In an intrinsically connected world, data is everywhere. It was always there, in the form of information. What evolved so powerfully over the last decade is technology that allowed us to handle, process and analyse the vast amount of data. In knowledge economies where information is in abundance, how we process them to derive meaningful insights governs every aspect of our life.

My personal journey in data began while I was working for the first ride sharing platform called Tootle in Nepal. We invested significant time and effort in collecting and analysing data to optimise business processes. We leveraged real time data to create effective visualisation which resulted in higher number of ride completion and reduced waiting times for rides. We suited economic models in customer data to create personalised marketing campaigns. This was a good display of the power of data directly driving business outcomes. The passion of becoming data driven only intensified here after.

The burning passion to learn more about data led us to start Moonlit Solutions. Through Moonlit, we have delivered data solutions to businesses in sectors of education, healthcare and entertainment. We have performed expansive data analysis and forecasting to solve business problems like new market identification, competitor analysis, customer segmentation, identify revenue leakages, reduce unnecessary costs, identify optimum sales which resulted in significant market growth. We have also created data visualisation to give information a visual context through appropriate graphs and maps, simplified complex data into meaningful insights, identified data points through proper business KPI definition and established hierarchies for smooth navigation of information. This has ultimately helped businesses with their strategic planning, decision making, fundraising etc.

Data has a pivotal role in fixing complex humanitarian and environment issues too. About a year back, we created a visualization tool to monitor the melting down of glaciers in Nepal and their corresponding impact in the river basins. The melting down of glaciers is bound to change the flow of the rivers, ultimately threatening the environment. A lot of river basins also have huge hydropower potential in Nepal. Pools of investment are being pumped in the hydropower sectors. A visualization tool as such can be substantial to calculate possible risks and work on the conservation of glaciers from both environment and economic standpoints.

To say the least, delivering a few handful projects on data was very fulfilling to me. The best part about working in data is the opportunity to explore everything like a child like curiosity. Working with diverse clientele amplified our own learning as a team. I often felt how little I knew. While learning about forecasting techniques, I seldom wondered how everything is only a probability. It is very easy to fall into the fallacy of forecasting. It is essential to apprehend that it is almost impossible to forecast the risk of rarest events. Even Pascal, the father of probability recognized the limitations of his methodology in real life. While betting on God’s existence, he did not consider the probability. He considered the consequences of being wrong and chose the least consequential option.

My personal take away from my data driven journey is to do the math but never let it transcend wisdom.

Few misconceptions we have had to come across in this market:

In these two years, we have reflected on the myths that we have come across working as a BI service company in Nepal. Here is a collection of those reflections which might evolve with time. But today, here we are:

Myth 1: Some organisations believe that they do not have enough data required to perpetrate analysis.

Organisations can in fact begin with identifying appropriate data points for the business. For example: Defining key business and socio-economic KPIs, asking right questions to gather data that could drive business outcomes. Additionally, many organisations in Nepal use traditional methods of storing data either in bulky registers, record files or low formatted excel sheets or doc. Data aggregation and automation is visibly a need but the willingness is so overdue. Choosing a proper data storage, replacing traditional techniques and identifying new data points is where organisations can kick off.

Myth 2: Organisations are relying a little too much on intuition and heuristic reasoning. 

While, heuristic methods of analysis are noble on their own. But using it as a compensation for not being able to and(or) too lazy to crack the deductive reasoning is not the smartest move. It also projects a lot more risk. A proper weightage should be given to both the methods. 

One of the clients we did market research for responded to our report saying they knew most of the findings of the research heuristically before the report was presented. We believe numbers may not reveal something drastically new always. Sometimes it is merely a quantitative proof or validation to our hypothesis.

Reminds me of Adam Grant in this context: “Intuition is a source of information, it’s a data point we should learn from, but it shouldn’t drive our decision making. We should incorporate it.”

Myth 3: Data Analysis is about making few fancy charts and crunching numbers. There are so many softwares today that allow darlings of visualisation and analysis in a click. Why expansive data analysis as a service?

While, it is inherently true that there are plenty of amazing softwares that provide great analysis and instant visualisation but business intelligence is so much more. Each organisation is unique and requires customised solutions. Software alone limits providing catered analysis to unique business problems. This is where data teams can add immense value. BI is also about building a culture of visual thinking, organisation consciousness and expanding one’s curiosity and comfort to constantly ask the hard questions. 

Dhiraj Rajaram from Mu Sigma prophetically makes a case about Organisation Consciousness: “Thinking about questions is more important than thinking about answers. Intelligence comes from an interconnected network of ‘answers’.  But an interconnected network of ‘questions’ in an organisation is its consciousness” 

Myth 4: There are some cultural impediments around the sharing of organisation data and privacy.

Some organisations demonstrate interest in data science but do not exhibit the readiness to share their data to outside teams. Their concern mostly is about privacy. They generally tend to think that the work can be done by one in-house data scientist. It can definitely get them started on data analytics but expecting one person to do the work of a team is not a practical approach. It is also not the best solution for the sake of privacy. 

Even building an in-house data team possesses a lot of challenges like huge cost of time and money, scarcity of expertise and a required focus. This can be a good option only if organizations are committed to manage the team for the long run.

Outsourcing data science will give an opportunity to get focussed data solutions from experts who have a diverse know-how of data driven decision making. Therefore, businesses need to share data and trust outsourcing to rip the benefits of innovation and efficiencies. They can establish trusted channels for data exchange (eg : building and maintaining VPCs (virtual private cloud) , servers and  other cryptographic network protocols like SSH ). They can also adopt data anonymization techniques like data masking, pseudonymization, data perturbation etc. Legal contracts, non-disclosure agreements (NDA) and prior work collaboration also ensures trust.

Myth 5: Data is only about technology.

Somehow, data analysis is still only thought of as a technical role in many workplaces. Although it does require a solid set of technical skills of a software developer, there is another important dimension to it. Jose Miguel puts it very aptly in his TEDx Talk “A profile of a data scientist should be something like a polymath, someone with a wide spectrum of knowledge in multiple domains.”

In fact I would emphasise that all computing fields require a mandatory social science perspective. A mindful data analysis can only happen when algorithms incorporate social and economic forces that shape our world. Furthermore, rising concerns in ethics and data gaps only stresses the importance of integration of social science and data science.

Few Challenges identified while working in Nepal 

Nepal is not a surplus economy

Although the WorldBank forecasts a GDP growth at 2.5% in the fiscal year (2021-2022), Nepal is not yet a surplus economy in terms of falling income, worsening inequalities, job losses etc.

Businesses here are not as competitive and consumers are not as demanding as they would be in developed economies. Innovation would be a top priority provided businesses felt more challenged from market competition. But, most businesses in Nepal feel their endeavours are apt already. They do not feel the urgency to look into differentiation. And differentiation is one prime spot, where harnessing data would come along.

In cases, where organisations look into data, they still lack the patience of believing the results it can bring and seek immediate financial ROI. But, analysis is a constant process of monitoring and evaluating organisational performance. Data Science/ML is in fact a zero sum game. We are at the brink of a fifth industrial revolution, companies who invest in it now will clearly win in future.

Organisations are struggling to understand how analytics can help them improve business results

Some organizations have tremendous volumes of data, but they struggle to understand how to use the already available data to optimize their business processes. Data science helps in this process. The result of data science & modeling is to identify patterns, understand short term & long term trends using predictive analysis, evaluate financial models to manage cost and plan sales, sift through online & offline data to stay competitive in the business.

Hence, it is important for organisations to first know what data they have and consult with data teams to figure out what can be achieved. Underestimating or overestimating data capabilities blindfoldedly is one blunder we have noticed in organisations.

Small Market Network Challenges

In Nepal, the network and the degree of separation between companies are small. We are a moderately growing economy with pockets of companies in few cities. This brings up unique challenges for a technology service company to build a relationship with a bigger client, since they might already have a close relationship with other service companies. The size of the clients are less in a small market hence the competition for a strong personal relationship & price plays a major factor here.

Having said that, this opens up a huge opportunity for outsourcing technology and data projects in Nepal from around the world. 

The ray of light

Here is an analysis done by Moonlit Solutions that puts  a spotlight on Nepalese software developers and trends in Nepal through a comparative study with the global trends in technology and coding landscape.

Nepal can in fact be one of the top destinations in the world to outsource technology as it stands tall in all three pillars of outsourcing – favourable productivity cost, pool of talents and workplace culture.