If you’ve ever set up a social listening query and then groaned at the junk data it pulled in, you’re not alone. Whether you’re new to Brandwatch or you’ve been wrestling with its Query Builder for a while, this guide is for you. We’ll cut through the noise (and the hype) to show you how to build smarter queries that actually give you what you need—and not a bunch of irrelevant noise.
1. Get Crystal Clear on What You Actually Need
Before you even log in, get specific about your goal. Are you tracking brand mentions? Industry chatter? Hunting for customer complaints? The more precise you are, the less cleanup you’ll do later.
- Write down your goal in one sentence. If you can’t, you’re not ready to start.
- Decide what doesn’t matter. Are you okay missing random memes that happen to mention your brand?
Pro tip: Vague goals (“see what people are saying about us online”) lead to messy queries and hours of filtering later. Don’t skip this step.
2. Map Out Your Keywords—and Their Landmines
Now, jot down every keyword, hashtag, and handle you think you’ll need. But don’t stop there. Think about the ways people don’t mention your brand:
- Common misspellings and abbreviations.
- Product names, campaign hashtags, and slogans.
- Nicknames or slang (if relevant).
- Handles and URLs.
But watch out for false positives. Some words sound unique in a meeting but are everyday language online.
- “Apple” isn’t always about tech.
- “Jaguar” could mean cars or cats (or the NFL team).
What to ignore: Don’t get greedy and add every possible variation. More isn’t always better—sometimes it’s just noisier.
3. Use Boolean Logic Like a Human, Not a Robot
Brandwatch’s Query Builder is built on Boolean logic (AND, OR, NOT). But if you just mash keywords together, you’ll get soup.
- Use “AND” to narrow your focus:
brandname AND (complaint OR refund) - Use “OR” to broaden it:
(brandname OR handle OR hashtag) - Use “NOT” to weed out junk:
brandname NOT "apple pie"
Parentheses matter. If your query logic is messy, your results will be too. Group things logically.
Common Pitfalls
- Over-nesting: If you’re five layers of parentheses deep, you’re making this harder than it needs to be.
- Too many NOTs: You can’t outsmart the internet. Excluding everything leads to gaps you might regret.
4. Filter by Source, Language, and Location
Don’t try to capture the whole internet if you only care about Twitter, or you only want English results. Brandwatch lets you filter by:
- Source: Twitter, Reddit, blogs, news, etc.
- Language: Useful for global brands or regions.
- Location: Country, region, or city (rarely perfect, since location data is often missing).
Honest take: These filters aren’t magic. If someone tweets in French from New York, but their profile says Paris, you might get weird results. But it still beats scrolling through stuff you don’t care about.
5. Exclude the Obvious Junk Early
This is where most people get lazy—and pay for it later.
- Block common spam words: “win a free iPhone,” “discount code,” etc.
- Exclude competitor noise: If you’re tracking “Apple,” you might want to weed out “Applebee’s.”
- Filter out retweets and links if they’re not relevant.
Pro tip: Run a test query, scan the first 100 results, and note the junk. Add those terms to your NOT list.
6. Test, Tweak, and Run Sample Queries
Don’t just hit “save” and walk away.
- Preview your query and look at a random slice of results. Is it mostly good, or mostly noise?
- Check for “leakage”: Are you missing posts you care about? Are you drowning in unrelated chatter?
- Iterate: Adjust keywords, Boolean operators, and exclusions as needed.
What to ignore: Don’t obsess over making your query “perfect.” Aim for “good enough” and revisit it once you’ve got real data.
7. Use Categories and Rules to Clean Up Further
Brandwatch lets you create categories and automated rules to tag or group mentions after they’re collected. This is handy for:
- Sorting by sentiment, topic, or product line.
- Filtering out customer service complaints from general buzz.
- Flagging posts for follow-up.
These tools don’t fix a bad query, but they can make a good one even better.
Honest take: If you’re spending more time building categories than you did on the query, you probably skipped a step above.
8. Set Up Alerts—and Don’t Overdo It
Alerts are useful if you want to know when something spikes (“10x more mentions of our brand today”). But if your query is noisy, your alerts will be, too.
- Keep alerts for actual emergencies. Don’t set up an alert for every blip.
- Check your query quality first. Junk in, junk out.
9. Audit Regularly (But Don’t Obsess)
Social media changes fast—so do slang, memes, and how people talk about your brand.
- Schedule a review every month or quarter.
- Look for new false positives or negatives.
- Add or drop keywords as needed.
Pro tip: Save old versions of your queries. Sometimes the “fix” you made last month wasn’t actually an improvement.
What Actually Works—and What’s Overhyped
Stuff That Works
- Simple, focused queries.
- Logical use of AND/OR/NOT.
- Regular manual reviews.
- Using test queries before full rollout.
Stuff to Ignore
- Trying to capture “everything.” You’ll get buried in data you don’t need.
- Overengineering with fancy logic. Clear beats clever.
- Assuming Brandwatch’s AI will sort it all out. The tool’s smart, but it’s not psychic.
Keep It Simple—and Keep Tuning
Don’t let the power of Brandwatch tempt you into building Rube Goldberg machines. The best queries are the ones you understand, can easily explain, and can adjust when things change. Start small, check your results, and tweak as you go. If you’re spending more time fiddling with your query than actually using the data, it’s time to step back.
Social listening is messy—don’t make it messier. Keep your setup lean, stay curious, and remember: the best insight comes from good questions, not just more data.