Case Study: Scaling TikTok Shop at Unilever Vietnam – From Emerging Channel to Seven-Figure Monthly GMV

Scaling TikTok Shop at Unilever Vietnam: A Behavioural Commerce Case Study

A professional case study on building a social-commerce growth engine through creator ecosystems, livestream operations, ROI-led investment, content toolkits, and behavioural commerce strategy.

Context

  • TikTok Shop was still a new and unproven commerce channel in Vietnam, with early strength mainly in fashion.
  • Shopee and Lazada remained the dominant ecommerce platforms, with more mature platform operations, traffic mechanics, and reporting habits.
  • TikTok Shop offered something different: discovery, entertainment, creator influence, product education, and checkout inside one ecosystem.
  • This made the platform potentially more “full funnel” than Shopee or Lazada, which often relied on external traffic sources such as Meta or Google.
  • For FMCG, the question was not only whether TikTok Shop could generate sales, but whether it could scale profitably and support brand growth beyond short-term conversion.
  • Vietnam became one of Unilever’s strongest TikTok Shop markets in Southeast Asia at the time, excluding Indonesia as a separate market.

My Roles & Challenges

  • Prove whether TikTok Shop could work for FMCG categories beyond fashion.
  • Understand whether investment could deliver both scale and ROI, instead of simply buying short-term GMV.
  • Manage P&L differently from mature ecommerce channels during the testing phase, while still keeping strict approval and governance.
  • Build a repeatable operating model for livestreams, short videos, creators, agencies, brand teams, ecommerce teams, and platform partners.
  • Scale creator participation without losing performance control.Improve data quality despite fragmented TikTok Shop dashboards, different reporting windows, brand vs creator views, and cancellation/return time lags.
  • Balance aggressive growth with profitability, especially as the channel moved from testing into acceleration.
  • Learn which product portfolios worked best for TikTok Shop, and avoid unnecessary overlap with traditional ecommerce channels.
Team collaborating around laptops at a shared workspace
Analytics dashboard displayed on a laptop screen

What made impact: Year 1

Test-And-Learn Foundation

  • Focused on Personal Care in the first year.
  • Treated TikTok Shop as a controlled growth experiment rather than a mature performance channel from day one.
  • Ran A/B testing across the first six months to understand what kind of content, creator activity, livestream structure, and investment level could generate meaningful commercial signals.
  • Used monthly ROI reviews and P&L reviews to decide the next month’s investment level.
  • Used quarterly reviews to support budget approval and align leadership on risk, learning, and scale potential.
  • Temporarily allowed a lower P&L threshold than mature ecommerce channels, with strict approval, because the strategic goal was channel learning and scale validation.
  • Tracked GMV growth, ROI, creator onboarding, creator performance, and platform activity volume.
  • Looked for green signals before scaling, especially sustained month-on-month GMV growth.
  • Once GMV showed strong acceleration, with monthly growth moving from around 30% to 100% month-on-month during the test phase, investment and activity were increased.

What made impact: Year 2

Scaling The Operating Model

  • TikTok Shop was not just another online shelf. It changed how people discovered, evaluated, trusted, and bought products.
  • Attention was created through short videos, livestreams, creator storytelling, and campaign moments, rather than only through search or banner placement.
  • Trust was built through creators who could demonstrate products, answer questions, and make the brand feel more human.
  • Product confidence increased when shoppers could see products used in real time, especially for beauty and personal care categories.
  • Urgency was created through livestream timing, campaign-day mechanics, creator prompts, limited offers, and high-traffic moments.
  • Friction was reduced because discovery, education, social proof, promotion, and checkout happened inside one platform.
  • Product fit mattered: not every ecommerce hero SKU was automatically a TikTok Shop hero SKU.Products that were easier to demonstrate, explain, compare, or dramatise often had stronger potential on TikTok Shop.
  • Creator selection mattered because different creators shaped different forms of trust: everyday relatability, expert-like recommendation, aspiration, entertainment, or celebrity authority.
  • The platform required brands to think less like static retailers and more like behavioural journey designers.
Silhouettes of business professionals in a high-rise office overlooking a city skyline
Figure 1: Year 1 GMV demonstration and KPIs tracking chart showing quarterly GMV growth and activity metrics
Figure 2: Understanding consumer's journey on TikTok Shop - BCG Shoppertainment framework showing themes, key attributes and key enablers

Impact overall

Moving From Channel Operations to Strategic Growth Engine

  • Moved from campaign-by-campaign activation to a more structured channel operating model.
  • Built monthly and quarterly review rhythms covering ROI, P&L, GMV, creator onboarding, and key platform KPIs.
  • Increased livestream frequency from campaign-led activity toward daily live operations.
  • Added 24-hour livestream coverage during major campaign moments to capture traffic peaks and urgency.Expanded short-video and creator-led content as a source of both discovery and conversion.
  • Developed creator incentive programmes to improve recruitment, activation, and retention.
  • Built brand-level sub-stores to support faster scaling and clearer portfolio management.
  • Separated ecommerce product lists from TikTok Shop product lists to reduce channel overlap and improve product-market fit.Worked with R&D on product demo kits so creators could explain, test, and show product benefits more clearly.
  • Created content guidelines to improve message consistency while still allowing creator authenticity.
  • Improved dashboarding with D&A support so decisions could be made faster and with fewer reporting gaps.
  • Built agency segmentation so each agency could contribute based on its strongest creator pool and execution capability.

Data And Decision-Making

Two men sitting at a desk with laptops, viewed through a window.

Challenges

  • TikTok Shop data was fragmented across multiple sources and formats.
  • Different dashboards used different time windows and metrics definitions.
  • Creator-side views and brand-side views did not always line up.
  • Cancellation and return windows made financial performance harder to read.
  • Fast growth made the problem more serious over time.

Solutions

  • Worked with D&A teams to build an integrated performance view.
  • Used quarterly business reviews to support alignment on data definitions.
  • Combined quantitative performance data with qualitative insight from creators, shoppers, and brand teams.
  • Kept the focus on decision usefulness, not perfect data.

Behavioural Commerce Lens

  • TikTok Shop was not just another online shelf. It changed how people discovered, evaluated, trusted, and bought products.
  • Attention was created through short videos, livestreams, creator storytelling, and campaign moments, rather than only through search or banner placement.
  • Trust was built through creators who could demonstrate products, answer questions, and make the brand feel more human.
  • Product confidence increased when shoppers could see products used in real time, especially for beauty and personal care categories.
  • Urgency was created through livestream timing, campaign-day mechanics, creator prompts, limited offers, and high-traffic moments.
  • Friction was reduced because discovery, education, social proof, promotion, and checkout happened inside one platform.
  • Product fit mattered: not every ecommerce hero SKU was automatically a TikTok Shop hero SKU.Products that were easier to demonstrate, explain, compare, or dramatise often had stronger potential on TikTok Shop.
  • Creator selection mattered because different creators shaped different forms of trust: everyday relatability, expert-like recommendation, aspiration, entertainment, or celebrity authority.
  • The platform required brands to think less like static retailers and more like behavioural journey designers.
A busy marketplace with various stalls and signs in Polish.

Insights

  • Scaled TikTok Shop from an experimental new channel into a seven-figure monthly GMV channel in EUR.
  • Delivered consistent double-digit growth across both year one and year two.Saw strong month-on-month GMV acceleration during the early validation phase, moving from around 30% to 100% growth month-on-month.
  • Built and scaled livestream operations, including daily livestreams and 24-hour campaign livestreams.Onboarded 5+ creator/service agencies.
  • Expanded the creator portfolio to around 10,000 creators.
  • Supported Personal Care in year one and Beauty & Wellbeing in year two.
  • Helped Vietnam become one of Unilever’s strongest TikTok Shop markets in Southeast Asia at the time, excluding Indonesia as a separate market.
  • Built a more repeatable model for TikTok Shop investment, creator scaling, livestream operations, content readiness, and data-led decision-making.

Reflection: Understanding Everyday Purchase Decisions from Brand and Consumer perspectives

  • The biggest learning was that TikTok Shop behaved less like a traditional ecommerce platform and more like a behavioural commerce ecosystem.
  • Scaling the channel required more than media investment. It required the right creator network, the right product portfolio, the right content tools, the right livestream rhythm, and the right decision cadence.
  • ROI had to be managed with both discipline and flexibility: too strict too early would limit learning, but too loose for too long would create inefficient growth.
  • Creator ecosystems need structure. Scale only works when creators, agencies, product kits, incentives, content guidelines, and reporting loops are managed together.
  • Data quality becomes more important as a channel scales. Early learning can tolerate messy data, but larger investment needs faster and more reliable reporting.
  • The strongest commercial results came when behavioural understanding and operating discipline worked together: knowing how people notice, trust, evaluate, and buy, then building a system that could support those behaviours at scale.