This article focuses on the Inbox features available in the Essential and Reputation plans. The Engage+ plan includes additional functionalities and enhanced tagging options.
🧠 How Our AI Determines if a Message Needs a Response
(a closer look at intent recognition)
At the heart of our response engine is intent recognition, powered by advanced Natural Language Processing (NLP). The AI doesn’t just scan for keywords—it analyzes the structure, tone, and meaning of each message to understand whether the person expects a reply from the brand.
What is “intent to receive a response”?
Our AI uses natural language processing (NLP) to understand the user's intent to receive a response, or in other words, it finds out if the user is looking for a reply from your brand. The AI is trained to detect various types of conversational intent, including but not limited to:
Information-seeking
“How much is this?” “Is this vegan?”
These are direct questions where the user expects clarification or details.Problem-solving intent, complaints or issues that warrant support
“My order hasn’t arrived.” “This doesn’t work.” "DM me"
These imply dissatisfaction or confusion, often needing support.Decision-making cues
“Trying to decide between this and that.” “I’m not sure if I should try this.”
These suggest the user is considering a purchase and might benefit from guidance.Implicit requests
“This looks amazing 👀” (on a product post) or “Wish I knew more.”
Sometimes the intent is subtle. The AI picks up on these cues and surfaces them as high-value engagement moments, even if no question mark is present.Conversational follow-ups
Comments like “still waiting…” or “any update?” often signal that a previous interaction needs closure or that the user feels ignored.
Once detected, messages with intent are tagged as Response Needed, so your team (or automated workflows) can jump in at the right moment.
🧡 What about positive messages that don’t ask anything?
That’s where Fan Community comments come in. These are not detected by the intent engine, but are tagged as "Response Needed" as well because they’re valuable opportunities for brand engagement.
Examples:
“Love this brand so much 💕”
“Just bought my fifth bottle!”
“This is why I stick with you guys!”
They may not always expect a reply from the brand—but replying makes a huge impact. That’s why we include them in the Response Needed queue: not for obligation, but for opportunity.
Actions You Can Take in the Smart Inbox
Organize your workflow with simple, powerful actions:
Reply: Respond directly to the comment.
Like/React: Acknowledge the comment quickly.
Star: Flag comments for follow-up in the Starred tab.
Archive: Remove non-priority comments from your active view.
Go to Comment: Open the comment directly in the native platform.
What Conversations Should You Prioritize?
Use the Smart Inbox and strategic filters to uncover the highest-value interactions. Here’s what to focus on:
1. Customer Complaints
Filter:
AI Response Needed+Negative SentimentGoal: Identify users who are dissatisfied and expect action or resolution.
Common Issues: Shipping delays, product problems, negative experiences.
2. Buying & Pricing Questions
Filter:
AI Response Needed+ Keyword FilteringSuggested Keywords:
pgsqlCopyEditbuy, purchase, price, cost, acquire, pay, offer, shop, value, expense, get, order, product, I want, how much, sale, discount
Goal: Capture users showing purchase intent to boost conversions.
3. Order or Shipment Questions
Filter:
AI Response Needed+ Keyword FilteringSuggested Keywords:
rustCopyEditorder, ship, tracking, delivery, deliver, paid, bought, parcel, post, postage, box, return, refund, didn't arrive, still waiting, FedEx, UPS, shipping, mail, address, transaction
Goal: Support users with fulfillment questions to maintain trust.
4. Fan Community Engagement
Filter:
AI Response Needed+Positive SentimentGoal: Deepen loyalty by engaging with fans posting positive feedback.
Examples: Praises, shoutouts, repeat purchase mentions.