24 days, and five holes he did not fall into

The testing lab's quote came back: twelve tests, ¥15,000.

Is that expensive? No idea. Has anything been padded? No idea. Can any of it be cut? No idea.

This is the kind of moment a cross-border seller hits every week — you have to make a decision, and you have nothing to base it on. You are not a testing engineer, you cannot tell EN71 from ASTM, and you have no second quote to compare against. Most people end up paying.

What follows is a record of a TikTok business lead spending 24 days handing moments like that to Optima. He runs several product lines across several overseas markets and has years of cross-border experience behind him.

Those twelve tests came to ¥9,400

He dropped a screenshot of the quote into Optima.

What came back had three layers: each test translated into plain language ("EN71-3 = screening for 19 hazardous elements"); which ones the target markets actually mandate and which are optional; and for each one, whether it could be dropped and on what grounds.

Three were cut: formamide, BPA, SVHC. The formamide one is the clearest case — it tests for a risk specific to EVA, and the product contains no EVA. Doing it changes nothing. ¥1,200.

¥15,100 down to ¥9,400, 37.7% off. Not from haggling. From separating what is required from what is not, and removing what should never have been on the list.

That is the reduction on one quote, not an average.

The more valuable part is what did not happen

Losing money in commerce is rarely the result of doing something wrong. Most of the time it is not knowing there was a hole there. Five were caught in advance over those 24 days, each with a specific mechanism.

That carrier's channel does not accept your warehouse. In the first round of freight comparison, one carrier's deferred channel was the cheapest and he nearly took it. The second pass found that the channel only accepts FBA addresses, not TikTok's FBT warehouses. Ship it, land it in the UK, fail to get it into the warehouse — returned or rerouted, two to four weeks and another ¥5,000–20,000.

That fact appears on no quote sheet. A carrier will not volunteer "my lane doesn't support your warehouse."

That category had seven or more international recalls in two years. Optima went and searched the recall databases on its own initiative. Without that, the normal sequence is tooling, sampling, testing — and only then discovering the compliance cost is far past budget, with ¥30,000–80,000 of tooling already spent.

"Without that warning I might have cut the tooling before finding out it couldn't pass — tens of thousands gone." — the operations lead, as recorded in the report

Four thousand units nearly went to destruction. The largest single figure here: from "write them off at a ¥170,000 loss" to "route them through FBT and recover ¥450,000."

One sentence nearly cost him a supplier relationship. He had to reply to a testing engineer in a group chat that included company executives. The draft said he was dropping SVHC and BPA — except the engineer's quote had never included either. Saying that in that room implies "I cut things you didn't even quote," which reads as not having read the quote at all. The wording went through six rounds before it was sent.

That one interests me most. It is not a technical problem — it is a social risk that a person under pressure is likely to miss. And it turns out to be checkable too.

What the 24 days produced

These items are not of comparable granularity — a 28-page competitor analysis and one short-video script are not the same order of thing. Adding them into a single headline number would be taking advantage of a reader who does not know this industry. So the table lists by area and gives no total.

Only one efficiency figure holds up: sourcing research ran 6–10× faster. At the conventional three to five days per product this batch would have taken 47–76 days; it took about eight. That multiple applies to sourcing research alone, not to operations as a whole.

What this actually shows

He was already experienced. Over those 24 days Optima did not teach him how to run a business — what it did was put the several kinds of expertise he could not possibly hold at once in front of him at the moment he had to decide.

A testing engineer's knowledge, the half-sentence a carrier will not say, a record sitting in a recall database, what one sentence weighs in a room with executives — none of these is hard on its own. What is hard is that they sit in five different fields and the decision has to be made this afternoon.

A person's knowledge boundary has always been the ceiling on how many storefronts they can run.

---

Material for this piece comes from the customer's operating record across a 24-day period. Every figure is given with its source and the limits of what it claims. At the customer's preference, the company and the individuals are not named.