There is heightened awareness of artificial intelligence and its potential to drive valuable business insights and generate market opportunities. However, amid growing demand for AI capabilities, most organizations remain unclear about the business problems they want to resolve.
To ensure they benefit from their adoption of data analytics, it is important that businesses first figure out why they need AI and establish use cases for its deployment. As it stands, AI projects often have high failure rates due to insufficient or inaccurate data, lack of relevant skill sets, and other factors.
Gartner estimates that half of IT leaders, through to 2023, will struggle to transition their AI projects past proof of concept to production.1 A key challenge is that most do not know how to go about establishing uses cases and are unsure as to what exactly they can and want to achieve from their AI initiatives.
Addressing this challenge will require an understanding of what actually is possible with data analytics as well as an awareness of the business problems the organization wants to resolve, specific to its domain and vertical. It may mean narrowing the problem statement or identifying use cases for which there is enough data to power the appropriate AI and machine learning models. It is only when an organization is able to establish clear business goals that it can then tap the opportunities AI and machine learning can yield.
For Amazon Web Services customers, this means working closely with AWS partners such as AspireNXT that can showcase existing use cases and explain what is possible with AI. Data also plays an important role in ensuring the success of AI projects. Businesses need to have the right data sets and know how to source the data. They will want to access quality data or figure out how to clean the data. The data will then need to be organized and may need to be converted into the right formats to be analyzed.
Organizations have to set aside time and resources to work through these issues and determine what kind of insights they can extract from the data. From there, they have to decide whether the data they have is sufficient to achieve their business objectives or if they need more, including cross-referencing data from external sources.
All data-related questions must be addressed before a proof of concept can be conducted. Organizations also need to identify the types of tools and services needed to support the project.
Tapping Data Analytics for Better Customer Retention
A Southeast Asian telco did just that when it approached AspireNXT for guidance on how to address limitations in its data infrastructure and improve its e-wallet. The mobile payment platform serves almost 9 million consumers across its domestic market. It was keen to tap big data to generate marketing insights and drive a better user experience for customers of its mobile payment app.
However, the mobile operator first needed to streamline a data infrastructure that had grown complex and fragmented over time. It also needed to find a more efficient way to ensure it remained compliant with the country’s complex reporting requirements even as its own business needs evolved.
AspireNXT put together a deployment roadmap for a data lake with AWS Lake Formation, which provides a centralized and curated repository that stores all data, both in its original form as well as a format-ready form for analysis. Establishing and managing data lakes can be tediously manual and complicated. AWS Lake Formation breaks down these complexities, enabling the telco to easily define data sources and the types of access and security policies that should be applied to each data.
The AWS platform resolves data silos and melds different types of analytics to generate deeper insights and drive better business decisions. It enabled the AWS customer to analyze petabytes of data with automated data quality pipelines, including notification systems.
The mobile operator now has full visibility across its users’ e-wallet journey, from when they sign up to when they uninstall the e-wallet. This allows the company to have a deeper understanding of customers’ behavior and interactions with the e-wallet, so it can enhance the app and improve user experience. Its data analysts also are able to run more analytics reports with at least 99% accuracy for all of its business units.
With the insights gained from the AWS deployment, the mobile operator was able to push its e-wallet users’ spending by 17%. It also improved customer retention with 50% more active users and reduced churn by 15%.
When organizations are clear about what they want to achieve and the business challenges they want to resolve from their AI deployment, they are more likely to see real returns on their investment. This further underscores the importance of having clarity on their end goals and establishing AI use cases that meet these objectives.
1 “How to Staff Your AI Team,” Gartner, Dec. 15, 2020
