Blog

How To Manage Data Quality for Indirect Procurement Spend

May 13, 2021 | Aparna Gude

The Johari Window was a technique developed by Joseph Luft and Harry Ingham in 1955. The intent was to help an individual better understand their relationship with themselves and with others.

As a framework, this technique has been used by psychologists and self-help forums to help people understand themselves, their interpersonal relationships, and group dynamics. In corporate settings, this technique has been used to help with team development and inter-group relationships. More recently, this framework has been adapted for a variety of purposes, including personal improvement, team building, corporate communications, client-customer interaction, etc.

The primary purpose of these adaptations is to help an entity realize what it does or does not already know and to increase its knowledge with respect to another entity.

In this blog, I am advocating the use of the Johari Window Framework to help organizations identify a plan of action in order to improve and manage their indirect procurement and purchasing data quality. The realization of  outcome (an improvement in the completeness and richness of data for better data to insights, better sourcing, procurement and purchasing activities) comes once an organization acknowledges and validates its knowledge and proactively and iteratively works to improve the quality over time.

The framework classifies the perceived data quality (by a customer) vs. the actual data quality (as surfaced in an analytic tool) into data quality quadrants:

The Open Area (The Arena): What is known to the customer and surfaced in an analytic insight

This quadrant represents the data with the highest data quality. This includes the data that a customer is capturing well and which the spend management tools can proactively use to provide valuable insights. The data consists of spend transaction details such as item (goods and services) descriptions and attributes, supplier details, manufacturer details, receipt and invoice details, and other transaction details. It also includes spend patterns that are known to the customer and are available in the data processed by the analytic tool. This spend is well tracked and monitored and can be easily and readily identified, enriched, classified, analyzed, and used to realize savings, cost analysis, predictive trends, and sourcing activities.

For indirect spend, we see 10-20% of all indirect procurement and purchasing spend in this quadrant. For most customers using any procurement application, it takes discipline, effort, and time to get their indirect spend in this quadrant. Ideally, customers need to have price agreements, catalogs, contracts with suppliers, reliance on purchase orders, tracking details of their invoices and receipts, and enriched item descriptions to start seeing their indirect spend progressively move into this quadrant.

The Hidden Area (aka The Facade): What is unknown to the customer but is surfaced in the analytic tool

This quadrant represents the data with good data quality. Spend transaction details such as item (goods and services) descriptions and attributes, supplier details, receipt and invoice details, and other transaction details are available in many systems and surfaced in the analytic tool even though these details may not be readily available or known to the customer.

This quadrant represents data where enough details were captured to initiate sourcing initiatives. The data can be used for insights with high confidence in spend management for procurement, even though the organization may be unaware of this spend. This scenario also includes what the customer may anecdotally suspect but not know with certainty.

This data is valuable in increasing a customer’s knowledge about their data and provides important insight into sourcing and procurement opportunities. Decreasing the organization’s facade empowers the customer to be more strategic while making sourcing and procurement decisions. An organization should spend additional effort in acquiring greater knowledge about this data and improving the quality further by enforcing procurement best practices (e.g. creating item catalogs and contractual agreements with suppliers).

The Blind Area: What is known to the customer but not surfaced in the analytic tool

This quadrant represents the data with medium/moderate data quality. It includes spend transaction details, supplier details, goods and service details, and attributes, and spend patterns that are only known to the customer. The details are not included or partially included in the data made available to the analytic tool. This scenario also includes what the customer may anecdotally know but not know with certainty.

While these indirect spend details may be recorded in an organization’s ERP, AP systems, or other expense tracking systems, they may not be entirely nor accurately available to the analytic tools for spend analysis. One possible cause is fragmented data, where the data could be coming from multiple data sources that track the data differently. Another example is if the transactional details may not have fully captured all the attributes for an item.

This data is not easy to identify, enrich, classify, or analyze. Until the attributes of this data are further enhanced, this data cannot be readily used for sourcing opportunities. Additional information about these items may be needed before any sourcing actions could be taken. It is essential to focus on gathering additional details about this data as portions of the data may exist within the organization. This data affects an organization’s primary objective, such as reducing costs in certain categories, improving relationships with suppliers, consolidating suppliers where there is too much fragmentation of spend, and so forth.

Related: Preparing Your Data for Xeeva Spend Analytics

A common example is when the data might be generically listed as “IT service.” Such generic descriptions do not indicate what the related part or service was, nor do they support a good classification of data. Decreasing the data in this blind area helps surface more meaningful and reliable savings insights and allows the customer to be more strategic while making sourcing, procurement and purchasing decisions. An organization should spend more effort in gathering and storing greater knowledge about this data.

Xeeva best assists in improving the data quality in this quadrant by leveraging its AI technology and its Item and Supplier universe.

The Unknown Area: What is unknown to both the customer and not surfaced in the analytic tool

This quadrant includes data that is of poor quality. It includes item descriptions, attributes, and spend patterns that are unknown to both entities. There are multiple causes to why this data is unknown:

  • Data is not captured by the organization – organizations often depend on integrators who provide partial/incomplete information
  • Different spend data sources such as ERPs vs. AP systems capture different item descriptions and attributes, that too is not captured completely nor in a consistent pattern
  • Data is too fragmented – organizations use multiple data sources. Each source tracks partial information about an item, such as item names or item descriptions. Organizations store or extract only partial information to provide for analytics
  • High volume of buyers – each buyer tracks their items differently; thus, there is inconsistency in data captured
  • Procurement best practices are not enforced – a process may not exist within the organization, or such practices may not be enforced

Related: Managing MRO spend: overcome challenges, strategize data, and be a step ahead with next-generation technology

This data cannot be identified, enriched, classified, or analyzed. In the majority of cases, this data cannot be used to derive any analytic insights. When used, this data gives only a partial and often an unreliable diagnostic view of spend.

The organization must make all efforts to focus on gathering more information about this data and implementing best practices and compliance procedures to minimize the data that falls in this quadrant. The larger the quadrant, the greater the risks to the organizational objective.

Johari Window’s Quadrants: The Data Quality Framework

Each quadrant reflects the data quality at a given point in time. The size of the quadrant changes with every refresh over time.

At launch, we notice that our customers have a large volume of data that falls in the “Unknown” quadrant. The data quality pattern is displayed as:

The Open Area is small, the Blind Area and Hidden Area are large, and the Unknowns are intimidating.

With every incremental refresh of data, our goal is to reduce the Blind area and the Hidden areas through continuous learning and collaborative discovery. To track and monitor the Unknown area to ensure it is not becoming bigger. The key focus is on expanding the Open Area.

Through shared and collaborative discovery, learning about customers, the goods and services they procure, how they procure, and how their items are utilized, we increase our knowledge of our customer’s data. By leveraging the AI and Xeeva’s Item and Supplier universe and the shared knowledge, our goal is to reduce the Hidden and Blind areas. In addition to the continuous learnings, we reduce the size of the “Unknowns” and the potential risks in achieving business objectives while leveraging best data to surface savings insights.

We advocate for every organization to spend effort in discovery and data curation in the Blind Area and the Hidden Area to be able to use it for strategic sourcing processes and indirect procurement initiatives.

Lastly, any organization must implement best practices and enforce compliance to keep the Unknown to a minimum. While ‘Unknowns’ can never be eliminated, the goal is to constantly research and discover the unknowns and keep them to a minimum level.

Summary & Next Steps:

  • Open Area: High-quality procurement and purchasing data with required item details and attributes provides very reliable data to savings insights. Data is good to be used for strategic sourcing process and item catalogs. Xeeva uses this data to create savings insights in Spend Analytics as well as strategic sourcing opportunities.
  • Blind Area: Good to moderate quality spend data with many required attributes – provides reasonably reliable insights. Data is good to initiate strategic sourcing activities with the intent to create supplier contracts and item catalogs. Xeeva uses this data to create savings insights in Spend Analytics as well as strategic sourcing opportunities including supplier consolidation.
  • Hidden Area: Low-quality data with missing attributes provides limited or unreliable insights. Xeeva will help gather additional information and data enrichment project to better curate the data. Enrichment will lead to better insights and additional opportunities for savings.
  • Unknown: Poor quality data with few or no details and attributes – provides none to unreliable insights. With Xeeva’s help, customer should focus on implementing and enforcing best procurement practices. Focus should be on gathering and curating this data before it can be used for spend analytics and sourcing initiatives. These efforts will help drive better data to insights outcomes.

Want to learn more about how to optimize imperfect data and make it more valuable?
Check out Xeeva’s Data Enrichment

Want to get in touch? Contact us today!