🛠️You've inherited someone else's Meltwater account. Here's how to make it yours.
Your account already has searches built — keywords, filters, saved queries you didn't set up. That's actually a head start. This guide walks you through how to audit what's there, improve what needs fixing, and connect it to work that matters.
Â
→ By the end of this guide, you'll know how to evaluate an existing search, use the AI Search Assistant to improve it, and connect it to dashboards, alerts, or reports — without starting from scratch.
Â
Step 1: Pick one search and open it
Orient yourself before you change anything
Â
Your Action:
- Go to ExploreÂ
- Click on one of the Saved Searches. Don't try to audit everything at once.

- Once opened, look at three things to get a picture of what was built:
- What keywords are being used in the search string
- What filters are applied (Source Type, Location, Language)
- How much volume (Total Mentions) is coming in

Â
Step 2: Read the results with a critical eye
Decide what's working and what isn't. This is your Quality Check.
Â
Scroll through the results and ask:
- Are these mentions actually about the right topic?
- Am I seeing a lot of unrelated content (noise)?
- Does anything obvious seem to be missing — big coverage, seasonal terms, related phrases?
Common patterns: A search titled "Home Depot – Patio Furniture" might surface generic furniture content that has nothing to do with Home Depot. Or it might miss phrases like "outdoor seating" or "backyard sets." Both are fixable.
Â
By the end of this step, you should have a rough sense of whether the search is broadly accurate, noisy, or missing coverage. That's all you need before moving on.
Â
Step 3: Use AI Search Assistant to Improve the search
Fix problems faster than manual editing
Â
Now that you know what's wrong, open the AI Search Assistant. Instead of manually rewriting the query, let it do the heavy lifting..
Â
Try prompts like:
“Make this search more specific to Home Depot patio furniture”
“What keywords am I missing for this topic?”
“Expand this search to include related product terms”
Â

Â
The AI Search Assistant will suggest keyword variations, identify noise sources, and flag gaps in coverage. You don't have to accept every suggestion — review them and take what makes sense.
If you're not sure where the query needs help, just describe the problem: "I'm seeing a lot of results that aren't about Home Depot specifically — how do I fix that?"
Â
Step 4:Â Validate, adjust, and save
Â
Before saving, use the preview panel to check whether the updated query is actually returning better results. Run the same mental check as Step 2:
- Is the noise reduced?
- Is the relevant coverage there?
Â
Make small adjustments if needed, then Save the search.

Â
DEFINITION OF DONE
The results reflect your actual topic. You're not spending 30 seconds per result asking "why is this here?"
Â
Now Put That Search to Work
Â
Once you hit save, the Explore Search page will open again. From here you will be able to take action and create outputs your team will use:

- Email Alerts (Bell Icon)Â - Create an email alert for this Search to get notified when something spikes or goes viral
- Daily Digest (Envelope Icon) - Create a Daily Digest Email so you can send scheduled summaries to stakeholders
- Reporting (Dashboard Icon) - Build a dashboard to monitor coverage over time -Â Build a dashboard to monitor coverage over time
- or if you have Mira Studio you can track trends over time.
Â
🚀 Want to Keep Improving?
Â
If you want to go further, here are a few next steps:
- Course: Start refining your searches with the Advanced Learning Path Course in Meltwater Academy
- Workshop: Watch the Advanced Trend Analysis workshop to uncover patterns in your data
- Playbook: Browse the Use Case Playbooks to find workflows that fits your goals
- Post: Learn more with the Using AI Search Assistant article
Â
Â

