Managed data
What is Data as a Service (DaaS) and why it matters
A clear definition of Data as a Service (DaaS): what it includes, how it differs from building your own infrastructure and why more companies adopt it.
Read articleKey takeaways
- Data quality is measured in dimensions: accuracy, completeness, consistency, validity, uniqueness, timeliness.
- Without quality, any report or AI model inherits the data’s errors.
- Measuring requires automated indicators and rules, not manual review.
- Quality is a continuous process, not a one-off project.
- It is especially critical for AI.
"The numbers do not match" is one of the most expensive phrases in any organisation. Behind it is usually a data quality problem that erodes trust in reports and compromises decisions.
The six dimensions
Data quality is not abstract: it breaks down into measurable dimensions. Management frameworks such as DAMA-DMBOK identify several, and these six are the most used in practice.
- Accuracy: the data reflects reality.
- Completeness: no mandatory values missing.
- Consistency: the same data matches across systems.
- Validity: it meets the defined format and rules.
- Uniqueness: no duplicates distort it.
- Timeliness: the data is available when needed.
How to measure it
Measuring requires turning each dimension into indicators and automated rules — percentage of complete mandatory fields, duplicate rate, out-of-range values — run continuously over the data flows, not in sporadic manual reviews.
How to improve it
Improvement combines prevention and correction: validate at the entry point so bad data never enters, and deduplicate, normalise and enrich what already exists. It needs clear owners — without a data owner, quality degrades over time.
Why it matters for AI
Data quality is especially critical for AI. A model trained or fed on incomplete, biased or inconsistent data produces unreliable results, however good the technology. "Garbage in, garbage out" still holds.
A model is only as good as the data it learns from — quality is the foundation of everything.
In summary
Data quality breaks into six measurable dimensions — accuracy, completeness, consistency, validity, uniqueness, timeliness — measured with automated rules, not manual reviews. It is a continuous process needing owners, and it is the foundation of reliable reporting and trustworthy AI.
Sources & further reading
Frequently asked questions
How do I start measuring data quality?
Pick the business-critical datasets and define measurable rules for the most relevant dimensions. Automate them and review the indicators regularly.
Is data quality IT’s responsibility?
It is shared. IT provides the tools and automation, but the business defines what a correct data point is and names owners per domain.
Is it a project or a process?
A continuous process. Sources change and data degrades, so quality must be monitored and maintained permanently.
What are the six dimensions?
Accuracy, completeness, consistency, validity, uniqueness and timeliness — the most used measures of data quality.
Why does it matter for AI?
AI learns and scales the data’s defects. Incomplete, biased or inconsistent data produces unreliable models, however good the technology.
How do I improve quality?
Prevent (validate at entry) and correct (deduplicate, normalise, enrich), with clear data owners and continuous monitoring.
Turn this data into results
Tell us what you want to achieve. Data Layer connects, processes and delivers the result up and running, with no infrastructure for you to manage.