Data Points

Data correctness problems are frequently related to the data entry step.  Distracted, inattentive or overworked staff can make errors in their work.  Their data entry software may lack real-time controls to prevent typos or real-time quality checks to validate the inputs.  Good software would have thoughtfully considered how these data errors could be introduced and sought to prevent them by mistake proofing (poka-yoke).  Unfortunately, there are many systems that do not detect these sorts of problems, let alone prevent them.  

A recent article brought an extreme data entry error to light in Wasatch County, Utah.

Quoting the article "The error? An erroneous overvaluation of a single home for almost $1 billion. The 1,570-square-foot house built in 1978 on 2 acres in an unincorporated area of the county was recorded in 2019 tax rolls with a market rate value of more than $987 million and an overestimate of about $543 million in taxable value. In reality, the property should have only had a 2019 taxable value of $302,000, according to county property records. That error - which the Wasatch County assessor explained possibly occurred when a staff member may have dropped their phone on their keyboard - has resulted in a countywide overvaluation of more than $6 million and revenue shortfalls in five different Wasatch County taxing entities."

The data entry software could have prevented such large amounts for that field or the software could have had a QC system to notify supervisors of the unusually large amount and have them doublecheck it.  Similarly, QC to detect misformatted addresses (such as a zip code that was not 5 digits) or designing the data entry software to restrict data to defined domains (such as a pick list of state abbreviations) can reduce the likelihood of errors.    

In the end, emphasizing to staff members the criticality of accurate data entry and checking the correctness and completeness of their work is the best defense.  Accurate, first time right work will avoid much rework and pain later.

What are your examples of data entry problems, and how did you prevent them from reoccurring?

------------------------------
Tom Gilligan
Pitney Bowes Software, Inc.
White River Junction, VT, USA
------------------------------

Great story Tom. I once got a $1999 water bill in a similar scenario. Luckily in that case, I don't think it affected anyone else's bill.

------------------------------
Teri Dickinson
Knowledge Community Shared Account
Shelton CT
------------------------------

Original Message:
Sent: 12-09-2019 12:03
From: Tom Gilligan
Subject: How might we prevent data entry problems?

Data correctness problems are frequently related to the data entry step.  Distracted, inattentive or overworked staff can make errors in their work.  Their data entry software may lack real-time controls to prevent typos or real-time quality checks to validate the inputs.  Good software would have thoughtfully considered how these data errors could be introduced and sought to prevent them by mistake proofing (poka-yoke).  Unfortunately, there are many systems that do not detect these sorts of problems, let alone prevent them.  

A recent article brought an extreme data entry error to light in Wasatch County, Utah.

Quoting the article "The error? An erroneous overvaluation of a single home for almost $1 billion. The 1,570-square-foot house built in 1978 on 2 acres in an unincorporated area of the county was recorded in 2019 tax rolls with a market rate value of more than $987 million and an overestimate of about $543 million in taxable value. In reality, the property should have only had a 2019 taxable value of $302,000, according to county property records. That error - which the Wasatch County assessor explained possibly occurred when a staff member may have dropped their phone on their keyboard - has resulted in a countywide overvaluation of more than $6 million and revenue shortfalls in five different Wasatch County taxing entities."

The data entry software could have prevented such large amounts for that field or the software could have had a QC system to notify supervisors of the unusually large amount and have them doublecheck it.  Similarly, QC to detect misformatted addresses (such as a zip code that was not 5 digits) or designing the data entry software to restrict data to defined domains (such as a pick list of state abbreviations) can reduce the likelihood of errors.    

In the end, emphasizing to staff members the criticality of accurate data entry and checking the correctness and completeness of their work is the best defense.  Accurate, first time right work will avoid much rework and pain later.

What are your examples of data entry problems, and how did you prevent them from reoccurring?

------------------------------
Tom Gilligan
Pitney Bowes Software, Inc.
White River Junction, VT, USA
------------------------------