Every month we ask people doing business on TikTok Shop Japan the same 10 questions, to record the market’s current “temperature”. Official figures add up what has already happened; they don’t show what’s coming. Outlook and action can only be learned by asking the people involved, so URITE runs this survey itself. Results are published free for everyone, and the formula is on this page.
Round 1 is in preparation; we’ll announce the dates once they’re set. The questions and the formula are all below. The survey itself is in Japanese.
We build three DIs (diffusion indexes: the share answering “better” minus the share answering “worse”), average them and add 50. “Don’t know” stays in the denominator (leaving it out makes the index swing). The three DIs are always published individually alongside the composite, because publishing only the composite would be dishonest.
Current DI = Q4 “increased” % − Q4 “decreased” % Outlook DI = Q5 “will improve” % − Q5 “will worsen” % Action DI = Q6 “increase” % − ( Q6 “decrease” % + Q6 “plan to stop” % ) URITE index = ( Current DI + Outlook DI + Action DI ) ÷ 3 + 50
Expanding. Sales, outlook and investment all lean toward growth.
Flat. Those growing and those shrinking roughly balance out.
Contracting. People and money are leaving the market. This shows up before weekly GMV does.
Measures the make-up of respondents; also used to identify prospects for stage 1 (matching).
Breakdown by size. Answers whether there’s still room for individuals.
Breakdown by category. Can be read against category trends in weekly GMV.
Core of the index ① Current-conditions DI
Core of the index ② Outlook DI. This is what official data can’t show.
Core of the index ③ Action DI. Shows the market’s temperature earliest.
A number no one publishes in Japan. Backs up a table of going rates.
“Two to four weeks” is an industry-media claim. We measure it to check.
Actual usage shares of the three main tools. Leads into an annual satisfaction survey.
An entry point for readers’ questions and material for next month’s articles.
We keep the questions fixed because changing them breaks the time series. If we add questions, we only append them at the end; existing numbers and options don’t move.
The median and most common range from Q7. No one in Japan publishes the affiliate commission rates actually in use.
The median from Q8. “Two to four weeks to respond to a change” is a claim by industry outlets, not a measured figure.
Q9: actual usage shares of Kalodata, FastMoss, Nint and EchoTik. URITE takes no referral fees from any tool.
If the total is under 30, we still publish the index but mark it “reference only, provisional”.
We always show the number of responses, overall and by segment.
The formula on this page is authoritative. If we change it, we update this page and record the change in its history.
Questions don’t change month to month. If one does, we state that the time series for that item breaks.
For example: “60% of respondents are sole proprietors, so companies may be under-represented compared with the actual market.”
Free-text answers that could identify someone are never published as they are.
The survey is conducted in Japanese for people doing business on TikTok Shop Japan (sellers, creators and support companies), and takes 2–3 minutes. Results are published free for everyone, calculated with the formula above. URITE takes no referral fees and doesn’t recommend any specific company. Results are never published in a way that identifies anyone, and segments with fewer than 30 responses aren’t published.
Go to the survey (Japanese)When we record a change, we post it on X first, and we write up each week’s numbers on note. Both are in Japanese. Email updates are in preparation.