A founder once walked me through a spreadsheet with just over a hundred rows. Each row was a conversation with a Japanese prospect: name, company, date, sentiment. The sentiment column was mostly green. Eighty-something positive responses, a dozen neutrals, a handful of polite declines. He had spent five months and a lot of flight money building that file, and he wanted my help with the last piece of the puzzle, which he described as pricing.

I asked him one question. How many of those hundred people had been asked to sign something, pay something, or carry something?

The answer was zero. Which meant the spreadsheet was not market research. It was a record of a hundred people being courteous to a visitor, and courtesy in Japan is not a weak signal, it is a non-signal. It carries no information at all, because the cost of producing it is zero.

This is the most expensive mistake I see in Asia expansion, and it is expensive precisely because it feels like diligence. You are doing the work. You are flying, meeting, taking notes, building a deck. And at the end of it you know almost exactly as much as you knew at the start, except you are five months older and your runway is shorter.

Saying yes to a foreigner costs nothing

I built consumer hardware in China for years, inside a venture where one product eventually became a line of eight. I sold banking innovation inside HSBC in Hong Kong, where my job was to put machine learning into lending decisions and personalization into products that had been sold the same way for decades. Different industries, different buyers, same pattern every single time.

When a foreign founder sits across from a local manager and describes a product, the local manager is not evaluating the product. He is managing a social encounter. Agreeing is cheap, graceful, and leaves every option open. Disagreeing is costly, requires justification, and risks embarrassing a guest. In Japan this is sharpened by a specific dynamic: the person in the room is almost never the person who decides, and he knows that you do not know that. He can be enthusiastic all afternoon because enthusiasm commits nobody.

I have sat in rooms in Tokyo, Shanghai and Hong Kong where the warmth was genuine, the follow-up emails were prompt, and nothing happened for eighteen months. I have also sat in rooms where I was cut off after ten minutes with a flat list of objections about integration, support hours and who carries the liability if something fails. The second meeting was worth fifty of the first. Objections are expensive to produce. Someone only generates them when they have started imagining the thing actually existing inside their organisation.

So the first rule I apply now, and I apply it brutally, is this: a signal only counts if producing it cost the other party something. Money, inventory, calendar time with a named internal owner, or reputation inside their own company. Everything else is noise, however green your spreadsheet looks.

In Asia, the channel sets the price, not the customer

Here is the part that founders resist hardest, because it contradicts everything they were taught about willingness-to-pay research.

In most Asian markets, your price is not discovered by asking buyers what they would pay. It is determined, almost arithmetically, by the economics of whoever carries you into the market.

I learned this the physical way, with hardware. When you build a consumer device in China, you start with a bill of materials and a factory minimum order quantity, and you think your job is to figure out what consumers will pay. Then you meet the actual structure. The distributor needs a margin to take inventory risk. The retailer needs a larger one. If you are exporting into Japan, there is often a trading company in the middle, and in some categories a second layer after that, each one with a number they will not move on because that number is what their own business model requires. By the time I had added up the stack, the shelf price was not something I had chosen. It had been calculated for me, and my only real decision was whether my landed cost left me alive at that shelf price or not. Across categories I have worked in, assume somewhere between 40 and 55 points of the end price disappears into the chain before you see a cent. If you built your plan on a direct-to-consumer margin, you do not have a pricing problem, you have an entry problem.

Software founders think this does not apply to them. It applies more. A SaaS product entering Japan through a systems integrator or a reseller is entering a world with its own fixed physics: the partner expects a recurring margin that is rarely under 25 points and often approaching 40, expects you to fund localization of the interface and the documentation, expects support in Japanese within Japanese business hours, and expects a named human they can escalate to. Those are not negotiating positions, they are the cost of them putting their own client relationships behind your logo. If a Japanese client relationship takes a decade to build, nobody risks it for your 15 points.

So when a founder asks me “what should we charge for this in Japan,” I tell them the honest answer is that I do not know and neither do they, and neither does any prospect they have spoken to. The number will come out of a conversation with a partner about how the partner makes money. Pricing a SaaS product for the Japanese market is a channel negotiation wearing the costume of a pricing exercise.

The sequence almost everyone runs is backwards

The standard playbook goes: research the market, size it, set a price, build the deck, then go find a partner who will sell at that price. That order guarantees you waste the first three steps, because the partner arrives last and immediately invalidates all of them. I have watched founders spend half a year on a pricing model and then rebuild it in a single afternoon after one serious conversation with a distributor.

Reverse it. Find the channel first. Let the channel’s economics tell you the price. Then, and only then, decide whether the market is worth entering at all.

That last clause matters. The honest outcome of this process is quite often “no.” When you run the channel stack and your product cannot support it, the correct decision is to not enter Japan this year, and that is a genuinely valuable conclusion reached in four weeks for the price of a few flights, rather than in eighteen months for the price of a local entity, a country manager and a localization project.

There is a second reason to put the channel first. Partners are the only people in the market who have already done your research. A distributor knows what competing products actually sell for, not what their websites claim. A systems integrator knows which enterprises have budget this fiscal year and which are in a hiring freeze. They know because they are wrong about it at their own expense. That knowledge is not available to you from a desk in Milan or London, at any price, in any report.

Run a four-week paid pilot in one market, not a six-month survey of five

Here is what I would do instead of the hundred conversations, and I would compress it hard.

Pick one market. Not Asia, not APAC. One. The temptation to survey Japan, Korea, Singapore, Taiwan and Vietnam in parallel feels efficient and is the opposite, because five shallow pictures tell you less than one deep one and you learn nothing transferable until you have gone all the way down in a single place.

In the first week, stop talking to end customers entirely and start mapping who already carries products like yours to them. Distributors, integrators, agents, industry associations, the regional arm of a trading house. You are looking for people whose existing business would be slightly better with your product in it, not people who would need to build something new.

In weeks two and three, have the margin conversation early and without embarrassment. Ask directly: what do you need to make on this, what do you expect us to fund, what does your end client pay today for the nearest alternative. Expect the answers to be uncomfortable. Then propose the smallest possible paid commitment that creates real exposure for them. A first purchase order, even a tiny one. A paid pilot with a named internal sponsor and a start date. A reseller agreement for a single named account. The size of the number is almost irrelevant. What you are testing is whether anyone will convert enthusiasm into exposure.

In week four, you count. Not meetings, not sentiment. Count signatures, invoices and inventory. If the number is zero after you have asked clearly and repeatedly, you have your validation result, and it is a real one. If the number is one, you have a market, and you now also have your price, because it was handed to you by the person willing to carry you.

A hundred conversations told that founder nothing. One partner willing to risk a little of his own reputation would have told him everything.

I write from twenty years of building businesses between Europe and Asia. If your company is facing this, start a conversation.