Shishir Khanal, the new foreign minister, has spent much of his first month repeating a familiar phrase. Economic diplomacy, he has told ambassadors and conference audiences, is at the centre of Nepal's foreign policy. It is not a new idea. Versions of the phrase have appeared in ministerial statements for years, mostly without consequence. His sincerity is not in question. Whether anything will come of it is though entirely another matter. The trade figures rather make the point on their own. Last year Nepal's deficit ran beyond Rs 1500 billion on exports of Rs 277 billion, four-fifths of which went to India. Strip out the surge in soyabean-oil re-exports (an arbitrage that Delhi could end with a memorandum) and the country's actual export base is somewhere around a billion dollars. For thirty million people. With an apparatus of forty-four foreign missions. The numbers have an almost theatrical quality to them, as though designed to embarrass the speeches. Swarnim Wagle, who has spent years on the question, makes a useful point about what Nepal can realistically sell. A landlocked country with poor infrastructure cannot profitably export cement or steel; the transport cost eats the margin before the goods reach a port. What it can sell is what economists call high value-to-weight: specialty agriculture, premium artisanal goods, software and services that travel by wire rather than container. The coffee bean, the cardamom pod, the line of code. These are the exports geography permits, and ones that numerous reports have already identified as goods with comparative advantage. The trick is to find the buyer. Finding that buyer is the primary job of a commercial mission. The interesting question is not who is sent to these posts, but what is asked of them once they arrive. Suppose for a moment that the ambassador in Washington filed a report each quarter listing American buyers introduced to Nepali exporters, sample shipments dispatched, trade orders placed, US firms approached on Nepal as a manufacturing site. Suppose those numbers were public and suppose one's continuation depended on them. Within a year the work of the embassy would be unrecognisable. The ambassador might well be the same person, but they would be answering a different question. An institution becomes, in the end, what you ask it to report on. Here, then, are five things one might reasonably ask: Trade show participation. International commerce, for the most part, transacts itself at trade fairs. Magic Las Vegas: 80,000 apparel buyers a year. Gulfood Dubai: 90,000 buyers from 125 countries. The Specialty Coffee Expo, which incidentally sets the prices that filter back to growers in Gulmi and Nuwakot, whether or not they are aware of it. Bangladesh sends roughly 2,300 firms to such events each year, helped along by commercial offices that subsidise the booths and pre-arrange the meetings. Nepal sends fewer than 20. The fix is not mysterious. Anchor investor recruitment. Vietnam did not industrialise by writing investment-promotion policy. It found Samsung. One company now accounts for about a quarter of Vietnamese exports. Bangladesh repeated the trick in the 1980s with South Korean garment manufacturers. The current China-Plus-One reshuffle with Apple, HP, Dell, Foxconn, all visibly reorganising their supply chains is precisely the sort of moment for which a competent commercial diplomat exists. Vietnam, India, Bangladesh, and Mexico are presently in those conversations. Nepal, having something to offer but no one tasked with offering it, is not. Buyer cultivation. A specialty roaster in Brooklyn pays 15 to 30 dollars a kilo for high-altitude coffee. The Nepali farmer who grew the bean received perhaps five. The gap between those numbers is, more or less, what an embassy is for. Quarterly cuppings for American roasters. Sample shipments through reliable freight forwarders. Factory visits arranged for serious buyers. Israel offers a useful template: its embassies in Singapore, Dublin, Berlin and elsewhere house dedicated economic and trade missions whose stated job is to find buyers and pair them with Israeli firms. Bilateral trade between Israel and Singapore now runs to roughly US$ 4 billion a year, in no small part because somebody at the embassy has been working on it every Tuesday for thirty years. The diaspora as something other than a remittance pipeline. Three to four million Nepalis live abroad and remit ten billion dollars a year, an arrangement the state has organised itself rather competently to receive. What it has not organised itself to use is what those Nepalis happen to know. A restaurateur in London knows British food distributors. An engineer in California knows which firms outsource what. An accountant in Dubai has procurement officers across the Gulf in his contacts list. The work consists in connecting any of this to a producer back home, and in tracking what comes of it. Friction. Most small export deals die not because the product is unwanted but because something obscure has gone wrong in the middle. A sample stuck in customs. A certification refused on grounds that nobody can quite articulate. A letter of credit that mysteriously fails to clear. An Indian customs officer at Birgunj applying a rule on Tuesday that he had not applied on Monday. An embassy with relationships in both directions resolves these things in days rather than months. The cumulative trade lost to friction nobody is responsible for clearing is rather a lot. There is a further reason for hurry. Nepal's IT industry, which crossed a billion dollars in service exports last year and is often invoked as the country's quiet success story, sits on a window that is closing faster than its champions admit. The same large language models that produced the celebration are now being assembled into agents that can do, in 2026, what an entire BPO floor did in 2024. Roughly two trillion dollars of market capitalisation evaporated from the global SaaS sector between January and February of this year on precisely this fear; SAP fell sixteen per cent on a single earnings call. The AI agents are not yet mature. They are, in the engineers' phrase, 'baby AGIs'. When they grow up and they are growing up at impressive speed the routine workflows that today's Nepali developers and customer-service teams sell to American firms will be done, more cheaply, by software that does not commute. Nepal has perhaps a few years to convert its current IT exporters into something more defensible: domain-specific products, regulated-industry tools, services with deep local context. That conversion will not happen without commercial offices abroad doing the unglamorous work of connecting Nepali firms to the buyers who might still pay for what they make. These are not new ideas. They are what Bangladesh's commercial offices, Vietnam's investment-promotion teams, Ethiopia's coffee diplomats and Israel's trade missions have been doing for years. The reason our embassies do not do them is that nobody asks. The performance template that lands on the foreign minister's desk records political dispatches, ambassadorial calls and dignitaries received, and so political dispatches, ambassadorial calls and dignitaries received are what the system reliably produces. Which brings us to the sixth thing, the one that makes the other five possible: rewrite the template. The Foreign Ministry adds the five categories above to the quarterly report, draws the targets from the trade ministry's already existing strategy, makes the numbers public, and uses them to decide who keeps a post and who does not. It requires neither legislation nor new institutions nor fresh money. The whole exercise can be accomplished by ministerial order on a slow afternoon. It would cost almost nothing. It would change behaviour within a quarter and the trade figures within a year. The interesting question is whether the current ministry can be persuaded that this rather modest reform is more useful than the considerably grander ones they would prefer to announce. Kharel is Investment Manager at NlBL Equity Partners