What does a successful data and AI strategy look like?

For any strategic data management leader, one of the primary steps is to develop a data and artificial intelligence strategy. This strategy must be fact - based, address the actual challenges faced by users, and use language that can be understood by executives, shareholders, and front - line data workers. The core of any data and artificial intelligence strategy includes the following four key elements:

• Data Governance
• Data Innovation
• Data and AI Analytics
• Data Culture

I. Data Governance

For most organizations, whether they are new to data governance or already familiar with it, the key to any successful project lies in building a non - intrusive way to reach the target audience. This is crucial for the Chief Data Officer (CDO) and requires careful consideration of the organization's willingness to assume data governance responsibilities. For organizations where data governance has not been formalized and there is no dedicated staff for data governance, it is necessary to integrate data governance into daily processes and even write it into job descriptions to ensure that front - line data workers can clearly understand data governance methods and contribute to the overall strategy while performing their daily work.

Considering this background, the overall approach to non - intrusive data governance is more of a value function rather than control. It helps users work within managed boundaries while providing data accessibility and availability, reducing the resistance to accessing the right data to drive insights and create value for the organization.

For the Chief Data Officer (CDO), one of the primary tasks in building a data governance framework is to develop strategies and standards that match the organization's risk profile. For example, in most organizations, data analysis is the largest data consumer. Hundreds of reports and dashboards present unique insights that need to be prioritized. Most importantly, users in the analytics ecosystem need to understand data usage guidelines, correct data access rights, and the availability of high - quality data that can be accessed through self - service.

Within the first 100 days, the Chief Data Officer needs sound data governance policies and standards to protect data, meet regulatory requirements, ensure the visibility of sensitive data, and enforce data quality business rules to ensure that the accessed data is suitable for use.

The process of formulating policies and standards should be carried out in collaboration with affected business stakeholders to avoid giving the impression of "command and control," which may lead to non - adoption by business units. A large part of data governance work involves communicating with people and creating a collaborative environment to actively showcase the leadership of business leaders. The Chief Data Officer (CDO) focuses more on creating value, not only supporting leaders to meet their immediate data needs but also empowering them to play a leadership role, understand the development roadmap, and identify potential risks due to lack of participation.

The key steps include formulating policies and standards to support the organization's data posture, such as formulating enterprise - level policies rather than just department - specific policies. Establishing and improving this policy library early is one of the most efficient tasks for the Chief Data Officer (CDO). The ultimate goal of data governance is to establish policies and standards, clarify data ownership, and ensure data quality. Although these goals are still somewhat abstract at present, they lay the foundation for achieving a positive return on investment (ROI), which will be discussed later in this article.

Summary of Key Points:

Create an enterprise and local data strategy and standard library

Develop data quality standards to provide high - quality data

Define data ownership by involving business stakeholders.

II. Data Innovation

For a specific organization, data innovation depends on how users can mine insights from existing data to address strategic application scenarios. For example, most banks or insurance companies operate in highly regulated markets. The hotel industry faces fierce competition from platforms like Airbnb but must also meet the needs of stakeholders to enhance customer loyalty. Similarly, the retail market aims to create real value for consumers while maintaining low prices and managing complex supply chains, suppliers, and products.

The success of data innovation largely depends on user profiles, their data needs, and the mapping of business use cases. I have communicated with global strategic data management leaders from different industries and regions and found a common point: many leaders tend to "cover everything." Based on my past experience, this phenomenon is particularly obvious in the proof - of - concept technology evaluation stage. Leaders often collect a large number of business requirements, but only a few truly drive these requirements to have the desired impact. The "cover - everything" approach can hinder the formation of a data - driven culture.

One of the primary steps in data innovation is to create a portfolio of use cases. Strategic data management leaders need to adopt a data product management mindset and clearly define the target audience, business problems, data scope, ownership, quality standards, etc. for each use case. Similar to a financial portfolio, use cases can be tracked for return on investment (ROI), thus avoiding intangible investments that cannot be achieved by governance alone.

Summary of Key Points:

Inventory use cases based on stakeholder surveys and mutual understanding are very important

Prioritize use cases based on their impact on organizational goals and focus on small - scale proof of value (POV).

Apply product management thinking: clearly define the problem statement, draw the target user profile, and outline the business impact.

Build a data community to ensure active participation, especially in the initial POV stage, to achieve quick results.

III. Data and AI Analytics

In any data - driven organization, data and AI analytics are the largest data consumers. Such a high - intensity ecosystem requires faster access to data, thereby improving data accessibility and availability. One of the main challenges reported by analysts is how to find and trust the right data for reports, AI, or machine - learning models.

For data and AI analytics, strategic data management leaders need to create value that can produce tangible results. The following key elements are required to achieve a positive ROI:

Data self - service: The goal of the data and AI analytics community is to mine insights and create AI/ML models. Access to clean and reliable data is the key to improving productivity. Organizations lacking self - service face challenges such as slow report generation, duplicate reports due to fragmented teams, and building and sharing AI/ML models using general data. A data self - service portal coordinated by strategic data management leaders can make it easier for analysis team members to use and access data.

Single source of truth: Creating a single source of truth requires registering the key data sources that the analysis team relies on. A data catalog helps organizations effectively understand and classify data, thereby achieving privacy protection, security management, internal and external data sharing, and project management. More importantly, a data catalog is the first step in building data products: data products are a collection of mapping key data to strategies, standards, and strategic use cases.

Gamification of the data analytics community: Not all members of the data analytics community contribute equally. Gamification can promote a sharing culture among analysts, scientists, engineers, and others. Strategic data leaders should identify key contributors, assign roles, and promote data reuse. Many organizations rely on fragmented data sources or Excel/SharePoint documents, which are difficult to scale as the data volume grows. Gamification helps users collaborate, improve data literacy, and generate applicable solutions that match strategic use cases. In this system, active contributors should be rewarded. The role of the Chief Data Officer (CDO) is to create value by improving data availability and accessibility to empower users.

AI increases the risk of losing daily data users who want to participate in strategic initiatives. Creating such an environment helps strengthen the overall data - driven culture.

IV. Data Culture

Building a data culture seems daunting because there is a risk that the value cannot be immediately realized. Many leaders confuse culture with business responsibilities, which may create the impression that data governance is for control rather than value creation.

A non - intrusive approach starts with a data literacy training program that does not take up too much of users' time and is adjusted according to users' tolerance. Based on my experience, successful data leaders collaborate with training and empowerment teams to promote data literacy programs and work closely with the human resources department to incorporate data literacy requirements into job descriptions, thereby ensuring that each relevant person assumes clear responsibilities.

Summary

By coordinating governance, innovation, analytics, and culture, strategic data leaders can transform data and AI into real value - driving assets for the organization. The following are four key points:

Establish clear policies and standards: Develop enterprise and local data policies, standards, and data ownership to ensure high - quality data and lay the foundation for a positive ROI.

Prioritize impactful data use cases: Inventory and prioritize use cases based on stakeholders' needs and organizational goals, and use data product management thinking to track ROI to avoid intangible efforts.

Enhance analytics capabilities through accessibility: Promote data self - service, build a single source of truth, and ensure data availability to accelerate insights, AI/ML development, and strategic decision - making.

Foster a collaborative data culture: Implement data literacy programs, engage the analytics community through gamification, and empower users to work within controlled boundaries to maximize value and drive adoption.

This article is from the WeChat official account “Data - Driven Intelligence” (ID: Data_0101), author: Xiaoxiao, published by 36Kr with authorization.