Since March 31, 2025, TikTok Shop has been officially available in France. This new e-commerce feature integrated into the TikTok app allows users to purchase products directly from their feed, without leaving the platform. For brands, this development presents a significant commercial opportunity, but also raises crucial strategic questions regarding marketing performance measurement and conversion attribution.
In this article, we delve into the effects of TikTok Shop on multichannel marketing attribution models and present two essential methods – deterministic reconciliation and probabilistic modeling – to effectively integrate this new conversion channel into a data-driven strategy.
TikTok Shop: A New Player in the Digital Purchase Journey
TikTok Shop allows brands to sell their products via native formats on TikTok: short videos, product carousels, and interactive live streams. Payment is made directly within the app, and part of the logistics can be handled by TikTok. Brands benefit from a centralised Seller Center for sales and performance tracking.
However, from a tracking perspective, this channel deviates from conventional standards:
- Purchases are made outside the brand’s e-commerce site.
- Users may have encountered multiple touchpoints before purchasing.
- Traditional tracking methods (UTM parameters, cookies, pixels) don’t allow direct attribution of conversions to the original marketing sources.
These factors present a real marketing attribution challenge: how can TikTok Shop sales be integrated into a reliable and actionable multichannel model?
The Limits of TikTok Shop’s Native Data
Here are the main types of data provided by TikTok Shop to sellers:
- Order data: date, amount, product reference.
- Customer information: name, address, phone number (often without an email address).
- Interaction stats: product views, clicks, conversions.
- TikTok touchpoint origin: live, video, product page.
However, TikTok doesn’t provide data on prior marketing sources: no UTM parameters, no acquisition channel data, and no referrer. As a result, it’s impossible to natively integrate these sales into a traditional web-based attribution model.
Still, two methods can overcome this opacity.
Accurate Customer Tracking with Deterministic Reconciliation: Wizaly’s Data-Driven Approach
Deterministic reconciliation relies on using personal data, especially email addresses, to link a TikTok Shop buyer to their previous sessions on the brand’s website.
Advantages of deterministic reconciliation:
- Web journey tracking: A user who visits a site via Google Ads and purchases days later on TikTok Shop can be linked through their email.
- Acquisition channel reconstruction: This approach enables attribution of part of the conversion to Google, Meta, or other intermediary channels.
- Complete conversion funnel view: It enriches the customer view with insights from multiple digital environments.
Example:
A user clicks on a Meta Ads campaign, visits a product page on the brand’s site, and logs into their account. A few days later, after watching a TikTok Shop live, they purchase the product directly on the platform.
Thanks to Wizaly’s deterministic reconciliation capabilities and the email provided at checkout, the platform links this conversion to the user’s prior journey. The sale value is then distributed between Meta Ads and TikTok Shop, more accurately reflecting the real impact of each channel.
👉 Without Wizaly, the conversion would be attributed solely to TikTok, skewing analysis and investment decisions.
👉 With Wizaly, the brand gets a full, precise view of the customer journey, optimises its media mix, and maximises ROI.
Wizaly Refines Marketing Attribution: From Individual Precision to the Power of ORA Modeling
When an email address isn’t available, the ORA method (Offline Revenue Attribution) offers a strong alternative. It relies on statistical modeling of purchasing behaviour and exposure to various marketing levers.
How ORA works:
- Analyse TikTok Shop conversions in volume and over time.
- Cross-reference with exposure history to paid and organic campaigns.
- Apply probabilistic models to estimate each channel’s contribution.
Example:
A user purchases via TikTok Shop after being exposed to a YouTube Ads campaign and a subway ad. Without personal data, ORA reallocates a share of the conversion to those two channels based on the likelihood of their influence.
Implications for Media Mix Optimisation
Not integrating TikTok Shop into your attribution model poses several major risks:
- Overvaluation of TikTok: The entire conversion is credited to TikTok by default, even if other channels played a role.
- Undervaluation of contributing channels: Google Ads, Meta, or email campaigns may be unfairly excluded from analysis.
- Biased budget decisions: Brands could misallocate investments based on incomplete data.
By applying reconciliation methods (deterministic or ORA), brands can fairly distribute TikTok Shop conversion value among the various channels that influenced the purchase.
Benefits of Wizaly Attribution:
- Fairer evaluation of TikTok’s role in overall performance.
- Media budget optimisation: allocation based on reliable cross-channel analysis.
- More strategic and profitable media mix management.
Conclusion: With Wizaly, Integrate TikTok Shop into an Advanced Attribution Strategy for Smarter Marketing Decisions
TikTok Shop marks a key milestone in the evolution of social commerce. As a native transactional channel within a social platform, it transforms the purchase journey and requires adaptations to marketing measurement models.
Brands that integrate TikTok Shop into their attribution via deterministic reconciliation or the ORA method gain a more accurate and actionable view of their performance. This integration helps them avoid misinterpretations and make better-informed decisions in their digital strategies.
“Integrating TikTok Shop into attribution models is not just a technical adjustment — it’s a strategic shift. As analysts, our role is to make visible what isn’t yet, and to give each channel its rightful share in performance.”
— Théo Notredaeme, Data & Media Analyst at Wizaly
