high-intent audiences.
- Pair custom audiences with lookalikes for scaling. Once you have a 1,000+ pixel-triggered audience, build a 1-3% lookalike and run it alongside your broad campaigns. The two will often complement each other.
- Avoid over-targeting in the first 30 days. Narrow audiences (e.g., “women 25-34 in top 5 cities with interest in weight loss”) usually lead to high CPMs and low delivery. Instead, target broadly and let the pixel learn.
Measuring and Refining
Targeting isn’t set-and-forget. Use these metrics to refine your approach:
- Cost per action (CPA) and return on ad spend (ROAS) are your north stars. If a targeting option delivers a low CPA, scale it.
- Click-through rate (CTR) and video completion rate (VCR) indicate creative-audience fit. Low CTR? Your targeting may be too broad or your creative weak.
- Frequency: If it climbs above 3-4, your audience is too small. Expand your targeting or refresh creatives.
- Use BI dashboards (like we do at Sagum) to visualize performance across targeting segments. This helps you spot patterns-such as a specific interest category that consistently underperforms-and pivot quickly.
Ultimately, the best way to target audiences on TikTok is to combine a data-first mindset with creative excellence. Start broad, test systematically, trust the algorithm, and refine based on real performance data. That’s how you gain traction, hit your goals, and scale-without wasting spend on audiences that don’t convert.
