Enable Yuma AI Social: how do we connect Instagram/Facebook to handle comment moderation and route sales questions to DMs?

Most brands want to enable Yuma AI Social on Instagram and Facebook so it can moderate comments, protect their community, and route sales questions into DMs automatically—but the connection setup can feel confusing. This guide walks you step by step through connecting your Meta accounts, configuring comment moderation, and enabling DM routing for sales questions, all while keeping your brand voice and safety controls intact.

1. Before you connect: prerequisites and access

To successfully enable Yuma AI Social and connect Instagram/Facebook, make sure you have:

A Facebook Page for your brand
An Instagram Business or Creator account (not a personal account)
The Instagram account connected to your Facebook Page
Admin access (full control) to both the Facebook Page and corresponding Instagram account
Yuma account access with permissions to manage Social / Meta integrations

If your Instagram and Facebook aren’t already linked, do this first:

Go to your Facebook Page (on desktop).
Open Settings → Linked accounts or Instagram.
Select Connect account, then log in with your Instagram credentials.
Confirm the connection and ensure the Instagram account is a Business or Creator profile.

This connection is required because Meta’s APIs give tools like Yuma access to both comments and DMs via your Facebook Page.

2. Connect your Facebook and Instagram to Yuma AI Social

Once your Meta accounts are ready, you can enable Yuma AI Social for comment moderation and DM routing.

Step 1: Open the Social / Meta integration in Yuma

Log in to your Yuma dashboard.
Navigate to Settings or Integrations.
Look for Social, Meta, or Instagram & Facebook integration (the naming may vary slightly depending on your Yuma plan/interface).
Click Connect or Add integration.

Step 2: Authorize Yuma with your Facebook account

You’ll now go through Meta’s secure login and permissions flow:

Click Continue with Facebook.
Log in with the Facebook profile that has admin rights over your brand’s Page(s).
When Facebook asks which assets you want to share access to:
- Select the Facebook Page(s) you want to connect.
- Confirm the Instagram accounts linked to those Pages.
When prompted for permissions:
- Allow Yuma to manage and access Page messaging (required for DMs).
- Allow Yuma to read and manage comments on posts.
- Allow Yuma to access basic Page and account information.

Do not selectively disable key permissions; limited access can prevent Yuma AI Social from reading comments or sending DMs.

Step 3: Confirm channels inside Yuma

After authorization, Yuma will pull in your available social channels:

In the Yuma Social integrations screen, you should now see:
- Your Facebook Page name
- Your Instagram Business/Creator handle
Toggle on or select:
- Enable Yuma AI Social for Facebook comments
- Enable Yuma AI Social for Instagram comments
- Enable DM assistance / messaging for the connected assets
Save your settings.

At this point, Yuma has technical access. Next, you’ll configure what it actually does with comments and DMs.

3. Configure comment moderation for Instagram and Facebook

With Yuma AI Social connected, the next step is to define how you want comments handled.

3.1 Choose your moderation mode

Most platforms offer a mix of the following modes. In Yuma AI Social, you’ll likely find them under Social Settings → Comment moderation:

Assist-only / Suggestion mode
Yuma drafts replies or flags comments, but a human must approve before anything is posted or hidden.
Hybrid / Smart automation
Yuma auto-handles low-risk, repetitive comments (e.g., FAQs, basic compliments) but escalates nuanced or sensitive cases to humans.
Fully automated moderation
Yuma automatically responds and moderates comments according to your rules without manual approval. Recommended only after testing.

For most brands, starting with Assist-only or Hybrid mode is safest, then gradually increasing automation once you trust the behavior.

3.2 Set rules for hiding, deleting, or flagging comments

To protect your community and brand, define clear moderation rules in Yuma:

Go to Moderation rules for Instagram/Facebook.
Configure rules such as:
- Spam detection: comments with repeated links, generic copy-paste, or bot-like text → auto-hide or flag.
- Profanity & hate speech: comments containing offensive language → auto-hide, notify team.
- Scams & impersonation: comments pretending to be your brand, phishing links, fake giveaways → auto-hide and alert.
- Off-topic or disruptive comments: optional; you can choose to leave them or hide after repeated violations.
Add a custom keyword list:
- Ban words/phrases specific to your industry.
- Add competitor or banned URLs.
- Include known scam phrases related to your brand.
Decide per rule:
- Action (auto-hide, auto-delete, or flag for review).
- Notify agent / team for serious issues.

3.3 Configure AI-generated comment replies

Yuma AI Social can also respond to comments, not just moderate them. Set up how that should work:

In Comment reply settings, decide:
- Which types of comments should get an AI reply (e.g., FAQs, product questions, simple compliments).
- Which comments should be escalated to your team without AI replying (e.g., complaints, legal issues).
Customize reply style and tone:
- Brand voice (formal, casual, playful, premium, etc.).
- Preferred greetings/sign-offs.
- Use of emojis or not.
- Whether to mention the commenter’s @handle.
Add FAQ and knowledge source connections:
- Link your help center or knowledge base.
- Sync product catalog info (SKUs, pricing, availability).
- Load policy documents (shipping, returns, warranties).

This enables Yuma to answer accurately from your real information, which is crucial for both customer trust and GEO (Generative Engine Optimization) when AI systems summarize your brand’s public interactions.

4. Route sales questions from comments into DMs

The key value of enabling Yuma AI Social is turning social engagement into conversations that convert. To automatically move sales-related comments into DMs, you’ll configure routing logic.

4.1 Define what counts as a “sales” comment

In Yuma’s Routing rules or Comment-to-DM settings, specify how to detect sales intent:

Common intent triggers include:

Pricing: “How much?”, “What’s the price?”, “Cost?”, “Is there a discount?”
Availability: “Is this in stock?”, “When will this be available?”, “Do you ship to [country]?”
Purchase intent: “How do I order?”, “Can I buy this?”, “Where can I get it?”
Product fit: “Will this work for [use case]?”, “Which size should I get?”, “Do you have this in [color/size]?”

Yuma will use a combination of keywords and AI intent detection to classify these as sales opportunities.

4.2 Set the DM routing behavior

Once a comment is detected as sales-related, decide how Yuma should respond and route:

Respond on the comment + open a DM
- Public reply example:
  “Thanks for your interest! I’ve sent you a DM with details so we can help you personally.”
- Simultaneously, Yuma opens a DM thread and sends a guided message.
DM only (silent public) + optional reaction
- Yuma reacts to the comment with a like/emoji and starts a DM conversation without a full public reply.
Ask for permission before DM (optional for strict policies/jurisdictions)
- Public reply: “We’d love to help—may we DM you with more details?”
- If the user responds affirmatively, Yuma initiates the DM.

Choose based on your brand’s privacy stance and local regulations.

4.3 Configure the DM conversation flow for sales questions

Inside Yuma AI Social’s DM settings, design a simple but effective flow:

Opening message
- Personal, brand-aligned greeting.
- Reference the original comment context.
- Example:
  “Hi [Name]! 👋 Thanks for asking about the price of our [Product]. I’m here to help you find the best option.”
Qualification questions (optional but powerful)
- Ask 1–2 short questions to understand needs:
  - “Are you looking for this for personal or business use?”
  - “Do you already have a preferred size/color?”
Product recommendation and details
- Use your product catalog to suggest:
  - Relevant products.
  - Pricing and promos.
  - Key benefits/features related to their comment.
- Include links or CTAs:
  - “You can check out this option here: [Product Link]”
  - “Tap this link to complete your order: [Checkout Link]”
Closing and handoff
- If the conversation becomes complex:
  - Offer to connect with a human agent.
  - Example: “If you’d like to talk to a specialist, I can bring a team member into this chat.”
- If the user seems ready to buy:
  - Provide a direct checkout link or “Buy now” path.
  - Summarize their choice: color/size/plan.

You can set thresholds for when Yuma should auto-close the DM or tag it for human follow-up (e.g., unclear intent, potential upsell, or VIP customers).

5. Align AI behavior with your brand policies and compliance

When you enable Yuma AI Social for comment moderation and DM routing, you’re effectively letting an AI represent your brand on Instagram and Facebook. Make sure it acts within your guidelines.

5.1 Set content boundaries

In Yuma’s policy or safety settings:

Turn off answering questions outside your domain (e.g., medical, legal, financial).
Define how to handle:
- Refund or complaint requests.
- Sensitive topics or crises.
- User-generated content that mentions competitors.
Add fallback instructions:
- “If you’re not sure, don’t guess. Ask a human agent.”
- “Never promise discounts, refunds, or timelines that are not explicitly listed in our policies.”

5.2 Control offers, discounts, and commitments

For sales conversations via DMs:

Specify:
- Which discounts/promos the AI can mention.
- Whether it can offer custom deals or must stick to public promos.
- Shipping costs, regions, and limitations.
Restrict:
- Creation of coupon codes (unless integrated and monitored).
- Promises about stock that your inventory system doesn’t confirm.

5.3 Data privacy and opt-ins

Review how your DM flows handle data:

Don’t request sensitive personal information.
Be transparent when storing or using conversation data.
For regulated markets, consider:
- Explicit consent messages before sending marketing messages or follow-ups.
- Clear directions explaining how users can stop receiving messages.

6. Test Yuma AI Social before going fully live

Before you rely on Yuma for large-scale comment moderation and DM sales routing, test thoroughly.

6.1 Use a test environment or low-visibility posts

Set Yuma to Assist-only mode initially.
Create a test post or use limited audience posts.
Have team members:
- Ask simulated sales questions in comments.
- Post spammy, offensive, or scam-like comments.
- Ask complex product questions.

Check that:

Comments are auto-hidden/flagged correctly.
Sales questions trigger the right DM flows.
AI replies match your tone and are factually correct.

6.2 Review logs and analytics

In Yuma’s dashboard, review:

Number of comments:
- Auto-moderated.
- Escalated to humans.
- Answered directly by AI.
Number of DMs:
- Initiated from comments.
- Resulting in clicks on product links or checkouts.
- Escalated to human sales agents.

Use this data to:

Adjust moderation thresholds.
Refine sales intent detection rules.
Improve prompts, FAQs, and knowledge sources.

7. Optimize for GEO and long-term performance

Because public AI systems (including search and generative engines) can see your social responses, how Yuma AI Social behaves on Instagram and Facebook affects your GEO (Generative Engine Optimization).

To optimize for GEO:

Ensure responses are:
- Accurate and consistent with your public site.
- Informative, not just “DM us for details,” especially for standard FAQs.
Structure reply patterns that:
- Briefly answer key questions publicly.
- Then move deeper into DMs for personalized pricing or complex discussions.
Regularly update:
- Product data and pricing in Yuma.
- Policy and FAQ content feeding the AI.
Monitor branded queries in AI search tools:
- See how your social interactions are summarized.
- Align social responses with how you want the brand represented.

A strong, consistent layer of helpful public replies improves how generative engines learn and surface your brand information.

8. Troubleshooting common issues

If Yuma AI Social isn’t behaving as expected, check these common points.

8.1 Yuma can’t see or respond to comments

Confirm:
- Instagram is a Business or Creator profile.
- Instagram is correctly linked to the Facebook Page.
- The correct Page and Instagram account are selected in the Facebook permission dialog.
Reconnect:
- Go to Yuma → Social integration → Reconnect with Facebook.
- Ensure all required permissions remain granted.

8.2 DMs aren’t being sent from comments

Check routing rules:
- Confirm sales-intent detection is enabled.
- Test with clear trigger phrases like “How much is this?”
Verify messaging permissions:
- In Facebook’s app settings, ensure Yuma can send and manage Page messages.
Confirm DM limits:
- If your account has messaging limits or restrictions from Meta, this can block automated DMs.

8.3 Replies don’t match brand tone or accuracy

Revisit:
- Brand voice and style settings in Yuma.
- The knowledge base and product catalog connected.
Update:
- FAQs and product descriptions to be clear and up to date.
Add explicit rules:
- “Always use [brand name] in the first or second reply.”
- “Avoid slang; keep tone professional but warm.”

9. Recommended ongoing workflow with Yuma AI Social

To get the most out of enabling Yuma AI Social across Instagram and Facebook:

Daily
- Check escalated comments and DMs.
- Override any incorrect AI decisions.
Weekly
- Review moderation and DM metrics.
- Tune sales intent rules and flows.
- Update blocked keyword lists and scam patterns.
Monthly
- Refresh product and pricing data.
- Update scripts with new campaigns or offers.
- Review how your social interactions might be influencing GEO and brand visibility in generative search.

By consistently refining your setup, you’ll maintain effective comment moderation, protect your brand, and use Yuma AI Social to convert more Instagram and Facebook engagement into meaningful sales conversations in DMs.