You could have the best jewelry ad in the world. But if Meta is showing it to people who buy $8 earrings from Shein... it doesn't matter.

Targeting is the part most jewelry founders set once, never revisit, and quietly blame the creative for.

This issue breaks down the 3 audiences every jewelry brand needs on Meta, why the order matters, and how to stop paying for clicks from people who will never buy your work.

CAMPAIGN BREAKDOWN

Most jewelry brands target once, half-heartedly, and move on. They pick a few interests, set their demographic, and let Meta figure out the rest. The result is budget scattered across people who have no intention of buying anything above $30.

The fix is a structured audience stack. Three audiences. Each one serves a different function. Each one belongs in your account at the same time.

Who it is: Website visitors from the last 60 days. Instagram profile engagers from the last 30 days. Your email list, even if it's small. Even 50 contacts is worth uploading.

Why it works: These people already know you exist. They've seen your work. The sale didn't happen yet, but the awareness did. Warm audiences convert at 3 to 6x the rate of cold traffic, which is why this audience is always your first priority.

How to build it: Create a Custom Audience from your website (pixel required), a Custom Audience from your Instagram account, and a Customer List upload from your email tool. Combine them into a single ad set. Keep the budget light here since volume is low, but watch ROAS closely. This is where your early returns come from.

Who it is: Website visitors from the last 60 days. Instagram profile engagers from the last 30 days. Your email list, even if it's small. Even 50 contacts is worth uploading.

Why it works: These people already know you exist. They've seen your work. The sale didn't happen yet, but the awareness did. Warm audiences convert at 3 to 6x the rate of cold traffic, which is why this audience is always your first priority.

How to build it: Create a Custom Audience from your website (pixel required), a Custom Audience from your Instagram account, and a Customer List upload from your email tool. Combine them into a single ad set. Keep the budget light here since volume is low, but watch ROAS closely. This is where your early returns come from.

Who it is: People who have never encountered your brand. You're finding them through layered interests and behavioral signals.

Why it works: This audience does the prospecting. It expands your reach and feeds the other two over time. It converts at lower rates, but it is how you grow beyond your existing audience.

How to build it: Layer 5 to 7 interests. Start with jewelry, handmade jewelry, and Etsy. Then add income or behavioral signals such as luxury goods affinity or frequent online shoppers. Then layer in complementary brand interests: Mejuri, Catbird, Gorjana, or similar DTC jewelry brands that share your customer profile. Set audience size between 500K and 3 million. Smaller gets you relevance. Larger gives Meta room to find buyers.

The order matters.

Build warm first, because it converts fastest and validates your creative. Build the lookalike second, because it scales your warm signal. Build interest-based third, because it feeds the top of the funnel once the other two are running.

BENCH INSIGHT

"For jewelry brands, lookalike audiences built from purchasers almost always outperform interest-based targeting. Your existing buyers are your best targeting data."

Interest targeting is a starting point. Lookalikes built from real buyers are a signal. Meta takes the behavioral fingerprint of every person who completed a purchase and finds more people who match. No interest stack you build manually comes close to the quality of that signal.

If you have 100 or more purchasers in your account, you have everything you need to build the most powerful audience available to you. Run it before you spend another dollar on cold interests.

REAL NUMBERS

These are directional benchmarks drawn from Meta performance data across jewelry and DTC product categories. Not guarantees. Use them as calibration, not targets.

Best for: Early returns, validating creative, recovering warm interest Volume is low. This audience is small. Do not over-spend here.

Your job: keep it running, refresh creative every 3 to 4 weeks to prevent fatigue.

Best for: Scaling efficiently beyond warm traffic Minimum source list: 100 purchasers. 500+ produces meaningfully better results.

Your job: let it run for 14 days minimum before drawing conclusions. Do not adjust early.

Best for: New buyer acquisition, feeding lookalike lists over time Lower efficiency is expected. This audience is prospecting, not closing.

Your job: test 2 to 3 creatives per audience. Watch CTR, not ROAS, in the first 10 days.

A realistic month-one scenario at $30 per day running all three ad sets: 5 to 15 purchases, mostly from warm and lookalike. Month two, the lookalike matures and ROAS begins climbing. Month three, you have enough purchase data to build an even stronger lookalike.

Each month builds on the last. The brands that stay consistent outperform the ones that restart every 30 days.

COPY CORNER

The same product speaks differently depending on who is about to buy it. Same sapphire. Same price. Same photograph. The messaging does the targeting work before Meta even decides who to show it to.

The product: Handmade sapphire pendant necklace, sterling silver, deep blue oval-cut stone, $185.

Same product. Same creative. Three opening lines. The buyer decides which one stops them. Your test tells you who is shopping.

AI PROMPT TOOLKIT

Paste this into ChatGPT or Claude. Fill in the brackets. It works best when you're specific about your best customers.

BENCH TALK

Meta has made Advantage+ Audience auto-enabled on new campaigns. The feature allows Meta to expand your targeting beyond the audiences you define if it believes broader delivery will improve results.

For low-AOV products, this sometimes works. For jewelry brands with an average order value above $80, it frequently dilutes targeting by pushing your ads to audiences with lower purchase intent and lower willingness to spend.

The behavior is easy to miss. Meta enables it quietly, and the campaign keeps spending. You see impressions. You see some clicks. The ROAS just never quite gets where it should be.

The fix: when building a new campaign or ad set, look for the Advantage+ Audience toggle in the audience section. Switch it to Original Audience Only. This keeps your targeting exactly where you set it and gives your purchaser lookalikes and warm audiences the precision they need to perform.

Manual audience selection still wins for premium jewelry. Until your account has substantial purchase history, do not let Meta expand beyond your defined audiences.

Ready to stop targeting the wrong people?

Targeting is where most jewelry ad budgets get quietly burned. Bad audiences make good creative invisible.

I work with a small number of jewelry brands directly: audience strategy, creative direction, full campaign management. If you'd rather be at your bench than inside Ads Manager, that's exactly what I'm here for.

Book your free 15-minute strategy call · jewelrybrandgrowth.com/p/strategy

Next in Bench Notes:

Your creative is the first thing they see. Three video formats that stop the scroll for jewelry brands, and how to shoot all three with your phone.