Macroeconomic events over the past few years have unfolded at a dizzying pace, creating uncertainty for many businesses. In the backdrop of this transformational change, businesses are now looking at ways to ensure the efficiency of operations, achieve customer satisfaction, and stay relevant and competitive. And data becomes the central thread across all these areas.
In 2022, Gartner identified data-centric artificial intelligence (AI), decision-centric data and analytics, and connected governance amongst the top twelve data and analytics trends. We believe these three trends will continue to dominate 2023 too as planning for business growth becomes more difficult with economic uncertainty looming ahead. Businesses will therefore need as much high-quality data as possible to power their decisions.
1. Data-centric AI: Greater focus on data
Traditionally when organizations have looked at AI and machine learning, data has been treated as a static artifact while the bulk of organization’s focus has been on the model. This model-centric AI approach — keeping the data fixed and iterating over the model and its parameters to improve performances. Model centric AI involves collecting and cleaning large volumes of data and using this data to run algorithmic models. As the name suggests, this approach emphasizes on the model rather than the data and the model needs to be constantly experimented with for better results.
Model centric AI, however, approaches present a set of problems. For example, not every organization has the capability or expertise to collect and process large volumes of data, the required technology can be costly, and the security concerns associated with vast data quantities becomes a compliance issue. Furthermore, multiple and different models may be in use within the same organization, leading to data inaccuracies and disagreements on which model to use.
However, the more recent trend of data-centric AI underscores the role of data and data quality in AI based applications. Data-centric AI focuses on collecting high quality data from the beginning in a way the data that will deliver the best results. Having higher quality data helps to improve the model too. Adopting a data-centric approach will help you overcome some of the challenges posed by model-centric AI:
- As the data that is collected is first cleansed, organizations will be able to access high quality, standardized and accurate data
- As teams do not necessarily have to experiment with models, your development processes become less time consuming. This will help with faster development of applications.
- Deciding on the exact types of data needed prior to commencing projects will enable consensus building, as this will be useful for determining the best model to adopt.
2. Decision-centric data and analytics: Achieving outcomes that matter
We often hear the phrase ‘data-driven decisions’ and understand the importance of using data to support decision making. Yet, what does decision-centric data and analytics mean? Just as data quality plays an integral role, so does defining the desired possible outcomes at the onset of a project. Adopting a decision-centric approach to data means first understanding all the possible business outcomes that the data could generate and planning for the follow up actions that will result from each outcome.
With this approach, organizations can focus on identifying the desired outcomes or results they want to improve before collecting the data that will enable the organization to achieve these. In other words, all the data you collect must influence a specific outcome (or decision).
Once your data and analytics functions are decision-centric, you are better placed to:
- Manage risks as you have already planned for several outcomes,
- Make fact-based decisions (i.e. literally ‘data-driven’) that are less susceptible to assumptions and bias,
- Make the right decisions confidently, and over time the practice of data-informing strategic decisions will become the norm,
- Explain the reasoning behind decisions with ease if required,
- Identify trends proactively and respond to them, especially since you would have access to historical data and have clearer insights about the development of certain trends and how they are likely progress in the future, and
- Evaluate your business performance over time as you have access to the right type of historical data.
3. Connected governance: Ensuring data integrity while responding to challenges
Discussions surrounding data invariably leads to questions about data governance and transparency. Data governance refers to the collection, management, usability, applicability, accuracy, and security of data. Gartner advocates a connected governance approach to data – a collaborative one that responds to business challenges and provides organizations with the flexibility to make decisions when market dynamics evolve.
A cross-functional connected governance approach empowers you to:
- Enact policies on data usage and access. By ensuring that authorized users have access to the necessary data, you facilitate team-wide access while protecting the data.
- Maintaining data integrity is key to ensure your data is relevant and of high quality. Your customers, too, will be encouraged because you only store data that is absolutely necessary for business operations – particularly personal data.
- Secure your data and prevent any data exposure through robust data protection policies.
- Integrate data across your organization’s teams and help to break down data silos.
- Comply with regional data privacy regulations such as GDPR, CCPA, LGPD, to name a few.
- Ensure data security, data integrity, and regulatory compliance so that your organization can be transparent with your stakeholders.
Ultimately, adopting data-centric AI, decision-centric data and analytics, and/or a connected governance approach will support strategies that maximize your business’s ROI. The world today is defined by its many volatilities; forecasting and anticipating responses to macroeconomic conditions are challenging for every organization. The importance of data is a constant – no matter the developments. A better understanding of the data required for desired outcomes and collaborative governance policies will provide you with the edge to pursue your growth strategies.
To find out how we can support your growth strategies by leveraging data and analytics, get in touch with us.