This post was inspired by a week of working from Ambitious Impact's office in London, and chatting with several of the startup charities there. While my experience is in the for-profit world, I think it's applicable to entrepreneurs working on impact-driven endeavors in lower-income places, both for-profit and non-profit. (Acknowledging that much of this advice will be obvious to entrepreneurs from lower-income countries; I am mostly writing for readers who like me are from high-income countries.)

On Speed and Focus

• Speed is critical. The faster you can get through the earlier stages, the higher chance you have to succeed.
  • Early stage endeavors are fragile, and frequently fail when the founders decide to give up—getting bored because progress is slow, not learning much and feeling behind in your career, running low on funds to pay yourself because you took too long to get a pilot going: all these are risk factors for deciding to quit and can be mitigated by focusing on only what is critical to get to the next stage of learning.
• Focus on your feedback loops—how can you learn fastest about how to make the greatest impact? What are your biggest uncertainties?
• Don't use your time going to conferences or networking unless you have a clear plan for how such efforts will help you in specific ways (e.g. fundraising).

On Learning Things By Going Places And Talking To People

A lot of people start by looking at the demographic and economic statistics (like population numbers, GDP, density etc). This is a good start, but it's easy to spend too long theorizing, when the best next step is actually to visit and observe things on the ground and in-person.

• In general, cultural knowledge is hard to represent and discover online/in literature but easy to encounter in life, and culture tends to have a huge effect on behavior.
• With experience you can learn to translate statistics into useful insights about what to do, but it is difficult to map this without lots of on-the-ground experience; and even then, your experience won't necessarily translate between different cultural contexts. Every country is different, every region is different.

On Regulation

• Governments are not likely to help you early on. Eventually they can be helpful, but dealing with governments will likely slow you down at first. Focus on what you can do without permission or with minimal permissions required.
• When you do require some permission, do your best to map out the structure of the regulator:
  • Who is the decision-maker? Who do they report to and who reports to them?
  • What are they paying attention to? What motivates them?
  • Think of the regulator as existing to defend against threats: what are the nature of the threats, how do they defend against them, and how can you show that what you're doing is non-threatening?
• In many places the rule of law is relatively weak; permission is de facto granted by people, not laws. If the people like what you're doing then they will find a way to enable it. Conversely, if they don't like it, even if the law says it is legal, they will continually throw up roadblocks, slow you down or stop you. This is the default state of affairs and the fact that we/I expected anything different is WEIRD. Viewing it as "corruption" is a fairly unhelpful/distortionary lens, leading you to flail and perhaps decide that you "need to bribe somebody" to get things done—but that's both illegal and unlikely to solve the problem longer term. Instead, think of it as stakeholder management: you must understand the motivations of each individual in the chain, and orient your strategy around achieving your goals while keeping your stakeholders happy.

On Product

• Technology is not an end in itself. It is a means. Sometimes people start treating it as an end and they lose their way. For example, cryptocurrency has pretty much always been pushed as an end rather than a means which is why it never took off.
• Do things that don't scale. In many cases you can offer your "product" initially without writing any code! This is a really good idea when you can pull it off because it naturally focuses you on solving whatever problems your users have rather than the self-created problems of software and tools. At some point (perhaps quickly) if you're making something people need, you will get blocked on tech; but you should be able to learn valuable lessons before starting, which will speed you up dramatically.
• Prioritizing tech accessibility can increase your reach. As of the early 2020s:
  • Smartphones are far more widespread than computers.
  • Low-income customers usually pay per megabyte of mobile data, and frequently will turn off their data to save money on apps background downloading.
  • Web pages tend to require more constant connectivity, but they also are easy for someone to try out/share.
  • Downloadable mobile apps are shareable locally with Bluetooth, not just app stores, and don't inherently require connectivity; but it's hard to convince people to download apps.
  • In low-income places, you'll find many more Android than iOS devices. And of those Androids, Chinese brands and knockoffs will prevail.
  • Lots of people still have non-smartphones ("feature phones").
  • Many people struggle with literacy, especially tech literacy.