Does Poor Data Quality Negatively Impact Stock Prices?

By admin

Dec

16

2025

Time to read

2 minutes

Read so far

For years those of us in the data management community argued that poor data quality has a substantial strategic and operational impact on corporations (and non-profit organizations). Unfortunately, due to the intangible nature of data and poor record keeping within organizations this argument has been difficult to support. Often, the negative public relations attributed to data issues that are under the control of management prevent truthful admission and are referred to as ‘computer glitches’ as if there was no human responsibility, almost as if it was an act of god, meaning no scientific measurement could be used to control these events.

Today, we are in a different place. Below I’ll reference a few recent examples in the news and academic studies that point out decreases in stock prices (valuation of a company’s current and future ability to beat the competition).

One example that comes to mind is Lamb Weston (LW), a potato processor, whose single day drop in stock price of 20% due to earnings miss attributed to a slower than expected ERP (enterprise resource planning system) transition.

Stock price: single day dramatic drop in stock price

An article describing the drop (permalink), said that LW stated that the ERP errors “reduced the visibility of finished goods inventories located at distribution centers, which affected our ability to fill customer orders, [which] in turn, pressured sales volume and margin performance.” 1

For those of you familiar with the Conformed Dimensions, created by DQMatter, we map common data quality problems using a framework describing data quality characteristics. This enables us to measure data quality, conduct retroactive measures to fix data quality but most importantly, identify needed preventative controls that ensure this doesn’t happen again.

The quote from LW directly referenced a lack of visibility which stems from either incomplete data or lack of access to needed data.

Here we've mapped the real-world data quality issues to the following dimensions and underlying concepts.

Conformed Dimension	Underlying Concept	Cause
Completeness	Record Population: This measures whether a row is present in a data set (table). Attribute Population: This measures whether a value is present (not null) for an attribute (column).	Typically, lack of transmission of data from legacy systems to new systems during migration/ integration of new systems. Can either be a lack of rows due to differing transaction dates or activity types that should have been moved but weren’t included.
Consistency	Equivalence of Redundant or Distributed Data: The measure of similarity with other sources of data that represent the same concept.	Typically happens when the full scope of work isn’t understood or that data received isn’t validated against the system of record.
Accessibility	Ease of Obtaining Data: This measures how easy it is to obtain data. Access Control: Access control includes the identification of a person that wants to access data, authentication of their identity, review and authorization to access required data, and lastly auditing the access of that data.	Bringing new systems online include a huge change management component with awareness, training and changes to access controls. If vendors or partners couldn’t see LW’s finished goods it’s a good chance that use of new systems was to blame, possibly due to training and access controls that prevented orders.
Timeliness	Manual Float: is a measure of the time from when an observation is made to the point it is recorded in electronic format. Electronic Float: is a measure of the time from when data is captured in an electronic format until it is accessed by a person.	Often, when systems are integrated, manual processes of data entry can lead to changes in lead time getting information in digital format (manual float), or when it can be accessed in the new system as opposed to the old system (electronic float).