Auto Validation
User Guide — Running Your Data-Quality Rules Across the Data and Reviewing the Results
| Last updated: August 2026
Contents
- Overview
- Key Concept: Rules, the Run, and the Score
- Finding Your Way Around
- The Report Header
- Re-Running and Downloading
- The Summary Tab
- The By Rule Tab
- The Defects Tab
- Understanding the Score
- Quick Reference
1. Overview
Auto Validation is where you see how your data measures up against the quality rules you built in Rule Pilot. It takes every active rule for the Enterprise Object, runs them across the object's data, and reports the results three ways: an overall summary, a rule-by-rule breakdown, and a drill-down to the exact rows that failed.
Where Rule Pilot is where you write and validate rules one at a time, Auto Validation is where you run them all together and read the outcome — the scorecard for the whole object. It answers "how clean is this object's data right now, and where are the problems?"
What you can do here
- See an overall score for the object at a glance.
- Break results down by target table — how many rows pass and fail in each.
- Break results down by rule — which rules are finding the most defects, and in which quality dimensions.
- Drill into the failing rows themselves — the actual records that broke a rule, with their keys and the error.
- Re-run the checks whenever the data changes, to refresh the results.
- Download the report as Excel or PDF to share.
WHERE TO FIND IT — Auto Validation is a tab on the Enterprise Object (it may be labelled Validation, or a name your environment has set for the layer). Open the object, choose the layer, and select the tab. It works on the Enterprise Object and mapping version you've selected in the tree on the left.
RULE PILOT WRITES, AUTO VALIDATION SCORES — The two go together. You author and activate rules in Rule Pilot; Auto Validation runs those active rules across the data and shows the result. A rule only shows up here once it's Active and the checks have actually been run.
2. Key Concept: Rules, the Run, and the Score
Three things make sense of everything on this page.
The rules come from Rule Pilot. Only rules that are Active (validated) take part — drafts and inactive rules are ignored. So what you see here reflects the checks you've actually put into force.
The run is the moment the rules are executed across the object's data. Auto Validation shows the results of the most recent run; it doesn't re-run automatically. When the data (or the rules) have changed, you trigger a fresh run with Re-Run (see section 5).
The score is how the results are summarised — the percentage of rows that pass. It's shown for the object as a whole, per table, and per rule, and it's colour-coded so a glance tells you where attention is needed. (The exact formula and colour bands are in section 9.)
THE REPORT IS A SNAPSHOT — What you're looking at is the result of the last run, not a live re-check. If you've just changed data or rules, the numbers won't move until you Re-Run.
3. Finding Your Way Around
Open the Enterprise Object, choose the layer, and select the Auto Validation (Validation) tab. Because the report is scoped to one object and one mapping version, you first pick those from the tree on the left; the report then loads for that selection.
The page has three parts:
- The header — the object name, key facts about the run (layer, version, tables, totals), an overall score gauge, and the Re-Run and Download buttons.
- The three tabs — Summary, By Rule, and Defects.
- The working area — the content of whichever tab is active.
| Tab | What it answers |
|---|---|
| Summary | How is the object doing overall, and which tables and rules are the weak spots? |
| By Rule | Rule by rule, how many defects did each find, and in which dimension? |
| Defects | Exactly which rows failed — down to the individual record and the rule error. |
PICK AN OBJECT AND VERSION FIRST — Until you've selected an Enterprise Object and a mapping version in the tree, the report shows a prompt to do so. The header's Layer and Version tell you what the current results belong to.
4. The Report Header
The header sums up the current run at a glance:
| Item | What it shows |
|---|---|
| Object name | The Enterprise Object the report is for. |
| Layer | Which layer the results belong to. |
| Version | The mapping version that was run. |
| Tables | The object's target tables covered by the run. |
| Total rows | How many rows were checked in all. |
| Defective | How many rows failed at least one rule (shown in red). |
| Score gauge | The object's overall quality score as a dial. |
To the right sit the actions: Re-Run and Download, covered next.
5. Re-Running and Downloading
5.1 Re-running the checks
Re-Run RulePilot schedules a fresh run of the object's rules against the current data. Use it whenever the data has changed, or after you've activated or edited rules in Rule Pilot, so the report reflects the latest state. The run happens in the background; the report refreshes shortly after, showing the new results.
5.2 Downloading the report
The Download button offers two formats:
| Format | Best for |
|---|---|
| Excel (.xlsx) | Working with the numbers — the summary, the per-rule breakdown, and defect rows in a spreadsheet. Built on the server so it handles large defect sets. |
| Sharing a fixed, readable copy of the report as it looks on screen. |
The file is named after the object, version, and date so downloads stay organised.
RE-RUN AFTER THINGS CHANGE — The report doesn't refresh on its own. New rules, edited rules, or new data all call for a Re-Run before the scores and defects are up to date.
6. The Summary Tab
The Summary tab is the object-level scorecard: a results table broken down by target table, followed by charts that point straight at the weak spots.
6.1 The per-table table
One row per target table, with:
| Column | What it shows |
|---|---|
| Table | The target table. |
| Total rows | How many rows were checked in that table. |
| Pass | How many rows passed every rule (green). |
| Fail (unique) | How many distinct rows failed at least one rule (red). A row that breaks several rules is counted once. |
| Score | The table's quality score, as a colour-coded pill. |
6.2 The charts
Below the table, three charts turn the numbers into a picture:
- Pass vs Fail by table — a stacked bar per table, green for passing rows and red for defective rows, so you see at a glance which tables carry the most defects.
- Defects by dimension — a donut showing where defects fall across quality dimensions (Accuracy, Validity, Completeness, and so on). A row that fails rules in more than one dimension is counted once, under whichever dimension ranks highest in your environment's dimension order (see the note below).
- Top problem rules — the ten rules finding the most defective rows, bar-coloured by score (green, amber, or red), so the biggest offenders are obvious.
WHICH DIMENSION "WINS" A ROW — When a row fails rules in several dimensions, it's attributed to a single dimension so it's never double-counted. Which one wins follows the priority order defined in your Data Quality dimension settings — it's configurable per environment, not a fixed list. Check that configuration to see the order in effect.
7. The By Rule Tab
The By Rule tab breaks the results down one rule at a time — the view for deciding which rules to act on.
7.1 The rule table
| Column | What it shows |
|---|---|
| Rule | The rule's name. |
| Dimension | The quality dimension it measures. |
| Column | The column it checks. |
| Tables | The tables it applies to. |
| Defective rows | How many rows failed this rule. |
| Score | The rule's score, as a colour-coded pill. |
| View failing rows | Jumps to the Defects tab, pre-filtered to just this rule's failures. |
You can sort and filter the table by any column to find, say, all the Validity rules or the rules with the lowest scores.
7.2 The score distribution chart
Above the table, a Rule score distribution chart shows how many rules sit in each score band (red, amber, green). A healthy object clusters to the right (green); a spread toward the left flags a lot of rules needing attention.
7.3 When the tab is empty
If no rules appear, it usually means one of two things: the rules aren't Active yet (validate them in Rule Pilot), or the checks haven't been run for this version yet (use Re-Run). The tab says as much when it's empty.
8. The Defects Tab
The Defects tab is the drill-down to the actual failing records — the rows you'd hand to someone to fix.
Each row shows:
| Column | What it shows |
|---|---|
| Table | The target table the row is in. |
| Row ID | The row's identifier. |
| Key 1 / Key 2 / Key 3 | The row's business keys, so you can find the record in the source. |
| Dim | The dimension of the failure. |
| Rule IDs | Which rule(s) the row broke. |
| Rule error | A description of what went wrong. |
You reach this tab either directly (to see all failing rows) or from View failing rows on the By Rule tab (pre-filtered to one rule). When a filter is active, the top of the tab shows it and offers Clear filter to widen back to all rules. You can also sort and filter the grid itself.
THE DEFECTS LIST IS CAPPED — The tab shows up to 500 rows at a time. When an object has more defects than that, the grid is showing a sample, not the complete set — narrow with the rule filter to focus, or use the Excel download (which covers a much larger defect set) for the full picture.
9. Understanding the Score
The score is the percentage of rows that pass, worked out as:
round( ( 1 − unique defective rows ÷ total rows ) × 100 )
A row that fails several rules still counts as just one defective row, so the score reflects how many records are clean rather than how many rule-failures occurred. A score of 100 means every row passed; a lower score means some rows have problems.
Scores are colour-coded consistently across the header gauge, the table pills, and the charts:
| Score | Colour | Reading |
|---|---|---|
| 90–100 | Green | Healthy — the large majority of rows pass. |
| 60–89 | Amber | Needs attention — a meaningful share of rows fail. |
| Below 60 | Red | Poor — a large share of rows fail. |
ONLY ACTIVE RULES COUNT — The score reflects the rules that were actually run — the Active ones. Draft or inactive rules in Rule Pilot don't affect it. If a score looks better or worse than you expect, check which rules are active and when the object was last run.
10. Quick Reference
A fast lookup for the most common actions.
| I want to… | Do this |
|---|---|
| Open Auto Validation | Enterprise Object → choose the layer → Validation tab |
| Load the report | Select the object and mapping version in the left-hand tree |
| See the object's overall score | Read the score gauge in the header |
| See which tables are worst | Summary → per-table table and the Pass vs Fail chart |
| See where defects concentrate | Summary → Defects by dimension and Top problem rules |
| See how each rule did | By Rule tab |
| See the rows a rule failed | By Rule → View failing rows |
| See all failing rows | Defects tab |
| Widen a filtered defect list | Clear filter on the Defects tab |
| Get the complete defect set | Download → Excel (covers far more than the 500-row on-screen cap) |
| Refresh after data or rule changes | Re-Run RulePilot |
| Share the report | Download → Excel or PDF |
| Fix an empty By Rule tab | Activate the rules in Rule Pilot, then Re-Run |
Source: OnCoor Auto Validation product documentation, written for end users. For the latest screens and options, always refer to the in-app interface.