Website Opportunity Scorer - Local Business Leads

Find the local businesses whose website needs work, and know what to pitch before you call.

Feed this Actor the results of a Google Maps scraper run (or a plain list of websites, business objects or CSV). For every business it checks the website with fast plain HTTP requests and returns a 0-100 opportunity score (higher = the site needs more work = better lead), a grade (A-F), a hot / warm / cold priority, the top 3 reasons in plain English ("Not mobile-friendly", "No HTTPS", "Copyright notice says 2017") and a suggested pitch angle with a one-line pitch that uses the business's own name, category, city and reviews.

Built for web design agencies, SEO agencies and freelancers who sell websites, redesigns, local SEO and maintenance to small businesses: kappers, loodgieters, restaurants, dentists, garages, shops.

• No website or only a Facebook/Instagram/Treatwell page? Scored 90-100 as a "new website" lead (cheaper event).
• Site down, 404, parked domain or "coming soon"? Scored 85-95 as a "website rescue" lead.
• Outdated, not mobile-friendly, no HTTPS? Redesign lead with the reasons to mention.
• Modern site with missing SEO basics, no click-to-call, no analytics? Local SEO / conversion / tracking pitch.
• Healthy site? Grade A, "maintenance / SEO retainer" angle, so you don't waste a cold call on a redesign pitch.

Also included: an HTML report sorted by score and a SUMMARY with the most common problems, CMS share and score distribution of the whole list.

How it works

1. Run any Google Maps scraper (for example compass/crawler-google-places) for "kapper Utrecht", "loodgieter Amersfoort", ...
2. Start this Actor and pick that run's dataset in the dataset picker (or paste the dataset ID).
3. Get one row per business, sorted by opportunity in the HTML report. Export to CSV/Excel or your CRM.

This Actor never scrapes Google Maps itself; it only reads the dataset you give it (read-only) and visits each business's own website homepage.

What is checked

All checks use the homepage HTML from plain HTTP requests (https and http), a TLS handshake and a technology fingerprint database. No browser, no proxy, no JavaScript rendering.

Area	Checks
Security	HTTPS, certificate valid / expired / wrong host / incomplete chain, expiring within 14 days, http → https redirect, mixed content
Mobile	Responsive viewport meta tag
Outdated tech	End-of-life CMS (Joomla < 4, Drupal < 9, TYPO3 < 11, Magento 1, PrestaShop < 1.7), WordPress < 6.0, jQuery < 3.5, Flash, table layouts and <font>/<center> tags, copyright year 3+ / 5+ years old, DIY site builders (Wix, Jimdo, ...)
SEO basics	Title (missing / length), meta description, H1, noindex, LocalBusiness structured data (schema.org and subtypes such as HairSalon, Plumber, Restaurant), Open Graph tags
Conversion	Contact / booking button or form, click-to-call (tel:) link, WhatsApp link, booking tool (Salonized, Treatwell, Calendly, ...)
Tracking	Google Analytics, Tag Manager, Matomo, Plausible, Clarity, Hotjar, Facebook Pixel, ...
Speed	Homepage response time and HTML weight; optional Core Web Vitals from real Chrome users (CrUX)
Broken / not a site	DNS failure, timeout, connection refused, HTTP 4xx/5xx, parked or for-sale domain, placeholder ("binnenkort online", "under construction", server default page, empty page), redirects to a social or booking platform
Not scored	Sites behind bot protection (Cloudflare challenge, SiteGround captcha, 403) or whose hosting firewall drops our HTTPS connection while http:// answers are output as blocked with no score and are not charged

Scoring rules

Every point comes from a documented rule (no AI). The score is the sum of all matching rules, capped at 100.

Whole-site outcomes (the score is set to this value):

Rule	Points
no_website - no website on the listing	100
dns_fail - domain does not resolve	95
parked - parking / for-sale page	92
platform_only - listing links to Facebook, Instagram, Treatwell, Linktree, ...	90
unreachable - timeout, connection refused, TLS failure	90
http_error - homepage returns 4xx/5xx	85
placeholder - coming soon / under construction / default server page / empty	85

Website rules (added up):

Rule	Points
no_https - no HTTPS	25
cert_invalid - certificate expired / wrong host / self-signed	25
no_viewport - not mobile-friendly	22
noindex - homepage hidden from Google	15
flash - uses Flash	15
outdated_cms - end-of-life CMS version	10
table_layout - table layout / deprecated HTML tags	10
copyright_old - copyright 5+ years old (3-4 years: 6)	10
old_wordpress - WordPress older than 6.0	8
no_title - no page title	8
slow - server response (time to first byte incl. redirects) > 3.5 s (2-3.5 s: 4)	8
cwv_fail - fails Core Web Vitals (add-on)	8
cert_chain - incomplete certificate chain	6
no_https_redirect - http:// does not redirect to https://	6
no_meta_description - no meta description	6
no_cta - no contact / booking button, form or booking tool	6
no_h1 - no H1 heading	5
no_local_schema - no LocalBusiness structured data	5
no_phone_link - phone number not click-to-call	5
old_jquery - jQuery older than 3.5	5
cert_expiring - certificate expires within 14 days	5
no_analytics - no analytics or tag manager	4
diy_builder - DIY site builder	4
mixed_content - http:// resources on an https page	4
heavy_html - homepage HTML > 1 MB	4
cwv_poor_lcp - LCP > 4 s for real users (add-on)	4
no_open_graph - no Open Graph tags	3
title_length - title shorter than 30 or longer than 60 characters	2

Grades: A < 15, B 15-29, C 30-49, D 50-69, F 70+. Priority: hot 50+, warm 25-49, cold < 25.

Pitch angle: follows the opportunity type (redesign leads get the strongest of security / mobile / modern redesign, SEO leads local SEO, conversion leads calls & bookings or tracking, healthy sites a maintenance / SEO retainer); otherwise the angle whose rules scored the most points (ties: the more urgent angle). Angles: new website, website rescue / rebuild, security & trust (HTTPS), mobile-friendly redesign, modern redesign, local SEO, more calls & bookings, speed optimisation, analytics & tracking, maintenance / SEO retainer.

Input

Field	Description
datasetId	Dataset of a Google Maps scraper run (dataset picker, ID, or Console/API URL). compass/crawler-google-places and Outscraper-style fields are recognised.
businesses	Websites as strings, or objects { "name", "category", "city", "website", "phone", "rating", "reviewCount", ... }. Maps scraper items can be pasted as-is.
urls	Plain list of websites.
csvText	CSV with a header row (English or Dutch column names: naam, website, plaats, telefoon, ...).
minScore	Only output and charge businesses with at least this score (default 0).
onlyCategories	Keep only categories containing these words, e.g. kapper.
includeNoWebsite	Output no-website / social-only businesses (default true).
skipClosed	Skip permanently closed businesses (default true).
maxBusinesses	Limit (0 = all).
checkCoreWebVitals	Core Web Vitals add-on (default off), see below.
googleApiKey	Optional own Google API key for the add-on (encrypted secret field).
maxConcurrency, requestTimeoutSecs, userAgent	Advanced.

All sources can be combined; duplicates (same place ID, or same name and website) are scored once. Leave the input empty for a free preview on a few example Dutch businesses.

{
    "datasetId": "<default dataset ID of your Google Maps scraper run>",
    "minScore": 30,
    "onlyCategories": ["kapper", "barbier"],
    "includeNoWebsite": true
}

Output

One row per business. Business fields first (only these 11 are ever copied from the input), then the score, reasons, pitch and every signal (flat, CSV-friendly). Shortened example:

{
    "name": "Kapsalon Voorbeeld",
    "category": "Kapper",
    "address": "Voorstraat 1, 3512 AJ Utrecht",
    "city": "Utrecht",
    "website": "https://www.kapsalon-voorbeeld.nl/",
    "phone": "030 123 4567",
    "rating": 4.7,
    "reviewCount": 87,
    "openingHours": "dinsdag: 09:00 tot 18:00; ...",
    "placeId": "ChIJ...",
    "mapsUrl": "https://www.google.com/maps/search/?api=1&query=...",
    "opportunityScore": 72,
    "grade": "F",
    "leadPriority": "hot",
    "opportunityType": "redesign",
    "topReasons": [
        "Not mobile-friendly: no mobile viewport tag, so phones show a shrunken desktop page",
        "No HTTPS: browsers mark the site as \"Not secure\"",
        "Copyright notice says 2016: the site looks unmaintained for 10 years"
    ],
    "pitchAngle": "mobile",
    "pitchLine": "Mobile-friendly redesign: most visitors of a kapper come from phones, and the current site is not built for them. Kapsalon Voorbeeld has 87 Google reviews (4.7★), so people are looking for it.",
    "https": false,
    "mobileViewport": false,
    "cms": "WordPress",
    "wordpressVersion": "4.9.8",
    "copyrightYear": 2016,
    "localBusinessSchema": false,
    "phoneLink": false,
    "hasAnalytics": false,
    "responseTimeMs": 1840,
    "scoreBreakdown": [{ "id": "no_viewport", "points": 22, "reason": "..." }],
    "chargedEvents": ["website-scored"]
}

Dataset views: Leads (score, grade, business, top reasons, pitch angle) and Website checks. The key-value store has OUTPUT.html (report sorted by score) and SUMMARY (JSON).

Privacy (GDPR)

The Actor is built to keep personal data out of your lead lists:

• Only these business fields are read from the dataset and output: name, category, address, city, website, business phone, rating, review count, opening hours, place ID, Google Maps URL.
• Emails, owner or person names, review texts and reviewer names are never read or output, even if your scraper dataset contains them (the dataset is read with a field whitelist, so review arrays and enrichment data are not even downloaded). Email addresses that appear inside a business name or on the homepage are replaced by [email removed].
• Pasted objects and CSV columns outside the whitelist (e.g. email, owner) are ignored, with a note in the summary.
• Only the public homepage of each business website is fetched.

You are responsible for how you contact businesses (e.g. B2B outreach rules in your country).

Core Web Vitals add-on (optional)

Turn on checkCoreWebVitals to query Google's Chrome UX Report for each site's origin: LCP, INP and CLS of real Chrome users over 28 days. Sites that fail get extra points and a speed reason. Most small local-business sites have too little traffic to be in CrUX; those are not charged for the add-on. Optionally psiFallback runs a PageSpeed Insights lab test for sites without CrUX data (slow, 10-60 s per site).

The Actor's own key is used by default (shared rate limit). For large runs paste your own Google API key (Chrome UX Report API enabled) in googleApiKey; it is stored encrypted and only used for your run.

Data: Chrome UX Report by Google, CC BY 4.0. This Actor is not affiliated with Google.

Pricing

Pay per event, you only pay for results:

Event	Price	When
Website scored	$0.003 ($3 per 1,000)	a website was fetched and scored (including broken, parked and placeholder sites)
No-website lead	$0.0005 ($0.50 per 1,000)	a business without an own website (none, or only a social / booking page)
Core Web Vitals check	$0.003	add-on on, and CrUX (or the lab fallback) returned data for the site

Free: blocked sites (bot protection), businesses below your minScore, duplicates, closed businesses, and the empty-input preview. Set a maximum charge per run in the run options: the Actor checks the budget before every site and stops cleanly.

Limitations

• Only the homepage is checked (no crawl): a contact page or schema on a sub-page is not seen.
• No JavaScript rendering: sites built fully client-side may look emptier than they are (they are usually flagged as an empty page / placeholder only when the HTML has no text, no links and no scripts).
• Speed is the server response time (time to first byte, including redirects) measured from the Actor's datacenter (US), not from a phone; it is a rough signal. Use the Core Web Vitals add-on for real-user speed.
• When a hosting firewall drops our HTTPS connection but http:// answers, HTTPS support cannot be verified: the site is scored from the http version without "No HTTPS" points (httpsInconclusive: true), or output as blocked when nothing loads.
• Some hosting firewalls block datacenter traffic: those sites are output as blocked, unscored and free.
• Copyright-year and placeholder detection are pattern-based (Dutch, English and German phrases).
• The score says how much the website needs work, not how good the business is.

FAQ

Does it work outside the Netherlands? Yes. The checks are language-independent; Dutch/English/German phrases are recognised for placeholders, parking pages and CTAs.

Which Google Maps scrapers are supported? Any dataset with a website field. compass/crawler-google-places (title, categoryName, website, totalScore, reviewsCount, url, placeId, openingHours) and Outscraper-style (name, site, full_address, rating, reviews, place_id, location_link) fields are mapped automatically.

Can I re-score the same list later? Yes, run again with the same dataset; each run is independent.

Can I get emails? No, on purpose. Use the website and business phone from the output for outreach.