New guide Knowatoa for Paid Ads Research
Ryan Castillo

Export Competitor Sentiment Data to Your Reporting Tools

• Ryan Castillo

We’ve had requests from agency teams for a way to pull competitor sentiment data into their existing reporting platforms. The data was visible in Knowatoa’s dashboard, but getting it into tools like NinjaCat, Looker Studio, or custom client reports required manual work.

That’s fixed. You can now export competitor sentiment as structured CSV or JSON data through our API, ready to feed into whatever reporting tool you’re using.

What You Get

The competitor sentiment export gives you numerical data on how AI search engines are talking about your site compared to competitors:

  • Sentiment counts — Positive, neutral, and negative mentions for your site and each competitor
  • One row per brand per AI platform — Your site plus competitors, broken down by ChatGPT, Claude, Perplexity, and the other AI services you’re tracking
  • Period comparison — Current metrics alongside previous period numbers with change percentages
  • Stable format — Same columns every time, so automated reports don’t break

How It Works

The data comes through two endpoints:

Site-level: GET /api/v2/sites/competitor_sentiment
Pull sentiment data for a single site. Perfect if you’re building custom dashboards or need to control exactly which sites export when.

Account-level: GET /api/v2/account/competitor_sentiment
Get all your sites at once. Built for agencies managing multiple clients who want a single pull for everything.

Both endpoints support CSV export (add format=csv) or flat JSON (add flat=true) for feeding into BI tools.

Built for Reporting Tools

If you’re using NinjaCat to build client reports, the flat JSON format maps directly into their custom data connectors. The CSV export works with Excel, Google Sheets, or any tool that reads standard CSV files.

Each row represents one brand (your site or a competitor) on one AI platform. You get current numbers, previous period numbers, and the change between them. The schema stays consistent between pulls, which means your automated reports keep working without manual updates.

Example: Connecting to NinjaCat

Setting up the NinjaCat connection takes a few minutes:

  1. Get your API key from Knowatoa’s Account Settings
  2. Add Knowatoa as a REST API data source in NinjaCat
  3. Point it at https://knowatoa.com/api/v2/sites/competitor_sentiment?site_id=YOUR_SITE&api_key=YOUR_KEY&flat=true
  4. Map the flat columns you need: competitor_name, series, positive_count_current, negative_count_current, sentiment_score_current
  5. Set your refresh schedule

Your NinjaCat reports will now pull fresh sentiment data automatically. Column names and date parameters are in the site competitor sentiment endpoint.

Period-Based Reports

By default, the site endpoint compares your latest refresh to the previous one. The account endpoint is always the latest refresh for each site. For a specific range on one site, pass date parameters to GET /api/v2/sites/competitor_sentiment:

?start_date=2026-01-01&end_date=2026-01-31

The API returns metrics for that period alongside an equal-length comparison period. The change percentages show up in your reports without extra calculation work on your end.

What’s Included

Every row includes:

  • Site and account info — Account ID, site ID, site name, external ID
  • Period dates — The date range these metrics cover
  • Brand details — Competitor ID, name, and whether it’s your site (marked is_target)
  • AI platform — Series ID and name (ChatGPT, Claude, etc.)
  • Sentiment metrics. Each one is four columns: _current, _previous, _change_points, and _change_percent. The metrics are mentions, positive_count, neutral_count, negative_count, the matching percentages, and sentiment_score (average from -1 to 1).

If you’re pulling data repeatedly for dashboards, dedupe on: site_id + period_start + period_end + competitor_id + series_id.

Sentiment Thresholds

The API count buckets are:

  • Negative: score ≤ -0.1
  • Neutral: score ≤ 0.1, and above -0.1
  • Positive: score above 0.1

A score of 0.1 is neutral. The dashboard chip and the scheduled wide CSV use five labels (Very Negative through Very Positive). Those labels are separate from these three counts.

Documentation

The full API reference covers parameters, response formats, and integration patterns:

The MCP server also exposes competitor sentiment through the get_competitor_sentiment tool if you’re working with Claude, ChatGPT, or other AI assistants that support MCP.

Who This Helps

This export is built for teams that need to get Knowatoa data into their existing workflows:

  • Agencies building automated client reports in NinjaCat or similar platforms
  • Marketing teams combining AI search sentiment with other metrics in Looker or Tableau
  • SEO teams tracking competitive sentiment changes over time in custom dashboards
  • Developers integrating AI search visibility data into internal tools

If you’re already pulling other Knowatoa data through the API, competitor sentiment works the same way — same authentication, same format options, same period comparison structure.

Competitor sentiment export is available now. If you don’t have API access yet, it’s included in Professional and Enterprise plans. Reach out if you need help setting up your first integration.

← Back to Updates