You wouldn’t build a house without a concrete foundation. So why are many technology leaders attempting to adopt GenAI technologies before ensuring their data quality can be trusted?

Reliable and consistent data is the bedrock of a successful AI strategy. Incomplete or inconsistent data prompts GenAI models to propose equally unreliable outputs, calling the basic utility of these technologies into question. Therefore, addressing organizational data issues before AI adoption – not after – is key.

Let’s discuss how leaders in the most data-driven departments, including procurement, can address data quality issues today and build a more sustainable path toward AI adoption.

Data Is the Bedrock of AI Functionality

GenAI requires accessible and reliable data to function effectively. These models are trained on vast amounts of information, and the correctness of that training set influences the model’s ability to understand and generate “correct” responses – that is, contextually valid and factually accurate answers.

Imagine a GenAI model built to auto-complete procurement contracts based on existing supplier data. The model’s ability to complete contracts in a high-quality and efficient manner depends on the organization’s supplier data quality. The contract generation process will run smoothly only if the organization retains access to reliable supplier data (i.e., with high-quality data, completed contracts will include wholly up-to-date and factual information).

It’s important to define the difference between high- and low-quality data. Markers of high-quality data include:

• Completeness: Capturing and ensuring the availability of essential supplier information
• Validity: Ensuring data aligns with predetermined formats and standards applicable to procurement tasks
• Consistency: Ensuring consistency in the recording and storage of supplier information, such as implementing a documented taxonomy
• Timeliness: Having the assurance of accessing the latest supplier information when making procurement decisions

Complete, valid, consistent, and timely data enables better business practices, including more beneficial GenAI integrations.

Successful GenAI implementations drive competitive advantages in several core functions. According to McKinsey, leading AI adopters cite the following as their top objective for GenAI:

• Increased value of current offerings (30%)
• Increased revenue (27%)
• New business and/or revenue sources (23%)
• Cost reduction (19%)

Overwhelmingly, these competitive advantages are realized through (1) product and service development and (2) risk and supply chain management, the latter of which significantly impacts an organization’s ability to succeed. GenAI tools can simulate risk scenarios through simple chat functions; analyze historical data and market conditions to identify potential risks; and support the supplier identification process by vetting thousands of data points to provide concise supplier insights. Each of these capabilities contributes to an organization’s ability to mitigate supply chain risks and associated financial penalties, including non-compliance fines.

However, procurement leaders must audit their current systems to attain these competitive advantages.

Common Data Missteps to Address

The nature of your organization’s underlying data problems will be unique. A comprehensive data audit is the most practical way to determine the appropriate next steps for your department or organization.

We’ll use the procurement department as our example, since data makes such a crucial difference here. Additionally, GenAI interest is incredibly high among supply chain leaders, with only 2% having “no plans” to integrate these technologies over the next 12 months.

A procurement data audit requires the assessment of existing data practices, including the identification of key data sources and stakeholders (e.g., suppliers and vendors). Analyze your data for completeness, consistency, timeliness, and availability by asking:

• Does my organization store duplicative or unnecessary supplier records, including outdated or irrelevant information?
• Is my supplier data consistently updated?
• Are supplier records available in a centralized location for all employees and technologies?

Many procurement leaders may already know of their organization’s data gaps. For example, if projects are often delayed due to difficulties identifying an alternative supplier, your supplier data is likely incomplete, scattered across various systems, or inconsistently updated. Don’t worry – this hurdle is common. The average sourcing and procurement operation takes nearly five weeks to identify a new supplier.

How to Organize Your Data for GenAI

High-quality supplier data enables leaders to identify suppliers much more quickly, unlock insights into spend analytics, and remove the need for manual intervention – ultimately improving the effectiveness of GenAI. Leaders must adopt technologies that consistently enrich and validate organizational data to unlock superior supplier data and insights – for example, a supplier data foundation.

Supplier data foundations solve many of the most pressing challenges procurement leaders face today, including unreliable data and lack of centralization. Improvements in this arena not only benefit an organization’s ability to meet tight sourcing deadlines and consumer demand, they also enable organizations to take advantage of GenAI’s competitive promise. Supplier data foundations ensure that the organization’s supplier data is comprehensive and trusted, providing GenAI integrations a solid starting ground for better outputs.

As such, procurement organizations looking to leverage GenAI technologies must establish a robust supplier data foundation to ensure their supplier data is up to par. By ensuring their supplier data is high-quality, accessible, and routinely updated, leaders can improve their GenAI outcomes, streamline critical processes, and make data-driven decisions. The importance of these functions cannot be overstated as we advance further into the era of AI.