Get Inspire insights from former attendees in our AMA discussion thread on Inspire Buzz. ACEs and other community members are on call all week to answer!

Location Data Knowledge Base

Data methodologies, and Release schedules.

Notice: Defect Identified within Dun and Bradstreet Summary Variables Q3 2021

CSalazar
Alteryx
Alteryx
Created

This data defect only impacts users that have the Q3 US Business Insights package installed:


After a deeper analysis we have discovered a defect in our US Business Insights package within our US Installer. This package provides users the ability to analyze business locations, types of business', numbers of employees and size of employer. We have determined this invalid data impacted our Employee and Establishment variable counts during our build processes. The variables that have outlying counts include:

  • Total Employees

  • Construction (23)

  • Construction (15-17)

  • Building Cnstrctn - General Contractors & Operative Builders (15)

  • Residential Building Construction (2361)

Attached below is a full list of variables discovered that have invalid counts.

Q3_2021_DnB_Defect_Variables

 

Resolution

We have resolved and validated this data defect for our upcoming quarterly data release (Q4 2021) for both our CA and US Business Insights. We have taken measures to prevent this from impacting our future Business Insights data package updates. The corrected variable counts are valid and will be delivered for our upcoming Q4 2021 Business Insights package release.

Please let us know if you have any questions. 



Additional Resources

Comments
mbarone
16 - Nebula
16 - Nebula

Thanks @CSalazar .  So for those of us that do time or historical analyses, are you saying the original Q3 CYDB will not be updated with the correct values?  Is there a possibility to have a one-time install that updates the CYDBs for Q3?  Maybe a file with a corrected CYDB that we can replace the original one with?  Is this possible?

mweik
Moderator
Moderator

Hello @mbarone , if you're primarily concerned with the CYDB version of this data, then you don't have to worry about these errors.  We identified the invalid values for our Allocate "Business Summary" variables only and haven't had any issues with our "Business List"/"Analytical File" version of this data.  I apologize for the lack of clarity in this post leading to the confusion.

mbarone
16 - Nebula
16 - Nebula

Thanks @mweik!  But isn't this is still a CYDB sitting out there with bad data though?  If not, what is the "Experian US 2021" data file then?

2022-02-15 08_15_59-Clipboard.png

 

 

mweik
Moderator
Moderator

The Experian US 2021 file is an Allocate dataset created by attributing data to a Census Block Group and leveraging the population and household counts for each Census Block centroid within a Census Block Group to provide a Block centroid weighted result for any polygon.  The data for our Business Summary variables originates as the same data as our CYDB file, but we use Alteryx to aggregate values for each block group, map out to the block group, census tract, county, and state. Since its not accurate to use household or population counts for calculating block centroid weighted results of firmographic data. The Business Summary data only maps to the Census Block Group and doesn't leverage Allocate's ability to estimate values using Census Block centroids.  This data is stored in Demo files and is more difficult to install without the entire Allocate dataset. We are only able to provide the current data for download as you have maybe seen we don't offer previous versions for download, and the Q4 2021 update was released on February 7th. I know this can impact those who are using the data to run a time or history based analysis.  We have traditionally not supported using this data for time/history analysis, or selling previous data updates, as the methodologies our data providers use for individual variables can change and new data models have been introduced between releases. 

mbarone
16 - Nebula
16 - Nebula

Wow, thank you @mweik !  That is probably the most detailed, concise, and very easily understandable explanation I've ever heard for this data.

 

I'm assuming what you just summarized is what the PDF "Allocate Block Centroid Retrieval Methodology.pdf" is intended to convey.  However, I only make that conclusion after re-reading the PDF in light of what you just said.

 

I would highly recommend that your summary above be added as a first "high level summary" paragraph to that document, and then everything beneath it be labeled as "further detail".

 

I understand much better now, thanks again!

 

 

mweik
Moderator
Moderator

I greatly appreciate the feedback @mbarone !  Yes that PDF is an attempt to explain this information but gets a little lost in its own details.  I will look into updating that document so that it can be understood quicker by others as well.  Thank you!