If you’re responsible for brand reputation or social media at your company, you know things can go sideways—fast. One tweet, one Reddit thread, and suddenly you’re in the middle of a PR mess. This guide is for folks who need to catch problems early, not just scroll dashboards all day. Here’s how to set up real, useful alerts in Brandwatch so you can stay ahead of the next crisis.
Why Alerts Matter (and When They Don’t)
If you’re reading this, you probably already get the value of a good alert—you want to know when things are heating up before it’s trending on Twitter. But not every spike is a crisis, and not every alert is worth your time. The trick is to set up alerts that actually matter, without flooding your inbox or Slack with noise.
Here’s what works: - Alerts that catch sudden spikes or weird patterns. - Alerts for specific keywords tied to real risks (like “recall” or “lawsuit”). - Alerts that go to the right people, not just to you.
What to ignore: - Generic “volume increase” alerts with no context. - Alerts for every mention of your brand (unless you’re a one-person shop). - Overly complicated workflows—if it’s too complex, no one will use it.
Step 1: Know What a Crisis Looks Like for You
Before you set up anything, get clear on what a crisis actually means for your brand. This isn’t a philosophical exercise—it’s about avoiding false alarms.
Questions to ask: - What topics or keywords would keep your boss up at night? - How much of a spike is “normal” for your brand? (A 50% increase might be a blip for a big brand, but a five-alarm fire for a startup.) - Who needs to know, and how fast?
Pro tip: Pull up old examples of issues your brand has faced. What were the early signs? Use those as a starting point.
Step 2: Set Up the Right Queries in Brandwatch
Alerts are only as good as the searches behind them. In Brandwatch, these are called “Queries.” Think of Queries as your filter—if you set them up poorly, you’ll get garbage out.
How to Build a Good Query
- Be specific. Don’t just search for your brand name. Add in common misspellings, product names, execs, and crisis keywords (e.g., “BrandName recall,” “BrandName lawsuit”).
- Exclude the noise. Filter out routine chatter, like job postings or ads. Use NOT operators to remove irrelevant terms.
- Test your query. Run it for the past week or month—are you getting the right stuff, or just junk?
Example Query:
("BrandName" OR "Brand Name" OR "BrandNmae") AND ("recall" OR "lawsuit" OR "complaint") NOT ("job" OR "career" OR "promo")
Don’t overthink it: Start simple. You can always tweak later.
Step 3: Create Your Alerts
Once your Queries are set, time to build the actual alerts.
Types of Alerts in Brandwatch
- Spike Alerts: Notifies you when mentions jump by a set percentage or number.
- Volume Threshold Alerts: Triggers when mentions cross a specific number in a time frame.
- Custom Keyword Alerts: Focuses on mentions with crisis-specific keywords.
- Sentiment Alerts: Flags sudden changes in negative sentiment (but be cautious—sentiment analysis is hit-or-miss).
How To Set One Up
- Go to the Alerts section in Brandwatch.
- Choose your Query. Pick the one you set up for crisis terms.
- Set your conditions:
- For spikes: “Alert me if mentions increase by 100% in the last hour.”
- For thresholds: “Alert me if there are more than 50 mentions in 30 minutes.”
- For sentiment: “Alert me if negative mentions double in a day.”
- Pick your delivery method: Email, Slack, or in-app. If you need to wake someone up, use something they actually check.
- Decide who gets notified: Don’t just send to everyone. Pick the people who can act.
Pro tip: Start with higher thresholds, then dial down. It’s better to miss one early mention than to train yourself to ignore constant pings.
Step 4: Test, Tune, and Actually Use Your Alerts
You’ve set up your alerts—but the job’s not done. You need to make sure they work in the real world.
- Run simulations: Ask a coworker to post a fake “crisis” tweet or forum post. Did the alert trigger?
- Check for false alarms: Are you getting too many alerts? Not enough? Adjust your thresholds.
- Review after real incidents: Did your alerts help, or did you find out about the issue elsewhere? Be honest.
What to ignore: Don’t waste time on sentiment-only alerts unless you’ve seen them work for you. They’re famously unreliable, especially for sarcasm or slang.
Step 5: Set Up a Response Workflow (Don’t Skip This)
An alert is only useful if you know what to do next. Document a simple process for when an alert hits.
- Who checks the alert first?
- Who decides if it’s a real crisis?
- Who responds, and how?
- Who escalates, and when?
Keep it simple: A Google Doc with phone numbers and steps is fine. Just make sure people know what to do at 2am.
What to Ignore (and What to Watch Out For)
- Don’t try to automate everything. No tool catches every nuance, and false positives are real. Alerts are a heads-up, not a solution.
- Don’t set and forget. Your brand, your audience, and the internet all change. Review your queries and thresholds every quarter.
- Don’t get hung up on “AI” features. Brandwatch’s natural language processing is decent, but it’s not magic.
Pro Tips for Staying Sane
- Less is more: Start with one or two key alerts. Expand only if you’re missing real issues.
- Train your team: Make sure everyone understands what triggers an alert and what to do next.
- Keep the noise down: If you start getting “alert fatigue,” dial your settings back. A missed non-crisis is better than missing the real deal because you’re ignoring your inbox.
Wrapping Up
Setting up crisis alerts in Brandwatch isn’t rocket science, but it’s easy to overcomplicate. Focus on what matters: clear queries, sensible thresholds, and a plan for what to do when the alarm goes off. Start simple, check what’s working, and adjust as you go. The goal isn’t to become a slave to your alerts—it’s to catch fires before they burn down the house.