Competitor Pricing & Packaging Monitor

Turn public SaaS pricing pages into normalized competitive intelligence—then detect the pricing and packaging changes that matter.

This Actor extracts plans, prices, currencies, billing periods, units, features, calls to action, free-plan and trial signals, annual discounts, seat-based pricing, and usage-based pricing. It stores a normalized per-URL baseline and compares every later successful run against that baseline. The result is a typed, evidence-backed change report instead of a noisy text diff.

Why use it?

Generic website monitors tell you that a page changed. Raw pricing scrapers return whatever text happened to be visible. Neither reliably answers the questions a pricing, product-marketing, sales, or research team needs:

• Which plan was added or removed?
• Did the price change, and for which billing period or unit?
• Which features moved into or out of a package?
• Was a free plan, trial, enterprise tier, annual discount, seat model, or usage model introduced?
• What public evidence supports the finding?

Competitor Pricing & Packaging Monitor answers those questions in a stable schema that works in datasets, dashboards, webhooks, MCP clients, and scheduled workflows.

Quick start

Run with no input for a safe built-in three-plan example. To monitor a real page:

{
  "pages": [
    {
      "url": "https://www.notion.com/pricing",
      "competitor": "Notion",
      "label": "Notion pricing",
      "tags": ["productivity", "benchmark"]
    }
  ],
  "renderMode": "auto",
  "captureScreenshot": false,
  "stateStoreName": "my-saas-competitors"
}

The first successful run creates a baseline. Schedule the same input daily or weekly; future runs compare against the named store and return unchanged or changed.

Better than a generic page diff

The comparison fingerprint is built from normalized plans, price semantics, feature lists, CTAs, and packaging signals. Navigation, footers, cookie banners, modals, chat widgets, scripts, and other common noise are removed before extraction. Copyright dates or rearranged whitespace do not create a pricing alert.

When a real change occurs, change.details contains typed evidence:

{
  "status": "changed",
  "competitor": "Northstar Analytics",
  "change": {
    "material": true,
    "severity": "high",
    "summary": "1 price change; 1 feature added.",
    "details": [
      {
        "type": "price-changed",
        "plan": "Growth",
        "field": "prices",
        "before": "$49 per user / month",
        "after": "$69 per user / month",
        "evidence": "Growth For teams operating repeatable growth programs. $69 per user / month ..."
      }
    ]
  }
}

What gets extracted

Each successful page record includes:

• named plans and descriptions
• numeric or custom prices
• normalized currency, billing period, price unit, and qualifier
• plan features and primary call to action
• free plan, enterprise plan, free trial, and annual-discount signals
• seat-based and usage-based pricing signals
• extraction confidence and warnings
• stable pricing fingerprint
• baseline, unchanged, or changed status
• typed before/after change details with evidence
• optional full-page screenshot key and normalized page evidence

Fast HTTP checks with browser fallback

renderMode: "auto" uses an HTTP-first strategy to keep routine monitoring fast and inexpensive. It switches to browser rendering when extraction confidence is too low or screenshot evidence is requested.

• auto — HTTP first; use Chrome for low-confidence or screenshot requests.
• http — fastest and cheapest; best for server-rendered pages.
• browser — start with Chrome; best for JavaScript-heavy pricing widgets.

If browser rendering fails, the Actor can retain a useful HTTP result. If no useful pricing structure reaches minConfidence, it returns an uncharged failed record rather than a misleading paid result.

Datasets and upstream integrations

Select an Apify dataset containing any of these URL fields:

• url
• pricingUrl
• link

Optional competitor, label, tags, contentSelector, and ignoreSelectors fields are accepted. Inline pages and dataset pages can be combined; duplicate URLs are checked once.

This makes the Actor suitable after a company-discovery scraper, CRM export, research dataset, or portfolio inventory.

Scheduling and state

Set a task or schedule with a stable stateStoreName. The Actor stores one compact normalized snapshot per URL in that named key-value store. A separate name creates an independent baseline set.

Use resetBaseline: true after an intentional scope or parser change. The current result becomes the new baseline and is not reported as a competitor move.

Evidence controls

• includeEvidenceText: true includes up to 20,000 characters of cleaned public-page text.
• captureScreenshot: true saves a full-page PNG in the run's default key-value store and returns its record key.
• Every normalized plan contains the public text used as its extraction evidence.
• Every material change includes before/after values and supporting evidence.

Screenshots require browser rendering and may increase run duration and platform compute.

Pricing

You pay only for successful pricing-page-analyzed events. Failed, blocked, empty, or low-confidence pages are returned without that value charge.

Apify tier	Price per analyzed page
Free	$0.015
Bronze	$0.014
Silver	$0.012
Gold	$0.010
Platinum and Diamond	$0.009

There is also a $0.002 Actor-start event that covers fixed browser-image startup and safety-validation cost. A 20-page scan costs $0.302 on the Free tier before platform compute or proxy charges, with normalized extraction and change detection included.

Privacy, rights, and network safety

• Limited permissions: read the selected dataset and public HTTP/HTTPS pages; write the output dataset, evidence store, and named baseline store.
• Localhost, private, loopback, link-local, and reserved network addresses are blocked, including redirects and browser subrequests.
• HTML is limited to 3 MB and screenshots to 8 MB per page.
• The Actor does not log into sites, bypass access controls, or accept cookies or credentials.
• It does not send page content to an external AI model.
• Use it only for public pages you are permitted to monitor, and respect applicable site terms, laws, and rate limits.

Important limitations

Pricing pages can personalize content by geography, account, experiment, or cookies. Some prices appear only after clicking toggles, configuring a calculator, or signing in. Extraction can therefore miss interactive, personalized, geographic, or login-only pricing.

Confidence is a structural extraction signal—not a guarantee of factual completeness. Verify evidence before making consequential pricing, legal, financial, or sales decisions.

Automate through the API

curl -X POST \
  "https://api.apify.com/v2/acts/gifted_wagon~competitor-pricing-monitor/runs" \
  -H "Authorization: Bearer $APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"pages":[{"url":"https://www.notion.com/pricing","competitor":"Notion"}],"stateStoreName":"my-saas-competitors"}'

The run response contains the default dataset and key-value-store IDs. You can also use Apify schedules, tasks, webhooks, API clients, Make, Zapier, n8n, or the hosted Apify MCP server.

Support

If a public pricing page is not extracted correctly, open an issue from the Actor's Issues tab. Include the run ID, URL, rendering mode, confidence, and a redacted result. Do not post credentials, private URLs, or confidential pricing in a public issue.