Using content suggestions and batch optimization

Review, edit, and publish AI-generated content improvements in bulk

Last updated: Nov 18, 2025

Content suggestions is where you review, edit, and publish AI-generated improvements to your products, collections, blog articles, and static pages. It provides a streamlined workflow for managing bulk content optimization.

Plan requirements

Starter plan: Create 1 optimization batch per month Growth plan: Create 5 optimization batches per month Business plan and above: Create unlimited optimization batches

When you create a batch, you'll see your current usage and limit. Batch limits reset monthly on your billing date.

Upgrade your plan

What are content suggestions

When you run content optimization agents or create optimization batches, the AI generates improved versions of your content. These suggestions appear in the content suggestions interface where you can:

  • View side-by-side comparisons (original vs suggested)
  • Understand why changes were recommended
  • Edit suggestions before publishing
  • Approve or reject in bulk
  • Track what's been published and what's pending

Think of it as your content review queue where AI-generated improvements wait for your approval.

Accessing content suggestions

To view your suggestions:

  1. Click Content Hub in the left sidebar
  2. Select Content suggestions from the menu
  3. You'll see all pending and published suggestions
Go to content suggestions

Understanding the suggestions table

The suggestions table shows:

Entity name - Which product, collection, or page the suggestion is for Entity type - Product, collection, blog article, or static page Field - Which content field (title, description, meta title, meta description) Status - Pending review, published, or rejected Confidence - AI confidence score (0-100) in the suggestion quality Date created - When the suggestion was generated Actions - Review, edit, publish, or reject

Filtering suggestions

Use filters to manage your review queue:

By status:
  • Pending - Suggestions waiting for your review
  • Published - Suggestions you've applied to your store
  • Rejected - Suggestions you've declined
By entity type:
  • Products only
  • Collections only
  • Blog articles only
  • Static pages only
By field:
  • Titles
  • Descriptions
  • Meta titles
  • Meta descriptions
Search: Type entity name to find specific suggestions

These filters help you batch review similar suggestions together. For example, you might review all pending product meta descriptions at once.

Creating optimization batches

To generate suggestions for multiple items at once:

Step 1: Start a new batch

  1. Click Create optimization batch at the top
  2. The batch creation modal opens

Step 2: Select what to optimize

Choose entity type:
  • Products
  • Collections
  • Blog articles
  • Static pages
Choose specific items:
  • All items (entire catalog)
  • Specific items (select from list)
  • Filter by performance (traffic, revenue ranges)
  • Select manually (pick individual items)
Choose which fields to optimize:
  • ☐ Title
  • ☐ Meta title
  • ☐ Meta description
  • ☐ Description (HTML description for products/collections)

Select as many fields as you want. Each generates a separate suggestion.

Step 3: Set optimization focus

Choose what the AI should prioritize:

SEO - Optimize for search engines
  • Keyword targeting
  • Search intent matching
  • Character limits for meta fields
  • Schema and structured data considerations
Conversion - Optimize for purchases/actions
  • Persuasive language
  • Clear value propositions
  • Call-to-action strength
  • Benefit-focused messaging
Readability - Optimize for clarity
  • Simple, clear language
  • Shorter sentences
  • Better structure
  • Easier to scan
Engagement - Optimize for interaction
  • Interesting, compelling copy
  • Emotional connection
  • Storytelling elements
  • Reader interest

Step 4: Review credit estimate

The system calculates total credits needed based on:

  • Number of items selected
  • Number of fields per item
  • Each optimization costs 3 credits
Example: Optimizing title and meta description for 50 products = 50 items × 2 fields × 3 credits = 300 credits

Review the cost before proceeding. You can reduce scope if needed.

Step 5: Create batch

Click Create batch to start generation.

Processing time:
  • Small batches (under 20 items): 2-5 minutes
  • Medium batches (20-100 items): 5-15 minutes
  • Large batches (over 100 items): 15-30 minutes

You can leave the page during processing. You'll get a notification when suggestions are ready for review.

Reviewing suggestions

Once suggestions are generated, review them before publishing:

View mode

Click any suggestion to open the review modal:

Original content - Current content on your store Suggested content - AI-improved version AI reasoning - Why changes were recommended
  • What was improved
  • What keywords were added
  • Why it's better for SEO or conversion
  • Confidence score explanation
Confidence score - How confident the AI is (0-100)
  • 90-100: High confidence - minor risk
  • 70-89: Good confidence - review recommended
  • 50-69: Moderate confidence - review carefully
  • Under 50: Low confidence - edit or reject
Metadata:
  • Character count (for meta fields)
  • Word count
  • Readability score
  • SEO score

Edit mode

To customize a suggestion before publishing:

  1. Click Edit in the review modal
  2. The suggested content becomes editable
  3. Make changes as needed
  4. Preview your edited version
  5. Click Save changes

This combines AI intelligence with your expertise, letting you refine suggestions to match your brand voice perfectly.

When to edit:
  • Suggestion doesn't match brand voice
  • Need to add specific product details AI doesn't know
  • Want to adjust messaging for your audience
  • Prefer different wording
  • Need to include specific keywords

HTML content

For description fields that support HTML:

Preview mode - See how the formatted content will look HTML editor - Edit the raw HTML if needed Visual editor - WYSIWYG editing for easier changes

The review modal renders HTML properly so you can see exactly how it will appear on your store.

Publishing suggestions

After reviewing, you can publish (apply to your store):

Publish individual suggestions

  1. Open the suggestion in review mode
  2. Click Publish at the bottom
  3. Confirm you want to apply this change
  4. The suggestion is marked as published
  5. Content updates on your connected platform within minutes

Publish in bulk

To apply multiple suggestions at once:

  1. Filter to show pending suggestions
  2. Select checkboxes next to suggestions you want to publish
  3. Click Publish selected at the top
  4. Review the list of changes
  5. Confirm bulk publish
  6. All selected suggestions are applied
Bulk publish is powerful - review carefully before using it. You can bulk publish similar suggestions (like all meta descriptions) after spot-checking a few for quality.

Auto-publish with confidence threshold

For high-confidence suggestions you trust:

  1. Select suggestions with confidence score above 90
  2. Publish them in bulk without individual review
  3. Monitor results to ensure quality

This saves time but requires trust in the AI. Start conservative and increase auto-publish threshold as you gain confidence.

Rejecting suggestions

If a suggestion isn't good:

Reject individual suggestions

  1. Open the suggestion
  2. Click Reject
  3. Optionally provide a reason:

- Doesn't match brand voice - Factually incorrect - Prefer original - Other (specify)

  1. Suggestion is marked as rejected

Rejection reasons help improve future suggestions by teaching the system what you don't want.

Reject in bulk

Select multiple suggestions and click Reject selected. Use this for suggestions from a batch that missed the mark.

Delete rejected suggestions

Rejected suggestions stay in the list for reference. To remove them:

  1. Filter by status: Rejected
  2. Select all
  3. Click Delete selected
  4. Confirm deletion

This cleans up your suggestions list.

Tracking published changes

After publishing:

Published status - Suggestions show when they were published Change history - View what changed and when Rollback - If you published by mistake, you can revert:
  1. Find the published suggestion
  2. Click three-dot menu
  3. Select Revert to original
  4. Confirm rollback
  5. Original content is restored
Performance tracking - After publishing, monitor product/page analytics to see if changes improved performance.

Batch management

View and manage all your batches:

Recent batches - See batches you've created Batch status:
  • Processing - Suggestions being generated
  • Complete - All suggestions ready
  • Partial - Some suggestions failed
Batch actions:
  • View all suggestions from a batch
  • Publish all from batch (if confident)
  • Reject all from batch (if quality is poor)
  • Delete batch (removes all suggestions)

This helps organize suggestions when you're running multiple optimization projects.

Best practices

Review before publishing - Always check suggestions, especially for important products or brand-sensitive content. Start small - Create a small batch (5-10 items) first to evaluate quality before doing your entire catalog. Edit when needed - Don't feel obligated to use suggestions as-is. Edit them to better match your needs. Provide rejection reasons - This feedback helps improve future suggestions. Publish strategically - Publish during off-peak hours in case something needs adjustment. Monitor results - After publishing, track performance to see if changes helped. Use confidence scores - Focus manual review time on lower-confidence suggestions. Batch similar items - Review all product meta descriptions together, then all titles, etc. Easier to maintain consistency.

Credit management

Content suggestions use credits when generated, not when published:

Generation cost:
  • 3 credits per field per item
  • Product title optimization = 3 credits
  • Meta title + meta description for one product = 6 credits
No cost for:
  • Reviewing suggestions
  • Editing suggestions
  • Publishing or rejecting
  • Deleting suggestions
Credit refunds:
  • If batch generation fails, credits are refunded
  • If individual suggestion fails, that item's credits are refunded
  • Published/rejected suggestions don't get refunds (generation completed successfully)

Troubleshooting

Suggestions not generating

If batch creation fails:

  • Check you have enough credits
  • Verify platform connection is active
  • Ensure items exist (not deleted from your store)
  • Try smaller batch size
  • Check if specific items are causing issues

Poor quality suggestions

If suggestions don't meet expectations:

  • Verify you selected appropriate optimization focus
  • Check if original content is complete (AI needs context)
  • Ensure product/page has enough existing content to work with
  • Try different optimization focus
  • Edit suggestions to improve them

Publishing fails

If published suggestions don't apply:

  • Check platform connection is active
  • Verify ConvertMate has write permissions
  • Ensure item still exists on your platform
  • Try publishing individual items instead of bulk
  • Check platform for errors or maintenance

Content not updating on store

If published but store doesn't show changes:

  • Wait up to 15 minutes for sync
  • Clear your store's cache
  • Check if platform caches content
  • Manually refresh the product/page in your store admin
  • Verify the update in ConvertMate shows as completed

Workflow examples

Example 1: Optimize top 50 products

  1. Create batch: Products, top 50 by revenue
  2. Fields: Meta title + meta description
  3. Focus: SEO
  4. Cost: 50 × 2 × 3 = 300 credits
  5. Review generated suggestions
  6. Edit any that need adjustment
  7. Publish all
  8. Monitor performance over next 30 days

Example 2: Refresh product descriptions

  1. Create batch: All products
  2. Field: Description only
  3. Focus: Conversion
  4. Review in batches of 10
  5. Edit to ensure accuracy
  6. Publish reviewed batches incrementally
  7. Track conversion rate changes

Example 3: SEO meta data improvement

  1. Create batch: Products with high traffic but low conversions
  2. Fields: Meta title + meta description
  3. Focus: Conversion (they're already getting traffic)
  4. Review high-confidence suggestions quickly
  5. Spend time on low-confidence suggestions
  6. Publish and monitor click-through rates in Search Console

What's next

Need help with content suggestions? Use the chat widget in the bottom-right corner or email support@convertmate.io.

Was this article helpful?

On This Page