Web Lead Extractor Agent Template

What it does: Turns a plain description of your ideal customer into a clean, deduplicated, scored list of matching businesses, scraped from Google Maps through Apify and saved for outreach or CSV export.
Best for: sales and SDR teams, agencies building client lead lists, growth and demand-gen, local and niche business prospecting
Apps used: Apify (Google Maps scrapers, via Composio)
Setup time: about 10 minutes (connect Apify, run a test search, publish)

Web Lead Extractor is a no-code agent that turns a plain description of your ideal customer into a clean, deduplicated, scored list of matching businesses, scraped from Google Maps through Apify and ready for outreach or CSV export. This template acts as a starting point: any app, instructions, and agents can be updated as required for your own use case, simply ask Max (Agent Builder) to update the workflow as needed. Last verified from workflow config on July 2026. Includes common failure modes and fixes.

Problem this solves

Building a targeted prospect list by hand means searching Google Maps, copy-pasting into a spreadsheet, and cleaning junk rows for hours.
Generic scrapers dump raw, duplicate, unranked data with tracking junk in URLs and invalid emails, so you still do the cleanup.
You want an outreach-ready list scoped to a real customer profile (category, location, rating, reviews), not "as many rows as possible."
You need cost control: scraping is billed per business, so pulling more than you need wastes money.

What this agent does

This template can:

Capture your Ideal Customer Profile (categories, location, intent keywords, rating and review thresholds, required fields, exclusions) and save it so you never re-type it.
Ask how many leads to fetch on every run, so you control spend before any scrape starts.
Map your profile to the correct Apify Google Maps actor input and run the scrape.
Clean every result: validate emails, standardize phone numbers, strip tracking junk from website URLs, keep only real social links.
Deduplicate businesses by their Google place ID so the same place never appears twice.
Score and rank each lead by quality (rating tempered by review volume) and apply your exclusions (competitors, closed places, wrong categories).
Save a ranked, outreach-ready lead list per business and per extraction date.
Enrich contacts (emails and social links) only when your profile asks for it, and only when the business actually publishes them.
Export any saved lead list to a clean CSV, built exactly to your column, filter, and sort spec.
Summarize each run honestly: places found, new vs. already-stored, how many passed your filters, and any gaps in the data.

How it works

You describe your target customer in plain language through the text trigger, for example "coffee shops in Wiesbaden, Germany with 4+ stars and 50+ reviews, I want name, phone, website and email, skip Starbucks."
The agent builds or confirms your profile (ICP). On the first run it collects the missing details and saves the profile to your workspace. On later runs it shows the saved profile and invites changes instead of asking again.
The agent confirms how many leads to fetch. Because you are billed per business scraped, it always asks for a target count (50 to 100 is a good start) and sizes the search to match.
The agent runs the scrape by mapping your profile to the Apify Google Maps actor, starting the run, polling until it finishes, then fetching the results.
The results are cleaned, deduplicated, scored, and filtered against your profile automatically, then saved as a ranked lead list (best matches first).
You get a summary of what was found, what was new, how many passed your filters, and where data is missing. The agent never invents phone numbers, emails, or ratings.
You export to CSV on request by naming the business (or the file) and telling the agent which columns, filters, and sort order you want.

Requirements

A SketricGen workspace with the Web Lead Extractor template cloned.
An Apify account connected to the agent through the Composio app marketplace (this powers all Google Maps scraping).
A small Apify credit. A free Apify account starts with roughly $5, enough for a first run of hundreds of places.
Your customer profile inputs: business categories, one location (city plus country), intent keywords, minimum rating and review thresholds, required fields (phone, website, email, address), exclusions, and a target lead count.

Setup guide

Clone the template and open it in AgentSpace.
Connect Apify. Open the apify tool node, click Connect, and authorize the Apify app from the marketplace (attach it to the Web Lead Extractor agent's tools). The template ships with the connection off until you authorize it.
Confirm the Web Lead Extractor agent node is wired to the Apify tool node on the tools edge.
Open the Playground and run a small test search, for example a niche and city with a target of 20 to 30 leads.
Confirm the agent asks for your profile details and the lead count, runs the scrape, and returns a cleaned, ranked list plus a summary.
Ask for a CSV to confirm export works ("export these as a CSV with name, phone, website, and score").
Publish to your channel when the results look correct.

Common issues and fixes

"Apify not connected" or tool disabled: Authorize the Apify app on the tool node and re-run. A run cannot start until Apify is connected.
Too few or no results: Broaden the location, use different or fewer keywords, or lower the rating threshold. The agent suggests this rather than inventing leads.
Missing emails: Normal. Not every business publishes an email, so coverage is usually partial. Email and social enrichment only runs when your profile asks for it, and adds a small per-place cost.
A run seems stuck or timed out: Google Maps scrapes take a few minutes, and enrichment takes longer. A timeout does not mean the run failed; the agent recovers the in-flight run instead of starting a costly duplicate.
Costs higher than expected: The number of leads you request is the main lever. Keep the target sensible and leave named decision-maker enrichment off on a free Apify tier, since it is charged per person.

Customization knobs

Adjust rating, review-count, required-field, and exclusion rules in your saved profile at any time by telling the agent what to change.
Change the target lead count per run (it is always re-confirmed, never silently reused).
Turn email and social contact enrichment on or off per profile.
Ask Max (Agent Builder) to add a downstream step (for example hand the list to an outreach agent), swap or add a scraper, or change how leads are scored.

Apps used

App	What it is used for	Typical permission scope
Apify (via Composio)	Runs Google Maps scrapers to discover matching businesses and, when your profile requires it, enrich emails and social links from business websites	Connection authorized on your workspace; start actor runs and read run datasets

The template uses three Apify actors under the hood: a Google Maps discovery scraper, an email-first Google Maps scraper (for email-centric profiles), and a website content crawler (for extra detail on a business site). You do not configure these directly; the agent picks the right one for your profile.

Use cases

Local business prospecting: Pull every plumber, dentist, or gym in a city that meets your rating and review bar. Best for local service sellers and SDRs.
Agency client lead lists: Build a fresh, deduplicated list per client niche and city, then export a CSV to hand off. Best for marketing and lead-gen agencies.
Niche or vertical outreach: Target a specific category (for example specialty coffee roasters) with intent keywords and exclusions. Best for founders and growth teams.
Feed an outreach campaign: Use the exported list as the input to a personalized email workflow. Best for teams pairing extraction with outbound.
Market and competitor mapping: See who operates in a category and location, with ratings and review counts. Best for market research and expansion planning.

Example prompts and outputs

Starter prompt (copy/paste)

Find me coffee shops and roasters in Wiesbaden, Germany with at least 4 stars and 50+ reviews. I want name, phone, website, and email. Skip Starbucks. Fetch about 50 leads.

Example 1: First run for a new profile

Prompt (input): Find dentists in Austin, USA with 4+ stars, I need phone and website, about 30 leads.

Output: The agent confirms the profile (category, location, rating, required fields, count), runs the scrape, then returns a ranked list of the top-matching dentists with name, category, address, phone, website, rating, and review count, plus a summary like "Scraped 30 places, 24 passed your filters, 24 with a phone, 19 with a website. Saved your list."

Example 2: Re-run for more leads

Prompt (input): Get me more coffee shop leads in Islamabad.

Output: The agent recalls your saved Islamabad coffee-shops profile, asks how many more to fetch, scrapes, drops any businesses already in your list, and appends the new ones, then re-ranks the full list best-first.

Example 3: CSV export

Prompt (input): Export the Austin dentists as a CSV with name, phone, website, rating, and a blank status column, sorted by rating.

Output: A clean CSV built to that exact spec from the cleaned data, ready to download and drop into your outreach tool or CRM.

Why you need this template

One chat request replaces hours of manual Google Maps searching, copy-pasting, and cleanup.
Every list is cleaned, deduplicated, and ranked best-first, so it is ready for outreach without extra work.
Your customer profile is saved, so every future run is a one-line request.
You stay in control of spend: the agent always confirms the lead count and only enables paid enrichment when your profile needs it.
Honest results: the agent flags thin data and missing fields and never fabricates contacts or ratings.