Categorization Strategies

Different business needs require different categorization approaches. This guide explains three core strategies—when to use each, how to configure them, and what results to expect.

Understanding the Trade-offs

All categorization strategies balance three factors:

Coverage: What percentage of your products get a category assigned

Accuracy: How often the assigned category is correct and appropriate

Specificity: Whether categories are detailed (most specific level) or general (may include broader categories)

No single configuration optimizes all three. Your strategy depends on which factors matter most for your store.

Strategy 1: Maximum Coverage

Priority: Get a category for every product, even if confidence is lower

Best for:

• Initial catalog imports where you’ll review results afterward
• Large catalogs where you need categories on everything quickly
• Situations where having an approximate category is better than none
• Products with brief or inconsistent descriptions

Configuration

Navigate to Settings in Categorify:

• Use best-guess category if AI cannot decide: ☑ Enabled
• Return only leaf categories: ☐ Disabled
• Instructions to the AI: Optional

How It Works

With “Return only leaf categories” disabled, Categorify can choose both specific and general categories. The AI picks the appropriate level based on how certain it is—specific when confident, general when uncertain.

With “Use best-guess category” enabled, even when the AI isn’t confident, Categorify returns the top-ranked option rather than no category.

This produces maximum coverage because you’ll get results in nearly every scenario:

• Green status badges when the AI confidently selects a category
• Yellow status badges when the system uses the top-ranked guess
• Gray status badges only in rare cases where no relevant categories exist

What to Expect

Coverage: Nearly every product will receive a category.

Accuracy: Green results will be fairly accurate, but yellow results may have lower accuracy. The top-ranked fallback mechanism prioritizes getting some category over perfect precision.

Specificity: You’ll get a mix of specific and general categories. Categorify chooses detailed categories when certain, broader categories when unsure.

Status badges: Expect mostly green and yellow results with very few gray results. Yellow represents products where the system fell back to its best guess rather than a confident AI selection.

When to Use

Use maximum coverage when:

• You’re categorizing a large catalog and will review results afterward
• You prefer to manually correct wrong categories rather than have missing ones
• Google Shopping or other platforms require every product to have a category
• Product descriptions are inconsistent or brief

Limitations

• Yellow results have lower accuracy than green results
• Some clearly incorrect categories will appear in yellow results
• Requires manual review and correction for production use
• Not suitable when categorization accuracy is critical

Strategy 2: Precision First

Priority: Only assign categories when highly confident they’re correct

Best for:

• Stores where accuracy is more important than coverage
• High-value products where miscategorization causes problems
• When you’re willing to manually categorize uncertain products
• Products with detailed, well-written descriptions

Configuration

Navigate to Settings in Categorify:

• Use best-guess category if AI cannot decide: ☐ Disabled
• Return only leaf categories: ☐ Disabled
• Instructions to the AI: Recommended for edge cases

How It Works

With “Use best-guess category” disabled, Categorify only assigns categories the AI actively selects with confidence. No guessing based on scores.

With “Return only leaf categories” disabled, the AI can choose broader categories when more specific ones are unclear, ensuring assigned categories are semantically appropriate.

This produces high accuracy but lower coverage:

• Green status badges when the AI confidently selects any category (specific or general)
• Gray status badges when the AI cannot confidently decide
• No yellow status badges (best-guess fallback is disabled)

What to Expect

Coverage: Significantly lower coverage than maximum coverage strategy—a meaningful portion of products will return no category.

Accuracy: Green results should be highly accurate since Categorify only assigns categories it confidently selects. No yellow results exist with this strategy.

Specificity: You’ll get a mix of specific and general categories, all actively chosen by the AI. The system chooses broader categories when more specific siblings are unclear, ensuring semantic correctness.

Status badges: Expect only green and gray results. Green indicates confident AI selection, gray indicates the AI couldn’t make a confident determination.

When to Use

Use precision-first when:

• Accuracy is more important than coverage
• You’re willing to manually categorize products that return no category
• Incorrect categorizations cause problems with your sales channels
• You have time to improve descriptions for gray-result products

Handling Gray Results

Products that return gray status badges need attention:

1. Improve descriptions: Add context, features, or clarifying details
2. Add custom instructions: Provide disambiguation rules for recurring patterns
3. Manual categorization: Assign categories by hand if description can’t be improved

Limitations

• A significant portion of products return no category initially
• Requires investment in description quality and custom instructions
• Higher manual effort than maximum coverage strategy
• Not suitable when complete coverage is required

Strategy 3: Specific Categories Required

Priority: Always assign the most specific categories available

Best for:

• Platforms that require detailed categories (no general categories)
• Product feeds with strict specificity requirements
• When business rules mandate specific classifications
• Catalogs where most products clearly fit a detailed category

Configuration

Navigate to Settings in Categorify:

• Use best-guess category if AI cannot decide: ☑ Enabled
• Return only leaf categories: ☑ Enabled
• Instructions to the AI: Recommended

How It Works

With “Return only leaf categories” enabled, only the most specific categories are considered—no broader categories.

With “Use best-guess category” enabled, if the AI can’t decide confidently between specific options, the top-ranked specific category is assigned.

This guarantees specific categories but sacrifices some accuracy:

• Green status badges when the AI confidently selects a specific category
• Yellow status badges when the top-ranked specific category is used as fallback
• Gray status badges only when no specific categories exist (rare)

What to Expect

Coverage: Very high coverage—nearly every product will receive a specific category. Gray results should be rare.

Accuracy: Green results should be fairly accurate. Yellow results will have noticeably lower accuracy since they represent forced choices between unclear options. Expect a higher rate of incorrect-but-specific categorizations in yellow results.

Specificity: All assigned categories will be at the most specific level—no general categories are possible.

Status badges: Expect a substantial number of yellow results alongside green results. Yellow indicates products where the system had to choose between unclear specific options using top-ranked scores.

When to Use

Use specific categories required when:

• Your sales channels reject general categories
• Business rules require the most specific classification
• You’re feeding product data to systems with strict category requirements
• You prefer specific (even if occasionally wrong) over general categories

Managing Yellow Results

With a substantial number of yellow results, accuracy management is critical:

1. Custom instructions reduce yellow results: Provide disambiguation rules for common unclear products
2. Improve descriptions: Add details that distinguish between similar specific categories
3. Accept some imprecision: Yellow specific results still meet platform requirements for specificity
4. Review before finalizing: Check yellow results before publishing to sales channels

Limitations

• High percentage of yellow results (lower confidence)
• Some incorrect specific categorizations in yellow results
• Requires robust custom instructions to manage unclear cases
• Cannot fall back to safe general categories
• Not ideal when semantic correctness matters more than specificity

Comparing the Strategies

Choosing Your Strategy

Ask yourself these questions:

1. Do my sales channels reject general categories?

• Yes → Specific Categories Required
• No → Continue to question 2

2. How important is accuracy vs coverage?

• Accuracy critical → Precision First
• Coverage critical → Maximum Coverage
• Both important → Specific Categories Required

3. Can I manually review and correct results?

• Yes, extensive review → Maximum Coverage
• Yes, some review → Specific Categories Required
• No, must be accurate → Precision First

4. How good are my product descriptions?

• Excellent, detailed → Precision First or Specific Categories
• Mixed quality → Maximum Coverage
• Brief, inconsistent → Maximum Coverage

5. What percentage of errors can I tolerate?

• Very few errors → Precision First
• Some errors acceptable → Maximum Coverage or Specific Categories

Testing Your Strategy

Before committing to a strategy:

1. Categorize a sample: Start with 50-100 products representing your full catalog
2. Review status badges: Understand your green/yellow/gray distribution
3. Manually verify accuracy: Check a sample of results against expectations
4. Observe patterns: Note which product types succeed or fail with each strategy
5. Compare against goals: Does this strategy meet your requirements?

If results don’t match your needs:

• Try a different strategy
• Improve product descriptions
• Add custom instructions for edge cases
• Use different strategies for different product types

Next Steps

After selecting a strategy:

1. Configure Settings: Apply your chosen configuration in Categorify Settings
2. Write custom instructions: Handle edge cases specific to your catalog
3. Review History: Monitor results and adjust based on patterns
4. Enable automatic categorization: Once satisfied, enable background processing

Your strategy isn’t permanent—adjust as your catalog, descriptions, or requirements change.