How to Use Data Science for Business Growth - Lessons from Data Scientists

Dominika Błaszak

Aug 3, 2021 • 6 min read
Man sitting next to his computer

Data science, machine learning, artificial intelligence - these terms may still seem like a fad or fancy buzzwords to some.

However, in reality these technologies are actual drivers of change in the way businesses operate and what experiences they provide to their customers. And these aspects directly translate into business growth opportunities and increased revenue.

We reached out to industry experts to discuss the role of data science and innovation for business growth. Here’s what we found out.

How did you use data science in your business’s digital transformation?

Ilidio Pereira, Manager of Customer Success at IVARE:

Now, you can't talk about the digital revolution and transformation without understanding the value of data. A company that has already embraced digital transformation as a strategy is able to understand and analyze market trends with the help of big data services and tools.

Customer buying patterns can be an example. By understanding them, a business can anticipate demand. Thanks to that, the most popular products are available, which positively impacts customer service. Big data is the perfect tool to gain insight from your customers.

Big Data can help you “foresee” your needs in advance and therefore provide better results in customer service. User problems can be minimized and resolved in time.

Arwin Sohrabi, Head of Data Science Platform Squad at Danske Bank:

Machine learning applications, including data science, are nothing new for Danske Bank. As one of the biggest banks in the Nordics, being at the forefront of technology is something we constantly strive towards. In areas as wide-ranging as fraud, risk, commercial banking, and beyond, data science is already employed to advance our products and services.

Advanced analytics directly and indirectly boost performance by increasing accuracy (data-driven rather than manual decisions), customer experience (more tailored offerings), and employee satisfaction (automation of tasks).

However, equally important is a solid foundation and infrastructure for providing and maturing advanced analytics capabilities in a modern enterprise. Data scientists should be able to work with model development in a professionally developed and tailored environment, without spending excessive time on other tasks.

Sam Brown, Ex AI Innovation & Intellectual Property Consultant at wefox, now self-employed consultant:

For wefox, how we used data to digitalize the insurance industry is more relevant than how we transformed ourselves; our vision was to transform the industry and our existence was only possible because the industry needed a digital makeover. We started as a digital company and created precedent for how incumbents should meet the digital era.

We used data science to optimize our actuarial algorithms, helping us generate more competitive insurance policies, provide better customer service by analyzing customer touchpoints, guide our product UI/UX offering using data on how users engage with our applications, and improve our operational efficiency and overheads by using data to enable automation of more internal processes like insurance claims processing.

We used real-time IoT data to help us better understand the risks and hazards in customers' lives, enabling better targeted product feature development. Data makes our prices more competitive and our product more engaging.

How did data science help you reach your KPIs?

Ilidio Pereira, Manager of Customer Success at IVARE:

Data science helped our company reach goals like:

  • Financial gains

A recent report by the Economist Intelligence Unit (EIU), based on a survey of more than 500 senior executives, mostly C-level, indicates that companies that make decisions based on data tend to do better financially.

  • Productivity increase: A similar study by Harvard Business Review found that data-driven companies not only make better decisions and leverage earnings, they also become faster and more productive.
  • More innovation capabilities: Business models in which data are important assets, and are even used to generate products and services, also tend to be more innovative. That's what makes startups so disruptive. This potential can also be better exploited by traditional companies in many segments of the market.

Sam Brown, Ex AI Innovation & Intellectual Property Consultant at wefox, now self-employed consultant:

The first is an operational efficiency KPI which we fulfilled and led the industry with. Wefox led the market for automating claims processing with quarterly figures ranging from 70-90% of all claims being automatically processed. As a reference, most incumbents currently manage 10-30% of claims automatically.

This is only made possible by data science where we can identify patterns in how to manage a claim with manual claims processing of the past and automate the claims which have a similar data structure. It would be almost impossible to make it much higher, as more complex claims have no precedents for data science to be supported with.

The next major KPI moonshot is reducing all unnecessary accidents by 100% and creating a safer world. We need IoT data to enable this, as it helps us determine when customers are likely to be in dangerous conditions that they might not be aware of and helps us create engaging content that can hopefully have more educational impact.

Prachi Bhide, Data Operations and Controls Manager at Deutsche Bank:

At Deutsche Bank’s Corporate Banking division we have adopted a business-case centric approach to data democratization. Our ‘Data Factory’ provides business users with ready-to-use, well-governed, and curated datasets for data science and analytics solutions. For example, we have successfully leveraged data science for holistic client-level data quality validation.

As with any large business handling a high volume of clients, reviewing and improving data quality can be very resource-intensive. This is often mitigated by doing batch sampling, but we aimed for something more comprehensive.

Using data science, we first clustered clients by profiling key features and then ran outlier detection to highlight deviations for further detailed, expert-led review. As a result, our subject matter experts could focus their resources in validating only the outlier cases. This has allowed us to perform comprehensive testing covering the entire population without slowing our delivery pace.

Justin Howell, CEO & Co-Founder at Rize:

At Rize we use empirical data to constantly evaluate the needs of our clients and test different avenues of approach for everything from product to marketing to sales.

By using data science, we're able to effectively iterate on what's working, throw away what's not working, and ensure we're using a customer-centric approach to give us the best chance at success.

Wrap up

As you may see, data science has many applications. It can be used by businesses to improve customer experience, establish competitive prices, improve UX/UI of digital products, boost performance by increasing accuracy and productivity, or prevent fraud. All of these areas have a direct or indirect impact on companies’ bottom lines. This means that data science is not another trend that will quickly fade away but, on the contrary, it may become one of the principles of modern businesses.
Your blog post content here…

Related topics

More posts by this author

Dominika Błaszak

codestories