If you are a Thai seller running Shopee and TikTok Shop in parallel, with one creator pool producing content that has to work on a shelf platform and a content platform at the same time, the decision in front of you is no longer whether to add another platform. It is how the same creators and the same assets get split across the places you now sell. Thailand is the first Southeast Asian market where content commerce passed shelf commerce on GMV in a single quarter, and the number is now well known: TikTok Shop took 48% of physical-goods GMV across the three main platforms in Q2 2026 against Shopee's 45%, with Lazada around 7%. Twelve months earlier the same comparison was Shopee 56% against TikTok Shop 30%. But the share line is not what should change how you run the two stores, because both are still growing — TikTok Shop did about $9.9bn in the first half against Shopee's $9.8bn, up 116% and 32% respectively, and across Southeast Asia as a whole Shopee still leads at 55%. What should change your operation is the category weight, because it says the two platforms are carrying different buyers. In the same first half, Shopee Thailand's category mix ran electronics 29%, fashion 24%, health 8%, while TikTok Shop Thailand's ran beauty and personal care 29%, fashion 23%, health 10%. Electronics is bought by comparing parameters, which is what a shelf page is good at; beauty is bought by watching a demonstration, which is what a short video is good at. So the question is no longer whether to be on both. It is what you are allowed to reuse, because a shelf platform and a content platform reward different things from the same product, and one asset served to both pays for the convenience twice over.
Price the reuse mistake in visits, because that is where it shows up first. Take a pool of 60 creators producing three assets a month, so 180 assets, averaging 40,000 views each — 7.2m views a month. Split the exposure roughly the way the market splits, six to four in favour of the content platform, and apply the rates each field actually produces: a demonstration asset on the content platform turns about 2.4% of views into product visits, and a specification asset on the shelf platform about 0.8%. Adapted properly, that is 4.32m views at 2.4% plus 2.88m at 0.8%, about 126,720 product visits. Serve one asset to both and both sides land near 1.0%, because the shelf buyer gets a demonstration without the parameters and the content buyer gets a parameter readout without the demonstration — about 72,000 visits. The gap is 54,720 visits a month. At a 2% order rate and 450 baht average order value, that is roughly 1,094 orders and about 492,300 baht of GMV a month, given up for the sake of not rewriting a brief. Add the second cost, which is the one that shows up in your team instead of your dashboard: without an adaptation spec, each of those 180 assets needs an extra round of explanation to make it usable on the other platform, and at twenty minutes each that is about sixty hours a month.
The share line is not the problem — the category weight is
Three mechanics explain why one asset cannot serve both platforms. The first is buyer intent: the shelf platform's mix is led by electronics at 29%, a category bought by comparing parameters, and the content platform's is led by beauty and personal care at 29%, a category bought by watching something work. The same product has to answer two different questions, and one asset only answers one. The second is frequency: TikTok Shop users in Thailand order about 11.8 times a quarter against Shopee's 9.9, up from 5.2 in the first quarter of 2023, so the content platform is not only bigger now, it is where the repeat habit formed — and repeat buyers respond to new demonstrations rather than to the same one relabelled. The third is enforcement, which is where reuse becomes risky rather than merely inefficient: the content policy of 22 May 2026 prohibits exaggerated performance claims, fake discounts, filter-altered before-and-after shots and false scarcity, and the product listing policy of 2 June 2026 makes a mismatch between a listing image and a video ad's claims a violation on its own. A demonstration that is fine on the content platform can carry exactly the claim that gets the shelf listing removed.
Two market facts are worth holding separately from the share line, because they shape what the two stores can promise. The Thai e-commerce market is about 1.15 trillion baht in 2026, roughly $36.2bn, with growth slowing from 7.6% to about 7% and e-commerce running near 30% of total retail, so this is a maturing market rather than a land grab. And the Trade Competition Commission froze commissions across six platforms for six months from September 2026, covering the 9.9, 11.11 and 12.12 peaks, which means fee-driven switching is off the table for the busiest part of your year — the split has to be won on content fit rather than on rate.
Step 1 — Split categories by field, not by platform
Sort the catalogue by field rather than by platform, because the field decides what the asset has to do. Categories bought by watching something work — beauty and personal care, fashion on a body, kitchen tools in use — belong to the content platform first, and the shelf listing there is a checkout page rather than a discovery page. Categories bought by comparing parameters — electronics, anything with a specification sheet, anything where the buyer already knows what they want — belong to the shelf platform first, and the content there supports the listing rather than the other way round. Fashion and health sit in the middle, and they are where most dual-track sellers waste the most: both platforms show them in the low twenties and high single digits respectively, which means neither is a clear call and both need their own version. Write the split down per SKU with one owner, and mark the middle group explicitly as needing two assets rather than one.
Use the buyer's question as the test rather than the product's category. Ask what the buyer has to see before they can decide: if the answer is a change on screen — how it looks on, how it works in a room, what the texture does — the content platform is the primary field. If the answer is a number — capacity, compatibility, dimensions, what is in the box — the shelf platform is the primary field. Products that need both are exactly the ones in the middle group, and they are the ones that need two assets rather than a compromise that serves neither.
One split list with a field per SKU is the whole artefact, and it only works if the person briefing creators opens it. register with DAMI and ask the team for the Thailand dual-platform SKU field split sheet so every SKU carries a primary field, a decision on whether it needs one asset or two, and the buyer question that decided it.
Step 2 — Write two adaptation rules, and keep the banned claims out of both
Write two adaptation rules, one per field, and treat the reused asset as a starting point rather than a finished one. For the content field, the asset has to earn the first three seconds: a hook that shows the change, a demonstration that runs long enough to be believed, and a spoken line that names the problem. For the shelf field, the same product needs a still that carries the specification — a first image with the parameters legible, detail shots that show material and scale, and captions that answer the comparison the buyer is actually making. Then apply the prohibition list to both, because this is where reuse turns into enforcement: no exaggerated performance claims, no fake discounts, no filter-altered before-and-after, no false scarcity, and none of the four claim categories the Thai FDA pursues — disease claims, rapid weight loss, sexual performance and absolute wording. Where a SKU sits in a TISI-mandated category, the reused asset also has to show the standard number, because the licence field and the physical label are checked separately.
Keep the demonstration and the claim separate in the brief, because that is the line enforcement falls on. A creator can show a product being used without asserting an outcome, and that is the version that survives on both platforms. The moment the script asserts an outcome — clears this, loses that, works faster than — it moves into claim territory, and on the shelf platform the listing image now has to match it or the mismatch is itself a violation. Give creators the demonstration and keep the outcome wording in a short approved list you review per SKU.
Step 3 — Run two sets of KPIs, because one total hides both failures
Run two sets of KPIs, because one combined GMV number hides both failures at once. On the content platform, measure what a piece of content is worth: attributed GMV per video, the share of creators who actually shipped an order, how many times an asset got reused, and dispatch performance during the peak hours, since the order clock starts the moment a video lands. On the shelf platform, measure what a listing is worth: search impressions, click-through, add-to-cart rate, and review score, because that is the chain a comparison buyer walks. Review them on the same day but never add them together, and never let a strong content number cover a weak shelf number — a SKU can be winning on video while its listing is invisible in search, and a combined total will show it as fine.
Hold the frequency number as the reason the content side compounds rather than substitutes. Thai TikTok Shop users order about 11.8 times a quarter against Shopee's 9.9, and both are up from 5.2 in the first quarter of 2023, so the content platform is not taking orders away from the shelf platform — it is adding occasions. That is why the two KPI sets should both be growing, and why a flat shelf number alongside a rising content number is a signal about the listing rather than proof that the shelf platform no longer matters.
Step 4 — Audit the reused assets weekly for the three things that decay
Audit the reused assets weekly, because reuse decays in three specific places and none of them show up as an error. First, the claim: a line that was acceptable when the asset shipped can become a problem when the SKU moves into a regulated category or when the FDA's attention shifts, so check the wording of your top twenty reused assets against the four prohibited categories. Second, the mismatch: a listing image and a video that have drifted apart as the product changed, because since 2 June 2026 the mismatch is a violation on its own. Third, the creative fatigue: an asset that has been reused past the point where it still earns the first three seconds, which you read as attributed GMV per video falling while views hold. Twenty assets, one pass, and record all three per SKU.
Put the audit on the same calendar as the peak calendar, because that is when reuse gets sloppy. The commission freeze across six platforms from September 2026 covers 9.9, 11.11 and 12.12, so the busiest six months of your year is also the period when you cannot solve a problem by switching platforms on rate. Two practical consequences: build the two asset versions before the peak rather than during it, and keep a named reviewer for the claim check, because in a peak week the claim check is the step that gets skipped.
How DAMI turns one creator shoot into two platform-ready assets
DAMI brings an 8-million-plus creator database, competitor and competitor-store creator search, automated outreach tasks, AI-assisted outreach, AI video creation, sample management, campaign management, social media management, store analytics, and multi-store team dashboards into one place. Three actions turn one creator shoot into two platform-ready assets, and each ends in something you can review rather than something you have to remember:
Find the creators whose content already works in your category on each platform
Use DAMI's competitor-shop creator discovery to pull the creators already producing for comparable shops on both the content and the shelf side, then compare them by field rather than by follower count. The output is two shortlists with the field each creator is strongest in, which is the input the split needs.
Track the two versions as two records with their own KPI
Use DAMI's store analytics to break GMV, orders and content shipped down by platform and by SKU, with the content-side and shelf-side metrics held separately. The output is one review that shows both chains, so a strong video number can no longer cover a weak listing number.
Manage samples per version so the reused asset has a source to check against
Use DAMI's sample management to tie each sample shipment to the SKU and to the version it produced, so the claim check and the mismatch check both have one place to look. The output is a per-SKU record that shows which asset came from which sample, which is what the weekly audit reads.
Frequently asked questions
TikTok Shop passed Shopee in Thailand, so should I move everything there?
No, and the category weights are the reason. Shopee's Thai mix is led by electronics at 29% while TikTok Shop's is led by beauty and personal care at 29%, and across Southeast Asia Shopee still leads at 55% of GMV. Move each SKU to the field it is bought in, not the platform with the bigger headline.
Can one video work on both platforms if I just cut it differently?
Only for the middle group, and even there it is the weaker option. A shelf buyer needs the parameters legible in the first frame and a content buyer needs a change on screen in the first three seconds, and one cut serves one of those. Categories with a clear field need two assets.
What claims should I keep out of the reused version?
Four categories the Thai FDA pursues — disease claims, rapid weight loss, sexual performance and absolute wording — plus the platform's own list from 22 May 2026: exaggerated performance claims, fake discounts, filter-altered before-and-after shots and false scarcity. Keep the demonstration and hold the outcome wording in an approved list.
Why does my listing get removed when the video was fine?
Because they are checked separately. Since the product listing policy of 2 June 2026, a mismatch between a listing image and a video ad's claims is a violation on its own, so a claim that passes on the content platform can remove the shelf listing. Audit the pair, not each asset alone.
Can I track both platforms on one GMV total?
You can, and it is the most expensive single number you keep. A SKU can be winning on video while its listing is invisible in search, and a combined total shows that as fine. Hold GMV per video, shipped creators and asset reuse on one side, and impressions, click-through, add-to-cart and review score on the other.
Conclusion
Layout is an allocation problem: the right SKU and the right creator in the right field, and everything above is about making that split an explicit decision rather than an accident. DAMI brings an 8-million-plus creator database, competitor and competitor-store creator search, automated outreach tasks, AI-assisted outreach, AI video creation, sample management, campaign management, social media management, store analytics, and multi-store team dashboards into one place.
Take your top twenty reused assets and check three things on each: whether the wording touches the four FDA categories, whether the listing image still matches what the video claims, and whether attributed GMV per video is falling while views hold. Fifteen minutes, and the third check tells you which asset has been reused past the point where it still earns the first three seconds.
Start with DAMI and put the split where the brief is written rather than where the platforms are compared.
