When considering globalization and expansion into emerging markets, the strategy slides and boardroom presentations often appear impressive: Millions of users, untapped opportunities, and excitement about scaling globally. For engineering leaders, we focus on the essential components — ensuring the technology functions reliably: Stable and reliable infrastructure, efficient code, predictable deployment, and all the elements that go into planning a technical rollout to deliver to new locations.However, the real challenge that we did not see or prepare for was about the subtleties of daily life in the areas we were expanding to — the realities of consistent, unreliable, and slow connectivity, the different devices people could afford, the way payments were made locally, the cultural context, and the worldview in which technology was understood and used.We observed users struggling in ways that our simulations did not predict, encountering invisible barriers where our code could not reach. At this point, we realized that the most challenging part of scaling is not engineering but empathy — having leadership that places as much value on human usage as the technology itself.Think Your Code Can Handle the Monsoon? Of course, you can predict how your cloud platform will behave under testable conditions, but how does that mean you are ready for the Mumbai monsoon season when the power grid is flickering. The mobile network is disrupted by the downpour? Or the chaotic elements of last-mile delivery when a local festival pops up and shuts down the area? What I learned is that significant roadblocks to scaling are not always technical bugs, but rather assumptions based on prior experience.From a Twilio-like (Communications API):Challenge: Consider the challenge of supporting a global platform that relies on SMS communication to deliver time-sensitive information (e.g., OTPs) for verification or notifications. As it expands into new markets where SMS cannot be delivered due to carrier issues (for example, India or parts of Africa), its growth prospects (and credibility) can deteriorate rapidly with few or zero SMS deliveries.Reality: By this time, the company learns that it has encountered traffic issues requiring registering sender IDs (which exist in India like the DLT system), blocked messages due to mandatory compliance with national DND registries, unreliable local carriers causing dropped messages, or finding routes that are reliable but expensive. In practice, typical international gateways often do not function.Leadership Decisions: Leadership should find time to become knowledgeable about local telecom regulations, build relationships with local SMS aggregators with enhanced carrier relationships, provide dynamic routing to select the best available path, and potentially develop fallbacks using either voice OTPs or unreliable data connections for push notifications (also, it is possible to consider WhatsApp-based messaging, where deliverability based on search results appears to be higher in some locations).Lesson: Communication infrastructure is not globally equitable or fair. Without working to understand telco local to international engagement, leading through regulations, and necessary channels for adaptation, success will only be met with barriers.Payment System ScenarioChallenge: Consider the challenge of a global e-commerce company or SaaS provider expanding to Southeast Asia, such as Indonesia and the Philippines, or Africa, and expecting end-users to pay by credit card. Almost every transaction showed a decline.Reality: It now finds out about low credit card penetration rates. Users predominantly pay using local e-wallets, bank transfers, mobile money (for example, M-Pesa), or cash on delivery. In addition, checkout flows need to integrate local payment methods and fail approximately 50-90% of the time for users due to limited mobile internet connectivity (see search results. Each country has its own regulatory requirements, resulting in fragmented markets.Leadership Decisions: Leadership should prioritize the integration of local payment methods. Equally important is the complete redesign of checkout flows for resilience to poor connectivity, for example, complete checkout for mobile money using USSD menus, only using minimum data on downloads, making callbacks server-side to confirm transaction status, even if the client disconnects during checkout, or payment methods that may need partial offline payment options. Compliance is ensured equally for each market. Lesson: Payment localization is not optional, and no technical or service solution needed for payments is robust enough to address unreliable connectivity issues during the payment process, where payments are actually made.Guiding Principles for Leading Across Cultures and Contexts These are not hard rules but mindsets that I think are very helpful during the complex challenges of scaling globally.Encouraging Different Perspectives in Design: Not designing in an echo chamber is critical. Be sure to ask for and include insights from engineers and team members who are working and living in your target markets. These are all useful perspectives to help you challenge your assumptions and identify blind spots early.Ask for Understanding through Real User Engagement: Simulators and metrics are helpful, but they do not replace the empathy of seeing real users engage with your product. Watching someone patiently wrestle with your app on a second-hand low-cost device or navigating a confusing workflow teaches you things you cannot capture in data.“Edge Cases” Can Be Mainstream: What is labeled as a minor issue or small user segment in one market may represent a core problem or a large portion of your users in another. Consider elevating the importance of engaging with these issues that are locally relevant, such as allowing SMS payments for unbanked users that turn them into banked customers, but in a different strategic context than your other market activities.Fight for Resiliency: Maintaining systems’ resilience under diverse conditions (e.g., offline sync, low-bandwidth mode) comes at a cost and typically also requires trade-offs with other priorities. The key is to communicate the value of these decisions and their long-term benefits in reducing risk, as risks can be framed as “not nice to haves” for sustainable growth. Acknowledge & CelebrateKey Takeaways: Recognizing the Digital Divide is within Evolution: Rapid digital evolution exists as they race to expand users, but at the same time, they must acknowledge the existence of large populations who are limited by factors (speed, reliability, cost, access, devices), and they cannot simply design for the top of the spectrum.Stress Test First: Designers need to stress test not only their servers for load but also stress test any assumptions about reliance on infrastructure and users’ ability to relate to that infrastructure’s reliability. For example, how does your organization design for situations where users are affected by power outages that impact their network, such as rain or extreme weather (e.g., monsoons)? Signals are affected by infrastructure issues, such as low-bandwidth signal strength. Testing during such periods may only reveal the expected duration of bandwidth limitations.When Things Go Wrong (Reflect and Rebuild Trust): Build transparency and accountability. Perhaps share learnings broadly (internally or even with a public post-mortem)? Owning failures lends credibility to the growing trust you are rebuilding. Embed lessons learned. Facilitating structured discussions (for example, “What If?” workshops that include variables of your markets) that help think through past failures into future planning to avoid repeated failures.ConclusionExperience in emerging markets reinforces one important lesson: our technology should adapt to the on-the-ground reality, not the other way around. There are challenges to overcome, and this is not a case of some applications dealing with intermittent connectivity, disruptive payment ecosystems, device take-up, or capability from which we can extricate ourselves. Resilience must be built in. When designing an effective offline mode for a product, consider how to adequately integrate local payment systems rather than replicating existing systems and undertaking the necessary stress testing relevant to on-the-ground conditions; these are not edge cases. These are critical aspects of building sustainable and significant products globally.The post Beyond the Code: The Real Work of Scaling appeared first on The New Stack.