Validations help you find errors in your workbook data. Obvious checks your data against rules you set and highlights cells with problems.

What Validations Do

Validations check if your data follows specific rules. When data breaks a rule, Obvious marks the cell with a colored indicator.

Three severity levels:

• Error (red) - Data is invalid or missing

• Warning (yellow) - Data looks suspicious but might be correct

• Info (green) - Informational highlight

How to Use Validations

1. Open Your Workbook - Click the workbook you want to check.

2. Ask Obvious to Validate - Type a prompt in the chat asking Obvious to validate your data.

Example prompts:

• Check if all email addresses are valid

• Make sure phone numbers are formatted correctly

• Find any missing required fields

• Check if prices are between $10 and $100

3. Review Flagged Cells - Obvious applies the validation rules to your sheet. Cells with problems show colored indicators:

• Red icon = error

• Yellow icon = warning

• Green icon = info

Hover over any indicator to see what's wrong.

4. Fix the Issues - Edit cells to fix validation errors. The indicators update automatically as you make changes.

Setting Up Automatic Validations

You can set up validations to run automatically. This is useful when you want the same validations to run every time without asking. Obvious will check your data whenever it is added or updated.

1. Go to Project Instructions: Open your project settings.

2. Add Validation Rules: Include your validation rules in the project instructions.

3. Be Specific: Clearly state what to validate. For example, "Automatically check if the 'Email' column contains valid email addresses."

4. Save Instructions: Save your project instructions.

Now, Obvious will apply these rules automatically.

Types of Validations

• Required fields - Flags empty cells that must have data.

• Email format - Checks if email addresses are valid.

• Phone format - Checks if phone numbers follow the correct pattern.

• Number ranges - Flags numbers outside your min/max limits.

• Unique values - Finds duplicate values that should be unique.

• Custom rules - Create your own validation logic for specific needs.

Common Use Cases

• Cleaning imported data - Check data quality after uploading a CSV or Excel file. Find formatting issues before you start working with the data.

• Quality control - Validate data before exporting or sharing. Make sure everything meets your standards.

• Finding incomplete records - Identify rows with missing required information so you can fill in the gaps.

Best Practices

• Start with automatic detection - Let Obvious suggest validations based on your column types. It can detect email fields, required fields, and number constraints automatically.

• Be specific in your prompts - Tell Obvious exactly which fields to check and what rules to apply. Instead of "validate my data," say "check if all email addresses in the Contact Email column are valid."

• Use severity levels wisely:

• Use error for data that must be fixed

• Use warning for data that looks odd but might be correct

• Use info to highlight specific values without implying problems

• Fix errors in batches - Sort or filter by validation status to fix similar issues together.

Tips

• Validations don't change your data. They only flag problems so you can review and fix them yourself.

• You can apply multiple validation rules to the same sheet. Each rule checks a different aspect of your data.

• If you need to remove validations, ask Obvious to clear them or replace them with new rules.