A little over 10 years ago, after the 2015 earthquake, I met a woman in Rasuwa, one of the hardest-hit districts, supported by a post-earthquake livelihood programme, who had built a small herd of goats. One season she sold six at once and received Rs 150,000 – a substantial sum for a household that had never seen that much money in hand. By any conventional measure, this was a success story. Except it wasn't, quite. Within a short time, the money was gone. She could not say where roughly a quarter of it went – nothing she could point to as an investment or even a memorable purchase. It had simply dissolved into the pressure of having, for the first time, more cash than she knew what to do with. On paper, this was a success. For me, it was a failure. A programme can hit every indicator on its log frame and still fail the person it was designed for, if what it leaves behind is a memory of having had money, not a changed trajectory. I have thought about that woman for a decade, because her story contains almost everything wrong with how the sector designs livelihood programmes – and almost everything right about what would work instead. Most livelihood programmes rest on one of two assumptions: that we can turn people into entrepreneurs, or make them ready for a job. Neither holds up. Entrepreneurship requires an appetite for risk no course can manufacture, and nobody can make anyone 'job-ready' except themselves – we can open doors, but readiness has to be built, not supplied. A programme that sees readiness as something to supply designs one-size-fits-all training: gather a group, deliver a curriculum, declare them trained. A programme that sees readiness as something built instead asks what it takes to help someone face the problems their situation will keep throwing at them – because that capacity, not the certificate, determines whether a livelihood survives its first real test. The 'one-size-fits-all' model looks efficient: train a batch, assume the enterprise starts itself. In practice, this gets the sequence backwards. Market viability, access to resources, and above all, follow-up and coaching after training are not optional extras, they are where the real work happens, and almost always what gets cut first when budgets tighten. Post-training support should be the major part of a livelihood initiative; training is the beginning, not the substance. It also matters who we are working with: people carrying generational, rooted disadvantage, not a temporary skills gap one workshop can close. Pretending otherwise sets everyone up to fail. A few principles follow, and they cut against how most programmes are designed. Individualise everything. Group-based work has its place but cannot substitute for individualised coaching. People do not fail or succeed as a batch; they need support calibrated to their circumstances, and to understand, by design, that failure is part of the process, with real alternatives available. Fewer people, better results. A programme that tries to reach everyone will, in practice, reach no one well. Working with fewer people, more intensively and for longer, produces outcomes more likely to last, and those results justify scaling, rather than scale being assumed from the outset. Nothing pays off tomorrow. Almost no enterprise returns income the day it starts. Programmes that implicitly promise fast returns set people up for the disappointment that makes them abandon the activity. Saving and wise investment are the real skill. Income is not the endpoint. What someone does with it – how much they save and reinvest – determines whether a livelihood becomes durable independence or a series of windfalls that change nothing. Any livelihood activity that does not deliver basic income from the outset, or on a clearly rising trajectory, will be deprioritized – not from a lack of commitment, but rational necessity. People redirect limited time toward whatever meets their needs. A cooperative can look financially strong on paper while the individuals it served quietly walk away, because it never gave them a reason, in terms they could feel, to stay. Which brings me back to Rasuwa. The mistake was never the goats. It was the lump sum. Handing someone Rs 150,000 at once, after years of little discretionary cash, asks enormous financial discipline of a person who never had the chance to practice it. The money did not fail her; the design did. What would have worked better is almost absurdly simple: sell one goat a month, or one every two, instead of six at once. A steady, predictable stream of income – one she could plan around and absorb into her household's rhythm – rather than a windfall arriving faster than any habit of saving could catch up to it. That respects how people actually build financial discipline: gradually, through repetition, not one large test they were never prepared for. That, more than any training curriculum, is what real economic empowerment looks like. Not a dramatic transfer of money or skills, but a patient, individualised, honestly paced process – one goat at a time. The author has worked for over three decades in disability-inclusive development, child rights, and community-based rehabilitation in Nepal.