Social Media skill 11 of 13
Social Listening and Reputation Monitoring SKILL.md
The Social Listening skill makes Claude, Codex or another AI agent track conversations about a brand, competitors and category across the web and turn them into action. It sets goals for care, product insight, competitors, trends, content ideas, advocates and risk, builds precise queries with exclusions, chooses sources, sets alerts and crisis signals, analyses themes with human review, routes insights to owners and reports share of voice and sentiment.
social-listening/SKILL.md, version 1.0.0, 1,508 words.
By Afeez Adebayo, Marketing, SEO & Growth Expert with 10+ years of experience.
Created 10 October 2026. Updated 10 October 2026. Free to use under CC BY 4.0.
Most conversations about a brand happen away from its own profiles: in replies to other people, forums, review sites, Reddit threads, TikTok comments, news articles and private groups that surface publicly. Social listening tracks those conversations to spot complaints early, find content ideas, understand sentiment, learn what customers say about competitors, discover advocates and creators and detect crises before they grow. Monitoring that only counts mentions misses the point; listening means turning what people say into action.
This skill sets up social listening: goals, queries, sources, alerts, sentiment and theme analysis, competitor and category tracking, crisis signals, routing to the right teams and reporting.
Operating principles
- Listen with a purpose. Each query should serve a goal: care, product insight, content, competitors or risk.
- Good queries beat big tools. Precise keywords, exclusions and languages decide data quality.
- Read, don't just count. Sentiment scores are rough; real insight comes from reading samples.
- Route insights to owners. Support, product, sales, PR and content each get what they can act on.
- Respect privacy and platform rules. Use public data and approved tools; don't track individuals inappropriately.
Listening goals and queries
| Goal | What to track | Example query terms |
|---|---|---|
| Customer care | Brand mentions with problems | brand name + "not working", "refund", "cancel" |
| Product insight | Feature requests, complaints, praise | brand + "wish", "would love", "missing" |
| Competitors | Competitor mentions and switching talk | competitor names + "alternative", "switching from" |
| Category trends | Topics and questions in your space | category terms, problem phrases |
| Content ideas | Questions people ask | "how do I" + topic |
| Creators and advocates | People recommending you | brand + "recommend", "love" |
| Risk and crisis | Spikes, negative themes, influential critics | brand + "scam", "lawsuit", "boycott", sudden volume changes |
Sources
| Source | Notes |
|---|---|
| X, Instagram, TikTok, Facebook, LinkedIn, YouTube | Coverage varies by tool and platform access |
| Reddit and forums | Rich, candid discussions |
| Review sites (G2, Trustpilot, app stores, Google reviews) | Product and service feedback |
| News and blogs | Coverage and reputation |
| Podcasts and video transcripts | Growing source where tools support it |
| AI answer engines | Check how assistants describe your brand (see AEO & GEO skills) |
When to use, and when not to
Use it to catch complaints and crises early, inform content and product decisions, track competitors, measure share of voice and find advocates.
Don't use it to monitor private individuals inappropriately, or as a replacement for direct customer research (the Product Marketing customer-research-interviews skill).
Inputs to ask for, in one message
- Goals for listening.
- Brand names, product names, common misspellings, handles and executives.
- Competitors and category terms.
- Languages and markets.
- Tools available (social listening platforms, native alerts, review monitoring).
- Teams that will act on insights.
Raw SKILL.md file
This is the exact file agents read, including the credit header and licence notice. Save it as social-listening/SKILL.md.
---
name: social-listening
description: "Set up social listening that leads to action: goals for care, product insight, competitors, category trends, content ideas, advocates and risk, precise queries with exclusions, sources from social and Reddit to reviews, news and AI answers, alerts and crisis signals, human theme analysis, routing to owners and reporting on share of voice and sentiment. Use whenever someone asks about social listening, brand monitoring, tracking mentions, online reputation, sentiment analysis, competitor monitoring on social, spotting a social media crisis or finding content ideas from conversations."
license: CC-BY-4.0
metadata:
author: Afeez Adebayo
author_url: https://afeezadebayo.com
author_linkedin: https://www.linkedin.com/in/afeez-adebayo/
author_x: https://x.com/za_icon
source: https://afeezadebayo.com/skills-md/marketing/social-media/social-listening/
category: marketing
subcategory: social-media
version: "1.0.0"
created: "2026-10-10"
updated: "2026-10-10"
---
# Social Listening and Reputation Monitoring
Most conversations about a brand happen away from its own profiles: in replies to other people, forums, review sites, Reddit threads, TikTok comments, news articles and private groups that surface publicly. Social listening tracks those conversations to spot complaints early, find content ideas, understand sentiment, learn what customers say about competitors, discover advocates and creators and detect crises before they grow. Monitoring that only counts mentions misses the point; listening means turning what people say into action.
This skill sets up social listening: goals, queries, sources, alerts, sentiment and theme analysis, competitor and category tracking, crisis signals, routing to the right teams and reporting.
## Operating principles
1. **Listen with a purpose.** Each query should serve a goal: care, product insight, content, competitors or risk.
2. **Good queries beat big tools.** Precise keywords, exclusions and languages decide data quality.
3. **Read, don't just count.** Sentiment scores are rough; real insight comes from reading samples.
4. **Route insights to owners.** Support, product, sales, PR and content each get what they can act on.
5. **Respect privacy and platform rules.** Use public data and approved tools; don't track individuals inappropriately.
## Listening goals and queries
| Goal | What to track | Example query terms |
|---|---|---|
| Customer care | Brand mentions with problems | brand name + "not working", "refund", "cancel" |
| Product insight | Feature requests, complaints, praise | brand + "wish", "would love", "missing" |
| Competitors | Competitor mentions and switching talk | competitor names + "alternative", "switching from" |
| Category trends | Topics and questions in your space | category terms, problem phrases |
| Content ideas | Questions people ask | "how do I" + topic |
| Creators and advocates | People recommending you | brand + "recommend", "love" |
| Risk and crisis | Spikes, negative themes, influential critics | brand + "scam", "lawsuit", "boycott", sudden volume changes |
## Sources
| Source | Notes |
|---|---|
| X, Instagram, TikTok, Facebook, LinkedIn, YouTube | Coverage varies by tool and platform access |
| Reddit and forums | Rich, candid discussions |
| Review sites (G2, Trustpilot, app stores, Google reviews) | Product and service feedback |
| News and blogs | Coverage and reputation |
| Podcasts and video transcripts | Growing source where tools support it |
| AI answer engines | Check how assistants describe your brand (see AEO & GEO skills) |
## When to use, and when not to
Use it to catch complaints and crises early, inform content and product decisions, track competitors, measure share of voice and find advocates.
Don't use it to monitor private individuals inappropriately, or as a replacement for direct customer research (the Product Marketing `customer-research-interviews` skill).
## Inputs to ask for, in one message
1. Goals for listening.
2. Brand names, product names, common misspellings, handles and executives.
3. Competitors and category terms.
4. Languages and markets.
5. Tools available (social listening platforms, native alerts, review monitoring).
6. Teams that will act on insights.
## Workflow
### Step 1. Set goals
One to three primary uses at first.
### Step 2. Build queries
Brand terms, misspellings, product names, competitor and category terms; exclusions to remove noise (for example common words that share the brand name).
### Step 3. Choose sources and tools
Match to where the audience talks; check platform coverage limits.
### Step 4. Set alerts
Volume spikes, negative keywords, influential accounts and crisis terms, routed to owners.
### Step 5. Review regularly
Daily for care and risk; weekly for themes and content ideas; monthly for trends and competitors.
### Step 6. Analyse themes
Read samples, tag themes, track sentiment trends with human checks.
### Step 7. Route and act
Support handles complaints; product gets feature themes; content gets questions; PR gets reputation issues; sales gets switching signals.
### Step 8. Report
Share of voice, sentiment trends, top themes, actions taken and outcomes.
## Crisis signals
- Sudden spike in mentions or negative sentiment.
- Influential accounts or journalists posting critically.
- The same complaint repeated by many people.
- Mentions of safety, legal or ethical issues.
- Hashtags or campaigns targeting the brand.
Agree escalation thresholds and owners in advance.
## Query example
```
("Acme Meals" OR "AcmeMeals" OR "acme meal kit" OR @acmemeals)
AND NOT ("acme corporation" OR "acme anvil")
Languages: English
Markets: UK, Ireland
```
Test queries on a sample of results and refine exclusions until most matches are relevant.
## Weekly listening report
| Section | Content |
|---|---|
| Volume and sentiment trend | Mentions this week versus last, with notable spikes |
| Top themes | Three to five themes with example posts |
| Care issues | Complaints found and resolved |
| Product and content insights | Requests, questions and ideas routed |
| Competitors | Share of voice and notable conversations |
| Risks | Anything to watch or escalate |
## Share of voice
Share of voice compares your mentions with competitors' in the same sources and period. Track it alongside sentiment, because a spike in negative mentions also raises share of voice.
## Finding advocates and creators
Filter mentions for positive recommendations and for people with engaged audiences in your niche. Thank them, invite them into the community or advocacy programmes and consider creator partnerships.
## Tool selection
Compare platform coverage, languages, historical data, alerting, sentiment accuracy, integration with support tools and price. Many teams combine a listening platform with native platform search, review alerts and manual checks of key communities.
## Crisis response steps
1. Confirm the facts internally before responding.
2. Pause scheduled posts that could seem insensitive.
3. Acknowledge publicly with what you know and what you're doing.
4. Route individual cases to support privately.
5. Update regularly until resolved, then review what happened.
Refine queries monthly as products, competitors and slang change.
## Edge cases and judgement calls
| Situation | What to do |
|---|---|
| Brand name is a common word | Use exclusions and combine with product or category terms |
| Small brand with few mentions | Focus on category and competitor listening |
| Several languages | Native speakers review samples; tools may misread sentiment |
| Sarcasm and slang | Human review; don't trust automated sentiment alone |
| Limited budget | Native alerts, Google Alerts, review site notifications and manual Reddit searches |
## Handoffs to related skills
- Community responses: `community-building`
- Reporting: `social-media-reporting`
- Creators and advocates: `creator-employee-advocacy`
- Competitive intelligence: the Product Marketing `competitive-intelligence` skill
- PR and reputation: the Content, Copy and PR `digital-pr` skill
- Brand accuracy in AI answers: the AEO & GEO `brand-accuracy-ai-answers` skill
## Worked example
A meal-kit company only watched its own comments. Listening queries for the brand, misspellings and "meal kit" plus complaint terms revealed a growing Reddit thread about late deliveries in one region, many TikTok videos praising a specific recipe and frequent comparisons with a competitor on price. Support contacted affected customers and logistics fixed a courier issue; content created a series around the popular recipe; product marketing updated pricing comparisons. Negative mentions about delivery fell, and the recipe series became a top performer.
## Output format
```
# Social listening plan: <brand>, <date>
## Goals
## Queries and exclusions
## Sources and tools
## Alerts and escalation
## Review cadence and owners
## Reporting template
```
## Common mistakes
- Counting mentions without reading them.
- Noisy queries full of irrelevant results.
- Trusting automated sentiment blindly.
- Insights that never reach product, support or content.
- No crisis thresholds.
## Questions to ask before you listen
- What decisions will listening inform?
- Which terms, misspellings and competitors matter?
- Who acts on each type of insight?
## Self-check before you answer
- Do queries serve clear goals with exclusions?
- Are alerts and escalation defined?
- Are themes read by humans, not just scored?
- Are insights routed to owners and acted on?
---
Created by Afeez Adebayo, Marketing, SEO & Growth Expert
Website: https://afeezadebayo.com
LinkedIn: https://www.linkedin.com/in/afeez-adebayo/
X: https://x.com/za_icon
Source: https://afeezadebayo.com/skills-md/marketing/social-media/social-listening/
Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). You may copy, share and adapt this skill, including for commercial use, as long as you credit Afeez Adebayo and link to the source URL above. Keep this notice in any copy or adapted version.
Workflow
Step 1. Set goals
One to three primary uses at first.
Step 2. Build queries
Brand terms, misspellings, product names, competitor and category terms; exclusions to remove noise (for example common words that share the brand name).
Step 3. Choose sources and tools
Match to where the audience talks; check platform coverage limits.
Step 4. Set alerts
Volume spikes, negative keywords, influential accounts and crisis terms, routed to owners.
Step 5. Review regularly
Daily for care and risk; weekly for themes and content ideas; monthly for trends and competitors.
Step 6. Analyse themes
Read samples, tag themes, track sentiment trends with human checks.
Step 7. Route and act
Support handles complaints; product gets feature themes; content gets questions; PR gets reputation issues; sales gets switching signals.
Step 8. Report
Share of voice, sentiment trends, top themes, actions taken and outcomes.
Crisis signals
- Sudden spike in mentions or negative sentiment.
- Influential accounts or journalists posting critically.
- The same complaint repeated by many people.
- Mentions of safety, legal or ethical issues.
- Hashtags or campaigns targeting the brand.
Agree escalation thresholds and owners in advance.
Query example
("Acme Meals" OR "AcmeMeals" OR "acme meal kit" OR @acmemeals)
AND NOT ("acme corporation" OR "acme anvil")
Languages: English
Markets: UK, IrelandTest queries on a sample of results and refine exclusions until most matches are relevant.
Weekly listening report
| Section | Content |
|---|---|
| Volume and sentiment trend | Mentions this week versus last, with notable spikes |
| Top themes | Three to five themes with example posts |
| Care issues | Complaints found and resolved |
| Product and content insights | Requests, questions and ideas routed |
| Competitors | Share of voice and notable conversations |
| Risks | Anything to watch or escalate |
Share of voice
Share of voice compares your mentions with competitors' in the same sources and period. Track it alongside sentiment, because a spike in negative mentions also raises share of voice.
Finding advocates and creators
Filter mentions for positive recommendations and for people with engaged audiences in your niche. Thank them, invite them into the community or advocacy programmes and consider creator partnerships.
Tool selection
Compare platform coverage, languages, historical data, alerting, sentiment accuracy, integration with support tools and price. Many teams combine a listening platform with native platform search, review alerts and manual checks of key communities.
Crisis response steps
- Confirm the facts internally before responding.
- Pause scheduled posts that could seem insensitive.
- Acknowledge publicly with what you know and what you're doing.
- Route individual cases to support privately.
- Update regularly until resolved, then review what happened.
Refine queries monthly as products, competitors and slang change.
Edge cases and judgement calls
| Situation | What to do |
|---|---|
| Brand name is a common word | Use exclusions and combine with product or category terms |
| Small brand with few mentions | Focus on category and competitor listening |
| Several languages | Native speakers review samples; tools may misread sentiment |
| Sarcasm and slang | Human review; don't trust automated sentiment alone |
| Limited budget | Native alerts, Google Alerts, review site notifications and manual Reddit searches |
Handoffs to related skills
- Community responses:
community-building - Reporting:
social-media-reporting - Creators and advocates:
creator-employee-advocacy - Competitive intelligence: the Product Marketing
competitive-intelligenceskill - PR and reputation: the Content, Copy and PR
digital-prskill - Brand accuracy in AI answers: the AEO & GEO
brand-accuracy-ai-answersskill
Worked example
A meal-kit company only watched its own comments. Listening queries for the brand, misspellings and "meal kit" plus complaint terms revealed a growing Reddit thread about late deliveries in one region, many TikTok videos praising a specific recipe and frequent comparisons with a competitor on price. Support contacted affected customers and logistics fixed a courier issue; content created a series around the popular recipe; product marketing updated pricing comparisons. Negative mentions about delivery fell, and the recipe series became a top performer.
Output format
# Social listening plan: <brand>, <date>
## Goals
## Queries and exclusions
## Sources and tools
## Alerts and escalation
## Review cadence and owners
## Reporting templateCommon mistakes
- Counting mentions without reading them.
- Noisy queries full of irrelevant results.
- Trusting automated sentiment blindly.
- Insights that never reach product, support or content.
- No crisis thresholds.
Questions to ask before you listen
- What decisions will listening inform?
- Which terms, misspellings and competitors matter?
- Who acts on each type of insight?
Self-check before you answer
- Do queries serve clear goals with exclusions?
- Are alerts and escalation defined?
- Are themes read by humans, not just scored?
- Are insights routed to owners and acted on?
On this page
- When to use this skill, and when not to
- Example prompts
- Example output
- What the agent needs from you
- How this skill hands off to related skills
- How to run Social Listening step by step
- Situations this skill handles
- What makes this skill different
- Practical tips from real projects
- Frequently asked questions
- Changelog
When to use this skill, and when not to
Use it to catch complaints and crises early, inform content and product decisions, track competitors, measure share of voice and find advocates.
Most conversations about a brand happen away from its own profiles: replies to others, forums, Reddit, reviews, TikTok comments and news. Listening that only counts mentions misses the point; the value is in what teams do with what people say.
Don’t use it to monitor private individuals inappropriately or as a replacement for customer interviews (use the Product Marketing Customer Research skill).
Example prompts
- “Set up queries to catch complaints early.”
- “What are people saying about us on Reddit?”
- “Track our share of voice against two competitors.”
- “Define crisis signals and escalation for social.”
- “Turn listening data into content ideas.”
Example output
The skill returns a fixed structure. A shortened example for an imaginary site:
# Listening plan: example meal-kit company, 11 October 2026 Found: Reddit thread on late deliveries in one region; TikToks praising one recipe; price comparisons with a competitor Actions: support contacted customers; logistics fixed courier issue; recipe series launched; pricing comparisons updated Result: fewer delivery complaints; recipe series a top performer
Themes are read by people, not just scored, because automated sentiment often misreads sarcasm, slang and context.
What the agent needs from you
The skill asks for listening goals, brand and product names with misspellings and handles, competitors and category terms, languages and markets, available tools and the teams that will act on insights.
How this skill hands off to related skills
Listening connects to the set: Community, Reporting and Creators and Advocates, plus Competitive Intelligence, Digital PR and the AEO and GEO Brand Accuracy in AI Answers skill.
How to run Social Listening step by step
- Install the skill as
social-listening/SKILL.md. - Set listening goals.
- Build queries with exclusions.
- Choose sources and set alerts.
- Review and analyse themes.
- Route insights and report.
Situations this skill handles
Listening setups vary. The skill handles these situations:
| Situation | What the skill does |
|---|---|
| Brand name is a common word | Exclusions and combined terms |
| Few mentions | Category and competitor listening |
| Several languages | Native speakers review samples |
| Sarcasm and slang | Human review |
| Limited budget | Native and free alerts |
| Sudden negative spike | Crisis steps and escalation |
What makes this skill different
Most listening programmes produce mention counts. This skill starts with goals, writes precise queries, reads the conversations, routes each insight to the team that can act and defines crisis thresholds in advance.
It also checks how AI answer engines describe the brand, a growing part of reputation.
Practical tips from real projects
- Listen with a purpose.
- Good queries beat big tools.
- Read, do not just count.
- Route insights to owners.
A worked example: a meal-kit company only watched its own comments. Listening queries for the brand, misspellings and meal kit plus complaint terms revealed a growing Reddit thread about late deliveries in one region, many TikTok videos praising a specific recipe and frequent price comparisons with a competitor. Support contacted affected customers and logistics fixed a courier issue; content created a series around the popular recipe; product marketing updated pricing comparisons. Negative mentions about delivery fell, and the recipe series became a top performer.
Goals shape queries. Customer care tracks the brand with problem terms such as not working, refund or cancel. Product insight tracks wishes and missing features. Competitor listening tracks rival names with alternative or switching from. Category listening tracks problem phrases and topics. Content listening tracks how-do-I questions. Advocate listening tracks recommendations. Risk listening tracks terms such as scam, lawsuit or boycott and sudden volume changes.
Sources include the main social platforms, Reddit and forums, review sites such as G2, Trustpilot, app stores and Google reviews, news and blogs, podcasts and video transcripts where tools support them and AI answer engines, which increasingly shape how brands are described.
Queries combine brand names, handles, misspellings and product names, exclude noise such as companies sharing a common word and are tested on samples until most results are relevant.
Crisis signals include sudden spikes in mentions or negative sentiment, critical posts from influential accounts or journalists, the same complaint from many people, safety, legal or ethical issues and hashtags targeting the brand, with thresholds and owners agreed in advance.
Crisis response confirms facts internally, pauses scheduled posts that could seem insensitive, acknowledges publicly with what is known and being done, routes individual cases to support privately, updates regularly and reviews afterwards.
Review cadence runs daily for care and risk, weekly for themes and content ideas and monthly for trends and competitors.
The weekly report covers volume and sentiment trends, top themes with examples, care issues resolved, product and content insights routed, competitor share of voice and risks to watch.
Insights go to the teams that can act: support handles complaints, product gets feature themes, content gets questions and ideas, PR gets reputation issues and sales gets switching signals from competitor conversations. Share of voice compares your mentions with competitors in the same sources and period, read alongside sentiment because a spike in negative mentions also raises share of voice. Positive mentions from people with engaged niche audiences point to advocates and potential creator partners, who are thanked and invited into community or partnership programmes. Tool selection compares platform coverage, languages, historical data, alerting, sentiment accuracy, support integrations and price, and many teams combine a listening platform with native search, review alerts and manual checks of key communities. Queries are refined monthly as products, competitors and slang change, and privacy and platform rules are respected throughout by using public data and approved tools only.
Native speakers review samples in each language the brand listens in.
Frequently asked questions
What is social listening?
Tracking public conversations about your brand, competitors and category to act on complaints, insights, opportunities and risks.
How is listening different from monitoring?
Monitoring counts mentions; listening reads them and turns themes into action across teams.
Which sources should be included?
Social platforms, Reddit and forums, review sites, news and blogs, podcasts and video where tools allow, and AI answer engines.
Is automated sentiment reliable?
Only roughly; sarcasm, slang and languages need human review.
What are crisis signals?
Sudden spikes in mentions or negativity, critical posts from influential accounts, repeated complaints and safety or legal issues.
Can small brands do listening without expensive tools?
Yes, with native alerts, Google Alerts, review notifications and manual checks of key communities.
Changelog
- 1.0.0, 10 October 2026: first release.
Use, adapt and share this skill under CC BY 4.0: credit Afeez Adebayo and link to this page.
All Social Media skills
- Social Media Audit
- Social Media Strategy
- LinkedIn Organic Growth
- X (Twitter) Organic Growth
- Instagram Organic Growth
- TikTok Organic Growth
- Short Video Scripts
- YouTube SEO and Channel Growth
- Social Media Content Calendar
- Community Building and Management
- Creator Partnerships and Employee Advocacy
- Social Media Reporting
