New VN TikTok Shop Funnel With PropellerAds: Zone-Based Strategy You Can Copy

Vietnam is quickly turning into one of the strongest performance GEOs in Asia, especially for social commerce. Users are buying straight from live streams, short videos, and in-app shops, not classic marketplaces. TikTok Shop sits right in the middle of this shift.

When our Asian partners started sending PropellerAds traffic to VN TikTok Shop via Onclick campaigns (through TikTok’s Spark), it created a powerful setup: you plug performance traffic directly into one of the hottest social-commerce “malls” in the country. Together with our Senior Account Strategist Karlina Berzina, we turned this into a practical, zone-based strategy you can reuse.

Offer Setup: A Whole Mini-Marketplace

Our advertiser’s TikTok Shop was structured as a multi-category store with three main sections:

  • Beauty: skincare and makeup products

  • Fashion: clothing and accessories

  • Gadgets: small electronics and smart devices

This reflects how TikTok Shop actually works in Vietnam. Beauty and fashion drive the volume, while gadgets help keep AOV healthy and attract impulse-driven scrollers.

Regarding this structure, it didn’t make too much sense to scale a single product, but find where and on which traffic the shop gets the best buyers in general. To do so, as Karlina, our partner’s account manager, explained, it was crucial to analyze zones, devices, and browsers first, and only then drill down into product clusters.

Step 1, Testing Vietnam Widely

The first step was a broad Onclick test for TikTok Shop with three key rules:

  1. Mobile-only: Android + iOS 

  2. Full-country coverage: all zones (districts/provinces), from large cities to smaller towns

  3. No early filtering: let the system explore browsers, OS, and zones before cutting anything

We were collecting data for analysis and kept an eye on:

  • CTR and engagement

  • Cost per checkout (CPCO)

  • Conversion rate (CR) and checkout completion

  • Day-to-day ROI stability

  • Refund/return risk by zone

Our internal impression data for VN confirmed what we already suggested: Chrome and Safari take about 80% of traffic, with in-app/social browsers (like Facebook) also relevant but smaller. 

Step 2: Searching for the Green Zones and Narrowing Down

After about 7-14 days, the team stopped looking at Vietnam as one huge GEO and started treating it as many micro-markets. Here is what it means practically:

  1. Analyzing zones

Each zone (district/province) was evaluated by CTR, CVR, CPCO, refund risk and ROI stability. Zones with high sales but significant percentage of refunds were deprioritized to protect profit. At the same time, stable and clean zones went to the whitelist.

  1. Building the whitelist

We assembled a whitelist of strong zones and focused on them instead of managing one large campaign with basic GEO targeting. 

  1. Custom bids for per-zone campaigns

Every shortlisted zone got its own bid/rate, because traffic in Vietnam varies.

Urban, Gen-Z, influencer-obsessed areas behave very differently from mid-income zones that click a lot but don’t buy consistently.

So we opted for precise zone-level bidding to cut off the low-performing traffic without accidentally killing the profitable traffic. It takes extra work, but has unbeatable benefits: 

  • Cuts about 20-40% low-quality zones and saves budget

  • Smooths daily CPA swings

  • Makes scaling more controlled because you know exactly where ROI comes from

Tthis is a structure you can literally copy-paste for other GEOs, not just VN.

Browsers and Devices: Chrome VS Safari VS Social

Browser + OS bundle also had a big impact on how users moved from Onclick to TikTok Shop and then checkout. From our tests, three clear patterns appeared:

  • Chrome brings the best volume
    Especially on Android, Chrome brought the most impressions and conversions and became the primary testing ground.

  • Safari is lower-volume, but pays well 
    iOS traffic had lower volume but higher AOV and better long-term buyer quality.

  • Social browsers perform, but are hard to predict
    Facebook and other in-app browsers showed good engagement but higher volatility. We kept them, but with tight control and only trimmed when data justified it.

In practice that led to structures like:

  • Separate campaigns for high-performing Chrome zones

  • Dedicated campaigns for “premium” Safari zones

  • Controlled testing budgets for Facebook browser traffic

This way, if performance shifts, you can tell whether a zone doesn’t work anymore or a specific browser/OS now behaves differently.

Scaling with Social Traffic via SmartCPM

Once the main TikTok Shop funnel got stable, the team added a second layer: Social Traffic optimization.

We didn’t want to rely only on native TikTok Shop traffic and started sending mobile Social Traffic (Facebook, Instagram) through PropellerAds SmartCPM, then applied the same zone-level logic to keep only the pockets that convert.

As always, we’ve had some rules for this test to follow:

  1. Mobile-only, latest versions
    Only Android and iOS latest builds were targeted to cut low-end devices and random desktop noise.

  2. High-impact browsers only
    We kept Chrome, Safari, and Facebook browsers and refused non-converting ones.

  3. All-social wide test with SmartCPM
    SmartCPM campaigns ran across different social placements (FB feeds, IG Stories/Reels, other social referrers) to find out which social pockets send real buyers, which zones pair best with social traffic and what the real zone-level price floor looks like.

SmartCPM handled dynamic price discovery per zone, while the team focused on reading patterns and whitelisting profitable micro-pockets.

Short term, this gives a clean cost baseline per zone and reveals unexpected gems (for example, a specific city + browser combo that loves beauty products).

Mid term, to add a second, diversified funnel into the shop, which makes scaling safer and less dependent on just one traffic type.

Key Takeaways for You:

  • Treat countries like clusters of micro-markets, not one GEO.

  • Run wide tests with a clear learning goal, then cut what doesn’t work.

  • Segment Chrome, Safari, and social browsers as separate audiences with their own bids.

  • Combine native social commerce (creators, lives, in-app traffic) with external performance traffic (PropellerAds Onclick + SmartCPM).

This approach turns TikTok Shop into a predictable performance channel you can scale on your terms. 

Join PropellerAds for relevant traffic and more actionable strategies that work! 

adindex