Skip to content

AI Classification

Formkove automatically reads every form submission and classifies it against tags you define. It also scores spam risk and detects sentiment, with no training data or manual rules needed.

This is a Pro and Agency feature.

  1. You define tags in plain language (e.g., “Customer complaint”, “Sales inquiry”, “Partnership request”)
  2. Formkove reads each submission and assigns applicable tags
  3. Tags are stored with the submission and shown in your dashboard

You don’t need to teach Formkove anything. Just describe what you care about, and it figures out the rest.

In your Formkove dashboard:

  1. Go to Forms → select your form
  2. Click AI Classification or Tags
  3. Add your tags as plain-text descriptions

Example tags:

Tag What it catches
Sales inquiry People asking about pricing, demos, or features
Customer support Issues, bugs, or account problems
Partnership Collabs, APIs, integrations, reselling
Urgent Anything that sounds like a deadline or emergency
Not spam Legitimate but unusual submissions you want to keep

Formkove is flexible. If a submission doesn’t match any tag, it simply gets no tags. No false positives forced.

Each form has its own AI settings, giving you fine-grained control over how submissions are classified.

The purpose tells Formkove what your form is about so classification is more accurate:

  • General contact — Catch-all for standard contact forms
  • Support — Bug reports, account issues, and help requests
  • Sales lead — Pricing inquiries, demos, and purchasing
  • Feedback — Product suggestions and user opinions
  • Job application — Resume and cover letter submissions
  • Custom — Define your own purpose description

Configure how the AI analyzes submissions. Each form can have one primary action:

Action What it does
categorize_and_filter_spam Tags submissions and filters high-spam entries
triage_by_urgency Prioritizes submissions by urgency level
route_to_department Tags by department (sales, support, etc.) for downstream routing
sentiment_only Detects positive, neutral, or negative tone without additional tagging

Tagged submissions can then trigger downstream automations via webhooks or Slack.

Tags are defined per form, so you can have different classification rules for your contact form vs. your support form. Tags can be:

  • Descriptive — e.g., “Urgent support request”
  • Categorical — e.g., “Bug report”, “Feature request”
  • Action-oriented — e.g., “Needs immediate follow-up”

You can combine tags to create rich classification rules. For example, a submission tagged “Sales inquiry” + “Urgent” could trigger a Slack notification and a webhook to your CRM simultaneously.

Every submission gets a spam score between 0 and 100.

  • 0–30: Likely clean
  • 31–70: Suspicious, review before acting
  • 71–100: Likely spam, consider discarding

You can filter your inbox by spam score and set up automatic rules (via webhooks or Zapier) to route or block high-score submissions.

Formkove detects whether a submission is positive, neutral, or negative in tone. This shows up as a label on each submission in your dashboard.

Use cases:

  • Route negative feedback to your support team
  • Surface positive submissions for testimonials
  • Track sentiment trends over time

AI analysis results are processed asynchronously:

  • Dashboard: Once classification finishes, results (tags, spam score, and sentiment) are stored with the submission and displayed in your dashboard.
  • Integrations: AI classification data is not delivered in real-time webhook or Zapier payloads, which only stream the core form submission. High-spam-risk submissions (score above 70) can be configured to be quarantined and filtered out before notifications are sent.

Formkove analyzes all text fields in your form: name, email, message, or any custom fields you add. It doesn’t read file attachments or structured data fields like dates and numbers.


To route submissions by AI tag via automation, see Zapier and Webhooks.