Data. It’s how Netflix builds entertaining content and it’s how Starbucks improves company performance. The importance of data cannot be overstated in the modern environment, especially in the business context. As the saying goes, data is the new oil. But how does one gain access to this oil? This is where companies like Zkewed come into the picture.
Started almost 3 years ago, Zkewed is a Sri Lankan based analytics service provider. The company caters to three main segments pertaining to data, data warehousing, data science and data visualisation. Its list of clientele includes some big names like Unilever, Expolanka, and IFS. But like many startups, Zkewed had a rocky start.
Putting on the entrepreneur’s shoes
Before co-founding Zkewed with Gaya Premachandra, Bemali Wickramanayake was part of Dialog Axiata’s strategic planning team. While at Dialog, Bemali had the opportunity to work with several companies. The experience made her realise that companies lacked proficiency in handling data and insights.
At the time Gaya worked as an analytics consultant at N*able. Following a discussion, the duo agreed that their combined effort might help offer companies an effective solution. Shortly after, both Bemali and Gaya chose to leave their jobs and leaped into the world of entrepreneurship. Thus, a startup called Zkewed was born.
Zkewed started off trying to offer data analytics services for SME companies. After all, SMEs in Sri Lanka account for 80% of all businesses. The initial plan was to go for a subscription model. Unfortunately, many of the SMEs were not receptive to the idea.
“We knew what to offer to our clients. But we didn’t know how to get that message across.”
Furthermore, some of these SMEs do not have access to their own data. For example, think of a Point of Sales system at a small retail chain. Businesses find themselves where software vendors restrict the SMEs from accessing the data related to PoS. This only made things more challenging for Zkewed.
It took them nine months to land the first client. By this time, Zkewed diverted its attention towards larger organisations. In just a little over a year’s time, Zkewed managed to crawl towards the breakeven point. This was largely attributed to the company’s lean operational structure, where everyone works remotely. During its first year, Zkewed had only one employee. Two years later, the team grew to seven, and still operates almost entirely remotely. The company has developed a culture of learning where its employees are encouraged to learn new technologies to better assist their clients.
For a data-driven culture
From a Lankan context, one of the main challenges for Zkewed was that companies often lacked the mindset and the capacity to dedicate towards data analytics. Particularly SMEs. This hurts businesses in the long run. For example, take an eCommerce platform in the fashion industry. Understanding insights like what design patterns are moving and what would customers buy together, would prove vital in strategizing a business.
But conversations with companies were not necessarily focused on the technologies themselves. It was also about the culture around data, how companies can enable key decision makers to use data. This was at a time when decision making around data was not so much common, not just with SMEs but the larger corporations as well.
Take the example of one FMCG conglomerate. Before Zkewed, an analysis of yearly forecasts with actual data was a manual process at the corporate. This in turn prolongs the decision-making period. In contrast, a data analysis implementation would not only improve end-to-end efficiency but save time in the process.
Opting for data analytics tools and processes often result in change at an organisational level. At times, this change is met with resistance because embracing data analytics may mean going through a whole new set of learnings. It could also mean that certain employee tasks might change moving forward. Hence, the hesitance. According to Zkewed, this can effectively be countered via migration training and offering hands-on experience of any of the proposed solutions.
Instilling confidence to do more with data
A company’s data usually resides in multiple silos. This means that different units of a business may perceive certain scenarios differently. For example, take a large-scale manufacturer with hundreds of products with multiple brands. In such a landscape the sales department may prioritise products differently to that of the production department. This marks a clear data inconsistency within the company. Hence, multiple versions of the truth.
Thereby, it’s important to instil uniformity and overall visibility of company data across the board. There are several pillars to the process. One of the important steps is to implement an appropriate data governance structure and bring existing data sources to a “data-friendly” format. In other words, data warehousing. But not every company opts for data warehousing at the first go. Why? Because data warehousing requires a lot of investment in time and money.
In such scenarios, Bemali states that companies would rather prefer to dive into data analytics with the existing data sources. Zkewed tackles with what they call data visualisation exercises where the team builds Business Intelligence dashboards for these companies. The idea here is to offer a better perspective on the value of data analytics within an organisation.
Once that is in the picture, the next step usually entails forecasting. This is where Zkewed brings in machine learning and artificial intelligence solutions. It might sound fancy, but Bemali notes that this basically recognizes patterns in the existing data. Furthermore, a common misconception among businesses is that implementing AI and machine learning solutions require heavy investments, something that only large organisations can afford. Bemali points out that this is not the case. In fact, a mere CSV file is sufficient to implement AI/ML based solutions.
An agile future
All in all, the Zkewed team sees themselves as a consulting firm on analytics. As Gaya puts it, “we are not bound by technologies” and it is about inculcating a data-driven culture within the organisation. He views engagements between Zkewed and its clients as agile. But this can be tricky at times. Particularly when such a company may be required to switch between different technologies for different clients. Zkewed team tackles this by going through training regularly. In situations of specific requirements, Zkewed would either opt for recruiting the required talent or train themselves with the required skills.
The advantage of going the consultancy route is that sometimes discussions alone might result in more efficient solutions. In the FMCG example, the company went ahead with implementing Microsoft Power BI as it already had an existing Microsoft Enterprise package. This package included a free Power BI offering.
Zkewed’s agile approach has also contributed to other parts of the company’s success. Gaya points out that the Year-on-Year growth during its third year of operations reached 100%. As a result, Zkewed has now set its ambitions beyond Sri Lanka and Australia, further into Southeast Asia.
Zkewed might have had a rough start. But Bemali and Gaya fondly look back at the journey, stating that the team learned many lessons on things that do not work as much as figuring out what works best for the company.