When I came to work at Lime as an Operations Manager, the business was scaling super fast. We were opening new cities each month, managing thousands of vehicles, and running complex logistics operations across a range of markets. The issue wasn’t having too many people, it was maintaining consistency, quality, and efficiency at scale.

The root issue with manual processes becomes apparent when you are scaled. While I was working at OYO, we were operating over 1.2 million rooms in 800 cities in 80 countries. The magnitude of our daily operations meant that even small inefficiencies in manual processes became gigantic bottlenecks when scaled up across our network. It may seem trivial to delay processing a property onboarding request by just five minutes, but when you’re handling hundreds of them a day across multiple markets, it adds up to hours of wasted productivity and income.

The execution bottlenecks I see tend to be found in three places. First, administrative and manual data entry procedures take disproportionate amounts of time that could better be used on strategic initiatives. Second, inconsistent practices across different team members create variability in quality and customer experience. Third, the inability to process information quickly enough to support real-time decisions puts companies at a significant competitive disadvantage.

At Lime, all of those challenges were particularly acute because our company is based on real-time optimization. Sending out vehicles, scheduling maintenance, and rebalancing all require millisecond decisions based on real-time information. Anything manual would simply fail to keep pace with the dynamic nature of our operations, especially as we expanded into new geographies with differing regulations, user behaviors, and infrastructure nuances.

The biggest insight I’ve gained from managing operations across different industries is that you must hardwire automation into your core business processes, not as an afterthought. When I was promoted to Senior Operations Manager at Lime, a top priority of mine was identifying what were the most critical processes to our success and most amenable to automation.

We started off by autopiloting our vehicle placement algorithms. Instead of letting operations managers rely on gut and rough reports to manually decide where to deploy scooters each morning, we developed data-driven platforms that filtered through historical usage patterns, real-time demand signals, and external factors like weather and events. It wasn’t merely about making it efficient, it was about making it consistent and optimized human operators could not reproduce manually across multiple markets at once.

The transformation was remarkable. Our utilization levels for cars increased significantly because we were placing inventory where and when it was most likely to be utilized. More importantly, we relieved our operations managers from routine decision-making to enable them to allocate time to strategic matters like regulatory compliance, community engagement, and planning for market growth.

Similarly, during my time at OYO, I developed process tools for technological transformation that optimized most of our business transformation process. Instead of needing to track project progress, resource allocation, and achievement of milestones in hundreds of projects running simultaneously manually, we engineered systems which provided us with real-time visibility and automatically warned us of likely issues before they became catastrophic problems.

One of the most important things that I’ve learned is that good automation is not about eliminating human judgment, it’s about augmenting it. The best automated systems I’ve ever implemented always allowed for human input and participation while eliminating the boring, repetitive, non-imagination, non-judgment-type tasks.

At Lime, our automated rebalancing systems are able to allocate vehicles more evenly across a city, but our operations specialists still take strategic decisions around special events, regulatory changes, and developing community relationships. The automation handles the routine optimization calculations so that human experts are freed up to do the relationship and strategy bits that require empathy, imagination, and context.

This collaboration between humans and robots is all the more crucial during phases of fast scale-up. Automated systems could handle run-of-the-mill deployment procedures and basic optimization when we were scaling up the business of Lime to new geographies, but we required seasoned operations managers to handle local regulations, community relationships, and adapting our standard processes to local contexts. The key is to develop automated systems that include clear escalation policies for unusual circumstances and are open about decision-making. Our personnel need to understand why automated systems are suggesting something so that they can participate in the process when things occur outside usual parameters.

The most revolutionary aspect of process automation is perhaps its ability to enable data-driven decision making at historic scale and speed. As General Manager, I oversee operations in a number of markets with varying profiles and challenges. To manually review operational data for these markets would be impossible, but automated analytics environments provide real-time insights that inform strategic choices.

Our automated reporting platforms constantly track important performance metrics in all markets, seeing trends, anomalies, and opportunities before any human analyst could. It’s not just a matter of efficiency, it’s a matter of competitive edge. When we can spot developing patterns of demand or operational problems within hours rather than days or weeks, we can react better than competing businesses that use manual analysis. At OYO, I found this skill to be crucial in dealing with the intricacy of rapid international growth. We needed to understand differences in performance between markets, identify best practices that could be replicated elsewhere, and solve operational issues quickly so that they didn’t impact customer satisfaction. Automated analytics systems facilitated this by continuously scanning massive amounts of operational data and pointing to actionable insights for management teams.

Practical execution of data-driven automation requires careful attention to data integrity and integration of systems. I have learned that automated systems are only as effective as the data they handle, and it is therefore imperative to put money in good data collection and validation procedures to guarantee the success of automation initiatives.

Effective process automation requires a managed process of balancing short-term operational needs with strategic long-term visions. Based on my working experience in different companies and industries, I have developed an effective model for identifying and ranking opportunities for automation.

The first step is to conduct an in-depth process audit to identify activities that consume significant time and effort with little strategic value. These tend to be routine administrative tasks, routine data processing functions, and routine decision-making processes that follow standard patterns. To conduct this audit, it’s necessary to involve front-line members who are knowledgeable about the day-to-day operating inefficiencies and issues that senior management might not see.

The second is to look at the business impact and technical feasibility of automating each process thus identified. Not every process performed manually can be automated, some require judgment, creativity, or interpersonal capabilities currently unavailable in current technology. The goal is to identify processes where automation will deliver substantive improvement in efficiency without damaging quality or customer experience.

Implementation needs to be phased and incremental, starting with high-impact and low-tech processes. This allows teams to build confidence and skills with automation tools while driving near-term value for the organization. We started at Lime with vehicle deployment optimization because it was tractable and fairly low-tech, and incrementally expanded into more complex areas like maintenance scheduling and customer service automation.

The final critical component of successful process automation is creating solid measurement and optimization frameworks. Automation isn’t a set-and-forget situation, it needs to be tracked, assessed, and tweaked constantly as enterprise conditions change. At Lime, we established key performance metrics measuring improvements in operational efficiency as well as quality aspects to ensure automation was not compromising service quality for speed. They include processing time, error rates, customer satisfaction ratings, and productivity levels among teams. Being measured on a regular basis highlights areas for additional optimisation and ensures automated systems are still aligned with business objectives.

Just as critical is sustaining feedback loops between humans and automated systems. Our staff members routinely offer observations regarding automation performance, special handling edge cases, and recommendations for system enhancement. This feedback helps sustain continuous system improvement and ensures that automation stays closely attuned to operating realities.

The cost structure should further include regular cost-benefit analysis to capture the return on investment in automation. Although improvement in efficiency will typically be apparent at once, the complete advantage of automation, such as improved decision quality, reduced error rates, and enhanced scalability, may take some time to take shape and requires systematic measurement to document.

My work at a succession of high-growth firms has convinced me that process automation early on provides long-term competitive advantages that gain momentum over time. Those firms that automate the guts of their businesses early on are in a better position to scale cost-efficiently, maintain quality, and adapt to changing market environments than those who remain committed to manual processes.

The automation we’ve built at Lime allows us to enter new markets more quickly and run more efficiently than competitors who have not invested similarly. When we open a new city, our automated systems can begin optimizing vehicle deployment, maintenance scheduling, and customer service processes based on local data on day one, while competitors will take weeks or months to manually build the same level of operational efficiency.

Process automation is actually all about building organisational capability for sustainable growth. By automating repetitive tasks through process automation, companies can release their human assets to focus on strategic initiatives, customer interaction, and innovation activities that generate sustainable competitive edge. Companies that embrace this first will be in a prime position to thrive in increasingly competitive and unstable markets.