Build a Public Business Prospect List From Search Results

1. Define which businesses belong in the list

Write your inclusion rules before you search. An inclusion rule states what a business must show to enter your list. An exclusion rule states what you will leave out.

For this guide’s proposed search, the scope is plumbing businesses that state they serve Toronto. This is a research scope, not a finding about any real business.

Use these steps to define your own scope:

1. Name the required service. State whether a general service claim is enough or whether you need a specific service.
2. Name the required location. State whether a business must have an office there or explicitly serve that area.
3. Require a public company website with evidence of business identity and fit.
4. Exclude directories, job advertisements, and pages that do not represent a business under your rule.
5. Define how you will record uncertain names, website relationships, or service claims.
6. Set a review point before you use the list for another purpose.

Keep an office location separate from a service area. A company can state one without stating the other. A Toronto address alone does not prove that the company offers your required service. A service page alone does not prove that the company serves Toronto.

The supplied buyer context asks about Google Maps scraping tools for local business leads. That wording makes location checks central to this method. However, it is one captured public search title and snippet. The full page was not checked in the supplied evidence. It does not prove search demand or establish Maps support. Public local-business question — retrieved August 11, 2026; title and snippet only, publication date unknown

2. Prepare a small, recorded search

A query is the exact text you submit to a search engine. A query panel is a short, recorded list of those queries. Begin with one query. Add another only when it addresses a clear gap in your research.

Save the query, requested country, requested language, and result limit before collection. These settings describe the search request. They do not establish a company’s service area.

An Actor is a hosted data collection program. The checked public default build of Chronometrica’s Google Search Scraper declares the input fields used below. JSON is a text format for storing named values and lists.

{"country":"CA","engines":["google"],"language":"en","maxResultsPerExecutedQuery":3,"queries":["plumbing company Toronto services"]}

This proposed input passed a check against the public build’s input rules. No Actor run occurred for this guide. The check confirms the input format and permitted values. It does not confirm available results, successful collection, or coverage. Public default build 1.0.15 — checked October 25, 2026; input rules and listed output fields checked, no live run

Replace the sample values as follows:

1. Replace the whole `queries` list with your intended query. Keep one query for the first collection.
2. Keep `engines` set to `["google"]` for this method.
3. Replace `country` if needed. `CA` requests Canada.
4. Replace `language` if needed. `en` requests English.
5. Set `maxResultsPerExecutedQuery` to a small limit. The checked permitted range is 1 to 100.
6. Save the complete input with your research notes.

A cap is the maximum result count requested for each executed query. The sample cap is three results. It is not a promise of three returned results or three qualified businesses.

You can start from the public Google Search Results Scraper Task. A Task is a saved input preset for an Actor. Its public page was checked, but its current saved settings were not verified. Inspect those settings and replace the sample values before use. Do not assume that the preset contains the input above. Google Search Results Scraper Task — checked October 25, 2026; public page presence verified, saved settings and runtime unverified

3. Keep search records separate from business records

A URL is a web address. A domain is the website name within that address. An observation is one recorded search result. Several observations or URLs can concern one business.

Keep a source table for the returned search records. Build your business list in a separate table. This preserves the evidence when several pages support one business.

Retain these documented fields without changing their source values:

A flag is a true-or-false value. Here, `isSponsored` records whether the result is marked as sponsored. Keep that distinction visible. Paid placement does not establish business fit.

Do not treat a title or snippet as a checked company fact. Use it to choose which page to inspect. The field name `canonicalUrl` also does not prove that the destination belongs to a particular company.

For the separate file-preservation task, see Can You Export All Google Search Results to a Spreadsheet?. Keep the export intact while you add your manual research in another table.

After your proposed collection, inspect its actual state before you use its records. Record whether it finished, failed, or was interrupted. Preserve any available explanation for a short or empty result set. These are reader actions; they are not reports of a run performed for this guide.

4. Check the business on its official website

Open each candidate website manually. Your goal is to establish a public business identity and its stated fit.

A public business identity is the name and website that the business uses in public. A legal entity is an organization recognized in law. This method does not establish legal identity merely because a website exists.

Use the same review sequence for each candidate:

1. Check whether the page represents the business itself or another publisher.
2. Record the business name as the website states it.
3. Record the address of the official page you checked.
4. Find a passage that states the required service.
5. Find a passage that states the required service area or office location.
6. Copy the exact relevant passages into your research table.
7. Record your inspection time separately from the search observation time.
8. Assign a review state and a reason.

The service and location evidence can come from different official pages. Keep both page addresses when that happens. A homepage may establish the public name, while a service page supports the service claim.

Do not replace an absent passage with an assumption. If a page states “repairs” but does not identify plumbing, leave the service question unresolved. If it lists an office but gives no service-area statement, apply the location rule you wrote before searching.

An official website supports a claim about what the business states publicly. It does not independently prove service quality, current capacity, or willingness to buy from you.

If the destination is unavailable, record that failure. Do not use the search snippet to fill the evidence gap. How to Build a Crawl Queue From Bing Search Results explains a related destination-review task. That guide uses Bing and builds a page collection queue. Here, the deliverable is a checked business list.

5. Resolve repeated names and website relationships

Choose one consistent public name for each checked business row. Keep other observed names in an alias field. An alias is another name used for the same candidate business.

Merge records only when the public evidence supports the relationship. Two pages on one domain can support one business. However, a domain is not automatically one legal entity. Two domains with similar names are not automatically the same business.

For each possible repeat, ask:

1. Do the official pages use the same business name?
2. Does an official page connect the names or websites?
3. Does the evidence support one public business identity under your counting rule?
4. What remains uncertain?

Retain all supporting source references after a merge. Do not delete the second service page merely because the business already has a row.

Keep unclear relationships in a separate review note. For example, a similar name on another domain may require another official-page check. Do not count it as a second qualified business until that check supports the decision. Do not merge it simply to reduce the count.

This approach keeps page counts and business counts separate. It also lets another reader review why you combined some records and kept others apart.

6. Save a business qualification sheet

Qualification means checking a business against your stated rules. It does not mean that the business is ready for contact.

Use one row for each supported candidate business identity. Keep unresolved source pages in the source table until you can connect them to a business. This prevents an inaccessible page from becoming an invented company row.

Save these fields in the business sheet:

Use “provisional fit” when the official evidence meets your declared rules at inspection time. Use “excluded” when evidence supports a reason to leave the item out. Use “needs review” when a required fact remains unknown.

Record only public company facts relevant to this decision. This method does not add private email addresses, personal phone numbers, inbox information, or owner-only account records. It also does not verify willingness to receive sales messages.

7. Worked example: six URLs support two businesses

Original fiction. All business names, page descriptions, passages, and counts in this section are invented for teaching. This is not an Actor run, customer result, or observed company finding.

The fictional rule requires a plumbing business that explicitly states it serves Toronto. A fictional search panel returns six distinct URLs. This example is separate from the proposed one-query input with a three-result cap.

On the fictional Alder service page, the reviewer finds: “Alder Plumbing provides plumbing repairs throughout Toronto.” On the fictional Cedar homepage, the reviewer finds: “Cedar Repair offers plumbing repair services in Toronto.” These invented passages show the kind of exact evidence to retain.

The reviewer creates two provisional business rows:

The source arithmetic is:

• 3 supporting official-page URLs + 2 excluded URLs + 1 unresolved URL = 6 discovered URLs.
• 2 Alder official-page URLs support 1 candidate business.
• 1 Cedar official-page URL supports 1 candidate business.
• 1 Alder candidate business + 1 Cedar candidate business = 2 provisional business rows.

U3 supports no additional business entry under this review. U4 is outside the rule. U6 remains an unresolved source, not a confirmed third business and not proof that no business exists.

The correct handoff is two provisional business rows, their exact fit passages, two exclusions, and one unresolved source. Six URLs do not equal six companies or six contacts ready for outreach.

8. Check the list before choosing the next action

Review the complete sheet against the original scope:

1. Confirm that every provisional business has official identity evidence.
2. Confirm that each service and location decision has a supporting passage.
3. Check that merged records retain their source references.
4. Keep uncertain names and domain relationships visible.
5. Separate discovered URLs, reviewed pages, and qualified business counts.
6. Preserve excluded and unresolved sources with reasons.
7. Assign a next review to each provisional business.

Keep missing values distinct from zero. A missing snippet is unknown text, not evidence of an empty page. A missing rank is not rank zero. An unavailable website is not a confirmed zero-service business.

Zero provisional businesses can be a valid result if you completed the stated review and none met the rules. If required pages remain unchecked, describe the review as incomplete. Do not report unknown outcomes as confirmed failures.

Choose the next action from the gap. Refine the query when it finds the wrong kind of page. Check another official page when a required claim is missing. Investigate identity when names or domains remain unclear. Keep any added search as a separate recorded sample.