If you sell white-label goods on TikTok Shop Indonesia from a cross-border entity, where search and product cards used to carry most of your orders, the change that hurt this year was not in your product or in your pricing. It was in where your traffic comes from. Sellers comparing this quarter with the last one keep reporting the same two things: the same ad budget now buys roughly two thirds of the volume it bought a month ago, and the product cards that used to sit on the first screen no longer do. The instinct is to treat it as a campaign problem and bid harder. It is not a campaign problem. Permendag 19/2026 took effect on 8 June 2026 and requires marketplaces to place Indonesian products, especially those from micro and small enterprises, at the top of search results, recommendation feeds and category rankings. A seller who holds an NIB and self-declares as local gets that placement without paying for it. A cross-border seller does not, and no amount of bidding rewrites a ranking rule. Which is why the fix is not more spend — it is finding a route to distribution that does not run through the ranking layer at all.
Put a number on the drop before deciding how much to spend fixing it. Take a SKU doing 900 visits a day at a 2.4% conversion rate and Rp138,000 average order value: that is roughly 22 orders and about Rp3m of GMV a day. A 25% fall in organic visits — the range most cross-border sellers report since the regulation landed — removes about 5 orders and Rp740,000 a day, or roughly Rp22m a month, from that one SKU. Multiply by the number of SKUs that depend on search rather than on creator content and you have the real size of the hole. Now put the other side of the equation next to it: a local creator video reaching 40,000 views does not compete for that ranking slot at all, because recommending a creator's video and ordering a product card are not the same mechanism. You are not trying to outbid a rule. You are trying to move a share of your distribution onto a surface the rule does not govern.
Why this drop is structural, not a bad week
The distinction that decides everything is which layer lost the traffic. Permendag 19/2026 binds search results, recommendation algorithms and product ordering — the surfaces where a shopper arrives already looking for something. It does not bind what a creator chooses to post, and it does not bind how that post gets distributed to people who were not looking for anything at all. That asymmetry is the whole opportunity. A local creator's video enters the For You feed on the strength of whether people actually watch it, not on whether the product behind it is domestically registered. When the viewer taps through, they arrive at your product card as a referred visitor rather than as a searcher, and the ranking preference never gets a chance to apply. This is also why the creators themselves have moved. Cross-border stores report invitation acceptance falling from roughly 15–20% to 10–15% since the rules landed, because creators can see that content for local stores gets recommended more readily. You are not arguing them out of their own interest. You are offering a product that holds up on the metric their video is actually judged on.
It is worth being precise about what did not happen, because the panic version of this story is wrong. TikTok Shop is still the second-largest e-commerce channel in Indonesia at roughly 32% of tracked GMV, against 23% a year earlier, and it is still growing about 40% year on year. What fell is the growth rate, not the base. That matters for the decision in front of you: this is not a market to exit, it is a market where the route to distribution changed and the old route got more expensive. The category data tells you where the pressure actually sits. Beauty and personal care is close to 60% of TikTok Shop Indonesia's tracked GMV, and across the first five months of this year that category fell about 10% on TikTok Shop while rising 14% on Shopee over the same period. If that is where you sit, the ranking rule is not your only problem, and creator content alone will not close the gap.
Step 1 — Split your traffic into four sources before blaming the algorithm
Open your store analytics and split visits into search, recommendation, shop tab and live, then compare each line for the last thirty days against the thirty before. Most sellers find the fall is concentrated rather than uniform: search and product-card impressions drop hard while recommendation holds or barely moves. That pattern is the signature of a ranking preference, and it tells you something useful — the part of your traffic that already runs on recommendation did not get worse, so the mechanism still works for you. If instead all four lines fell together, you are looking at an account-level issue such as a performance score or a policy flag, and creator content will not fix it. The split takes about fifteen minutes and it decides which of the next three steps is worth doing at all.
Write the split down per SKU rather than per store. Store-level averages hide the thing you actually need: inside one shop, the SKUs that were already carried by creator content usually held their volume while the SKUs that lived on search lost it. That gives you the priority list for step two, and it also gives you a baseline — without a per-SKU baseline you cannot answer the only question that matters in step four, which is whether the creator work moved anything at all.
Before you start reaching out it helps to have the screening criteria written down rather than held in your head. register with DAMI and ask the team for the local-creator shortlist template and get the local-creator shortlist template we use for this — it carries the three checks from step two and the seven-day measurement sheet from step four as fillable fields.
Step 2 — Judge local creators on content fit, not follower count
Three things are worth checking on a local creator before follower count enters the conversation. First, category concentration over the last thirty days: open their showcase and count how many of the last twenty product-carrying videos sit in your category. A creator whose last twenty videos span eight categories sells everything and converts on nothing. Second, the questions in the comments. Questions about size, material, delivery time and how the thing actually works mean the audience is shopping; comments that are only praise mean the audience is watching. Third, whether the scenes in their videos are the scenes your product gets used in. A creator who films in a cramped family kitchen is a strong fit for a kitchen product and a weak one for a gym product, regardless of how many people saw either video.
Set the bar in writing: require at least twelve of the last twenty videos to sit in an adjacent category, and require real shopping questions in the comments of at least three of them. Sellers who skip this and buy reach instead end up with videos that are watched and never tapped, which is the worst possible outcome here — watch time without a tap produces no referred traffic, so it does nothing at all about the ranking problem you started with.
Step 3 — Pull the creator pool from local stores that are already winning
The fastest way to build a local creator list is to stop searching for creators and start searching for stores. Find three to five Indonesian-registered stores in your category whose volume has held or grown since June, open their product pages, and list the creators who have posted about those products in the last thirty days. Every name on that list has already passed the test you actually care about: they make content for your category, they are willing to work with a store in it, and their videos are currently being distributed. You are not guessing at fit — you are reading it off the market. Then work backwards from each creator to the other collaborators on that store, which usually doubles the list without running another search.
Two filters keep the list honest. Drop anyone whose last post was more than fourteen days ago, because an inactive creator will not ship a video inside your campaign window however good the fit looks. And drop anyone whose audience sits outside the regions your logistics actually serve well, because a video that sells into an area where your delivery takes nine days produces refunds rather than repeat buyers. Both checks take seconds on a profile, and both remove a category of waste that only becomes visible weeks later.
Step 4 — Prove the lift over a seven-day window, not with affiliate GMV
Affiliate GMV tells you that a creator sold something. It does not tell you the thing you actually needed to know, which is whether your organic reach recovered. Measure that directly: pick one SKU, record its daily non-affiliate organic visits for seven days before the first video goes live, then for seven days after each video publishes, and compare the two seven-day means. A rise in non-affiliate organic visits after a creator video is exactly the signal you are looking for, because it means recommendation is feeding the product card rather than only the video. If affiliate GMV rises and organic visits do not move, you have bought sales but not distribution — and next month starts from the same hole.
Run this on three to five SKUs before scaling it. One SKU can move for reasons that have nothing to do with the creator — a price change, a competitor running out of stock, a public holiday. Three SKUs moving the same way inside the same window is a result you can spend against. Allow four weeks from first outreach to first reading: roughly a week for outreach and sample dispatch, a week for shooting, and two weeks of publishing. Sellers who read the number after five days conclude that the method failed when in fact only one video had shipped.
How DAMI turns a competitor's creator pool into your outreach list
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 cover the loop above, and each one produces something you can look at rather than something you have to take on trust:
Find the creators already selling for local competitors
Open competitor and competitor-store creator search in DAMI, pick three Indonesian stores in your category, and pull every creator who posted about their products in the last thirty days. The output is a named list with contact routes, category history and recent activity, so step three stops being a manual hunt through the app.
Screen and reach out without rebuilding the list every cycle
Run the three checks from step two across that list in one pass, then set an automated outreach task with your brief and sample terms attached. DAMI's 8-million-plus creator database and AI-assisted outreach keep the follow-up running while you do something else, and every reply lands in one thread instead of three inboxes.
Watch the seven-day window per SKU
Use store analytics to track non-affiliate organic visits per SKU against each publishing date, so the step-four comparison is already on a chart when you need it instead of something you rebuild by hand from exports.
Frequently asked questions
Does working with local creators fix the ranking preference itself?
No, and it is worth being clear about that. It does not change where your product card sits in search. What it changes is how many people arrive at that card without searching. Treat it as adding a second entrance rather than repairing the first one — the search route stays as expensive as the regulation made it, and you are building a route that does not go through it.
How many creators should I start with?
Invite twelve to twenty and expect four to eight to actually post. Acceptance for cross-border stores has been running around 10–15% since the rules landed, so a list that yields three videos from ten invitations is performing normally rather than badly. Budget for the invites, not for the acceptances.
Is it worth paying more to a creator who asks for it?
Only after step four. Before you have a seven-day reading you are paying more for an outcome you have not measured, which is how creator budgets drift upward without any evidence attached. Once you have the reading, pay more to the creators whose SKUs actually moved and let the rest go.
Do I need an Indonesian entity for this to work?
No, and that is precisely the point. Creator distribution does not check your registration; it checks whether the content holds attention. That is a different test from the one the ranking rule applies, and a considerably more winnable one for a cross-border seller.
Can I just do all of this in a spreadsheet?
For the first cycle, honestly yes. A dozen creators, three checks and seven days of visits fit in a sheet, and if that is the scale you are running this quarter then a sheet is the right tool and we would not try to sell you past it. It stops being enough when the list passes roughly fifty creators across several SKUs and you are re-checking activity by hand every week, or when you need the same screening run for a second market. That is the point where the tool earns its keep, and not before.
Conclusion
Distribution is the one input you cannot buy back at last year's price, and everything above is about getting it from somewhere the price did not move. 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.
Open your store analytics, split the last thirty days of visits into search, recommendation, shop tab and live, and write down the three SKUs where search fell hardest. That is fifteen minutes, and it turns everything above from general advice into a list you can act on tomorrow morning.
Start with DAMI and start the next cycle from traffic you did not have to bid for.
