What is a product score?

πŸ§ͺ Preview feature! This feature is available for your experimentation, and we'd love your feedback. It may change before its final generally-available form.

How can Atlan help you build trust in your data products? Product scores!

Based on the principles of data as a product, product scores can help you signal the accuracy and completeness of your data products, helping build trust in them. The product scorecard in Atlan enables you to:

  • Identify and prioritize data quality issues
  • Quantify the business impact of data quality
  • Drill down for detailed analysis and take action
  • Share insights across stakeholders

Components of a product score

Atlan calculates and assigns a product score to your data products based on a preset criteria of metadata completeness. Atlan evaluates metadata enrichment on the data product, output ports, or a combination of both. Using a weighted scoring method, values are automatically assigned on the basis of how well your data product satisfies the six principles of data as a product. Atlan scores your data product on a scale of 0-5, with 0 being the lowest and 5 being the highest score.

The six principles of data as a product are:

  • Discoverable: The discoverability of a data product is based on the availability of aggregate metadata. Atlan quantifies discoverability in the form of glossary terms linked to the data product and output ports.
  • Understandable: A data product is considered to be understandable when it has defined and curated contextual metadata to help your data consumers understand the product better. Atlan quantifies how understandable a product is in the form of:
  • Addressable: A data product is considered to be addressable when it has well-documented owners and points of contact. Atlan quantifies how addressable a data product is in terms of owners assigned to the data product and its output ports.
  • Secure: A secure data product will clearly signify the sensitivity of data. Atlan quantifies how secure a data product is on the basis of sensitivity classifications on the data product and tags attached to its output ports.
  • Interoperable: Interoperability of a data product is determined on the basis of the visibility and completeness of technical lineage between data assets.
  • Trustworthy: Trustworthiness of a data product is determined on the basis of certificates and data contracts attached to the data product.

A product score is not set in stone, and will change depending on the completeness of metadata enrichment. The score can help you understand how to improve your data product to make it more useful for your data consumers.

Atlan currently does not allow you to either configure the score or define your own criteria.

Scoring rubric

Atlan adheres to the following scoring rubric to assign product scores:

Principle 0 1 2 3 4 5
Discoverable No terms linked to product β€” β€” β€” β€” At least 1 term linked to product
Understandable

No description or README on product or output ports AND no README on output ports

Description on product AND
≀ 10% of output ports have descriptions but no READMEs

OR

No description on product or output ports AND product has a README

Description on product AND 11%-40%
output ports have descriptions but no READMEs

OR

Description on product AND
< 10% of output ports have descriptions AND product has a README

Description on product AND 41%-70% output ports have descriptions but no READMEs

OR

Description on product AND 11%-40%
output ports have descriptions AND product has a README

Description on product AND 71%-90% output ports have descriptions but no READMEs

OR

Description on product AND 41%-70% output ports have descriptions AND product has a README

Description on product AND >90% output ports have descriptions but no READMEs

OR

Description on product AND 71%-90% output ports have descriptions AND product has a README

Addressable No output ports in the product

Product does not have an owner AND
≀ 10% of output ports have owners

OR

Product has an owner but output ports do not have owners

Product has owners AND 11%-40%
output ports have owners
Product has owners AND 41%-70% output ports have owners Product has owners AND 71%-90% output ports have owners Product has owners AND >90% output ports have owners
Secure Product does not have sensitivity classification AND no tags attached to output ports

Product does not have sensitivity classification AND ≀ 10% of output ports have tags attached

OR

Product does not have sensitivity classification AND
no tags attached to output ports

Product has sensitivity classification AND 11%-40%
output ports have tags attached
Product has sensitivity classification AND 41%-70% output ports have tags attached Product has sensitivity classification AND 71%-90% output ports have tags attached Product has sensitivity classification AND >90% output ports have tags attached
Interoperable None of the output ports have technical lineage ≀ 10% of output ports have technical lineage 11%-40% of
output ports have technical lineage
41%-70% of
output ports have technical lineage
71%-90% of
output ports have technical lineage
> 90% of
output ports have technical lineage
Trustworthy Product neither has a certificate nor a contract β€” β€”

Product does not have a certificate but has a contract

OR

Product has a certificate but no contract

β€” Product has a certificate and contract

Scoring method

To calculate the product score, Atlan uses a weighted scoring method:

  1. Assign scores for each principle β€” Atlan assigns a score based on each of the principles to determine a product score within the range of 0-5.
  2. Set weights for each principle β€” Atlan determines the weight of each principle based on the completion rate of metadata enrichment.
  3. Calculate weighted scores β€” the score for each principle is multiplied by its weight. For example, if Trustworthy has a score of 4 on a scale of 5 and a weight of 30%, the weighted score would be 4 * 0.3 = 1.2.
  4. Sum up the weighted scores β€” Atlan adds up the weighted scores for each principle to arrive at a total score, which is displayed on the data product.
  5. Interpret the score β€” you can use the total weighted score to evaluate the data product's overall alignment with the principles of data as a product. A higher score will indicate closer alignment with the principles and metadata completion.

Related articles

Was this article helpful?
0 out of 0 found this helpful