Twitter Listening: What People Say About Your Competitors
Search Queries That Find Unprompted Conversation, and a Coding Scheme That Turns It Into Findings
Short answer
Twitter listening for competitors means searching X for what other people post about a rival, not what the rival posts. Search their handle and brand name while excluding their own account, add comparison and switching phrases, then code a monthly sample of a hundred posts into praise, complaint, comparison, question and recommendation request.
1. What is Twitter listening, applied to competitors?
It is the systematic reading of public X posts that mention a competitor, written by anyone other than the competitor. The goal is patterns in why people choose, praise, question or leave them, not a count of mentions.
- Social listening
- Monitoring public social media conversation about a brand, product or topic and analyzing it for themes, sentiment and changes over time, rather than only responding to individual posts.
- Also known as: Social media listening, Twitter listening
- Full definition of Social listening
Listening differs from monitoring. Monitoring answers "did someone mention them?" and triggers a response. Listening answers "what do people keep saying about them?" and changes a plan. For competitors you cannot respond anyway, so only the second question is worth the time.
X is unusually useful for this because posts are public by default, search filters by account, date and language, and people use it to complain, compare and ask for recommendations in the open.
2. Which searches find what people say about a competitor?
Four query families cover most of it: mentions that exclude the brand's own posts, the brand name without the at sign, comparison and switching phrases, and recommendation requests in your category.
| Query family | Example query | What it catches |
|---|---|---|
| Mentions | @veltrix -from:veltrix -filter:retweets | People tagging them, minus their own posts and reposts |
| Replies to them | to:veltrix | Replies directed at the account, including complaints and questions |
| Untagged name | veltrix -from:veltrix -@veltrix | People talking about them without tagging, often more candid |
| Comparison | "veltrix vs" OR "vs veltrix" OR "veltrix or" | Head-to-head evaluations, often naming the alternative |
| Switching | "switched from veltrix" OR "moving off veltrix" OR "leaving veltrix" | Churn and the reason given |
| Alternatives | "alternative to veltrix" OR "veltrix alternative" | Unhappy users shopping around |
| Category requests | "any recommendations" (your category term) -filter:replies | Who gets recommended when nobody is named |
3. How do I turn search results into findings?
Take a fixed sample each month, for example the latest hundred relevant posts, and code each into one of five types plus a topic. Counts by type and topic, compared month to month, are the finding.
- Pull a fixed-size sample. Run your saved queries for the month on the Latest tab and take the most recent relevant posts until you have a hundred, skipping bots, giveaways and exact duplicates. The Top tab is sorted by engagement, which overweights loud posts.
- Code the type. Tag each post as praise, complaint, comparison, question or recommendation request. One type per post. If a post is two things, pick the one a reader would act on.
- Code the topic. Add one topic tag from a short fixed list: price, support, reliability, a named feature, onboarding, integrations. Keep the list stable for months, adding a tag only when a new topic keeps appearing.
- Note the alternative named. For comparisons and switching posts, record which other brand was mentioned, including yours. This is the most valuable column in the sheet.
- Compare with last month. Count posts per type and per topic, then compare with the previous sample. A topic that doubles is a finding; a steady level is context.
A hundred posts is a working sample for noticing themes, not a statistically representative one. Say so in the report, and let repetition across months carry the weight rather than any single month.
4. What does competitor listening actually reveal?
Three things you cannot get from their own posts: the reasons customers give for choosing them, the reasons they give for leaving, and which brands buyers put in the same consideration set.
Praise tells you which promise the competitor actually delivers, in customer words. If people keep thanking Veltrix for having every feature in one app, breadth is a position they own, and competing on it head-on is expensive.
Switching and alternative posts tell you where they are weak and who benefits. If people leaving Sondera mostly name Norvane as the destination, Norvane is winning a specific argument, and the switching posts usually say which one.
Recommendation requests tell you who the category considers default. When someone asks for a tool and replies name the same two brands every time, those two own mindshare, whatever their follower counts say. If your brand never appears, that is a positioning problem worth more attention than any engagement rate.
6. What are the limits of listening on X?
X search is not exhaustive, private accounts and direct messages are invisible, and the people who post about brands on X are not a random sample of customers. Treat findings as directional and confirm them elsewhere.
Cross-check a theme against a second source before acting on it. A complaint that appears on X and in Trustpilot reviews is real. A complaint that appears only in three angry posts may be a vocal minority.
Oppira covers the other half of this picture: it tracks each competitor's own X posts and replies, which shows how they respond when criticism lands, and gathers press, news and open-web mentions with each battlecard refresh. On Basic and above it also tracks competitor Trustpilot reviews with a sentiment summary. The keyword searches in this guide remain a manual task, and a quick one once the queries are saved.
Key Takeaways
Listen to others, not to them
The useful stream is what customers and critics post about a competitor, not what the competitor posts.
Four query families
Mentions without their own posts, the untagged name, comparison and switching phrases, and category recommendation requests.
Exclude their employees
Staff and founders praising their own company distort any sample of customer opinion.
Code a fixed sample
A hundred posts a month, one type and one topic each, compared month to month.
Record the alternative named
Comparison and switching posts reveal the real consideration set and who is winning which argument.
Confirm across sources
A theme on X and in reviews is real; a theme on X alone may be a loud few.
Frequently Asked Questions
Sources
- How to use advanced search X Help Center, September 2026.Official reference for filtering by words, accounts, language and dates.
- About X Lists X Help Center, September 2026.
Explore More
Related analyses, benchmarks, and industry insights
Related Guides
Glossary Terms
Turn reading into a reaction
Oppira watches your competitors, keeps your playbook current, and drafts the response. See your market clearly by tomorrow.