8A Pro • Lead Knowledge Page← Operations Home
Build & maintain leads

Find → qualify → rank → contact → update.

This page is the instruction manual. The home page is where you act.

Open sources.csvOpen parsed.csv
Before You Start • Connect Your Tools

Use the same 8A Pro account everywhere.

Account to use:
Google Drive, Google Drive for desktop, and the Google Drive connection inside ChatGPT should all point to this same account.

1. Sign into Google Drive

  1. Open Chrome and go to Google Drive.
  2. Confirm the profile icon in the upper-right shows the 8A Pro Google account.
  3. If another Google account is active, switch accounts before opening or editing project files.
  4. Open the 8A Pro project folder and confirm you can see index.html, leads.html, website.html, parsed.csv, services.md, assets, and data.

2. Install Google Drive for desktop

This makes the 8A Pro Drive files appear in Windows File Explorer so the project feels like a normal folder on the computer.

  1. Download and install Google Drive for desktop from Google.
  2. When asked to sign in, use 8aproservicesllc@gmail.com.
  3. After setup, open File Explorer and look for Google Drive in the left navigation.
  4. Open My Drive → GPT → 8a-pro-site and confirm the same project files are visible there.

3. Connect Google Drive to ChatGPT

  1. Open ChatGPT in the account Juan and Vianey will use for this work.
  2. Open ChatGPT Settings and find Apps / Connected apps.
  3. Find Google Drive and choose Connect.
  4. When Google asks which account to use, select 8aproservicesllc@gmail.com.
  5. Review the Google permission screen and allow the Drive access ChatGPT requests so it can search and read the project files.
Important: Google Drive for desktop and the Google Drive connection inside ChatGPT are two different connections. Installing Drive on Windows does not automatically give ChatGPT access.

4. Test the connection

Do not assume the setup worked. Ask ChatGPT to prove it can see the project.

Test prompt

“Find services.md in the 8A Pro Google Drive project and tell me the services listed there. Then tell me the names of the files beside it in the project folder.”

You are connected when: ChatGPT can find services.md, read it, and identify the surrounding 8A Pro project files without you uploading them manually.
1 • Find & Download Data

Use sources.csv to find a database, then bring the data to ChatGPT.

Purpose: sources.csv is your list of websites and databases that may contain useful permit, bid, award, development, or contract data. You do not normally look for individual leads inside sources.csv itself — you use it to decide where to go get fresh data.
  1. Open sources.csv.
    Pick one of the listed sources. Start with a High-priority source or one that is due to be reviewed.
  2. Open the website in the URL column.
    Use what_to_check and recommended_filters as your instructions for what to search for on that site.
  3. Find the site's downloadable data.
    Many permit, contract, award, bid, and public-data websites let you export results as CSV, Excel/XLSX, or another spreadsheet file. Apply useful filters first so you are not downloading the entire universe.
  4. Save the downloaded file into the 8A Pro Google Drive.
    Put the raw download in the project data area so ChatGPT and both of you can find the same file later. Keep the original download instead of manually rewriting hundreds of rows.
  5. Ask ChatGPT to examine the downloaded file.
    Tell ChatGPT which file you downloaded, which source it came from, and ask it to compare the rows against services.md and the lead signals described on this page.
  6. Have ChatGPT propose the best leads for parsed.csv.
    Ask for a ranked shortlist with the project, contractor or owner, location, source, likely 8A Pro need, fit score, reason, and next action. Review the suggestions before allowing parsed.csv to be updated.
  7. Update parsed.csv with only the useful leads.
    parsed.csv is the curated result — not a copy of the raw download. The best, most realistic opportunities belong there; the noise stays out.
Example prompt after downloading data

“In the 8A Pro Google Drive, examine the new data file I just downloaded. Use services.md as the business filter and the lead signals in the Lead KB. Find the strongest realistic opportunities for 8A Pro. Rank them from best to worst and explain why. Show me the proposed changes to parsed.csv before you make any changes.”

The whole loop: sources.csv → open a data website → filter/search → download CSV/XLSX → save to Google Drive → ask ChatGPT to analyze → review the shortlist → update parsed.csv.
2 • Maintain the map

Add a new source when you find a useful place.

  1. Go to the first empty row in sources.csv.
  2. Enter a plain-English source name and category.
  3. Record the jurisdiction / area it covers.
  4. Explain what to look for and why the data helps.
  5. Add a suggested review frequency, useful filters and the URL.
  6. Save and confirm the row is still there.
Rule:a reusable website/database belongs in sources.csv. A specific project/company belongs in parsed.csv.
AI • Turning raw data into signals

You are not asking AI to find “tree jobs.” You are asking it to notice activity that may create work.

Concept: A permit, award, planning application, or contractor record is a signal. AI helps read a large batch of those signals, identify patterns, compare them to 8A Pro’s services, and reduce hundreds of rows into a short list worth human review.

Think one step beyond the permit name.

An electrical permit does not mean Juan should do electrical work. A sewer permit does not mean 8A Pro should install sewer pipe. They may still be useful because they can reveal that a larger construction project is starting.

Electrical permit May signal a remodel, commercial build-out, apartment project, or new construction. Look for the owner, GC, address, project value, and related permits.
Sewer / plumbing permit May signal site development, multifamily construction, a major addition, or redevelopment where exterior cleanup, vegetation, trim, punch-list, or maintenance work may follow.
Reroof / exterior permit This is a more direct signal. AI should look for fascia, trim, flashing-adjacent work, caulking, roof maintenance, contractor name, and whether the company does repeat commercial work.
Tree / vegetation / site work Usually a strong direct signal. Look for clearing, pruning, roadside maintenance, weed control, hazardous fuels, landscaping, forestry, or property-maintenance scopes.

A simple AI parsing workflow

  1. Give AI the source data.
    Upload the CSV/XLSX or provide the rows you exported from the permit, bid, award, or planning system.
  2. Give AI the business filter.
    Tell it to use services.md and Juan’s confirmed capabilities. AI should not invent new services.
  3. Tell AI what signals to notice.
    Examples: multifamily, new construction, reroof, exterior alteration, site development, sewer/electrical permits tied to larger projects, vegetation, roadside work, forestry, grounds maintenance, large primes, out-of-area contractors working locally.
  4. Ask for a ranked shortlist, not a giant summary.
    Have AI return only the strongest candidates with project, address, owner/contractor, likely 8A scope, fit score, reason, source row, and next research step.
  5. Human review decides what enters parsed.csv.
    AI proposes. Juan and Vianey confirm whether the scope, geography, company, and next action make sense.
Example prompt to reuse

“Review this permit/contract dataset using services.md as the business filter. Do not assume 8A Pro performs electrical, plumbing, sewer, or other licensed trades. Treat those permits only as possible indicators of larger construction activity. Look for projects in or near our service area where tree/vegetation work, exterior trim, fascia, roof maintenance, caulking, cleanup, punch-list, or local field support could plausibly be needed. Rank the best 10. For each one, show the source row, project, owner/contractor, signal you noticed, likely 8A scope, fit score 1–5, why, and the next thing Vianey should verify before adding it to parsed.csv.”

Important: AI is a filter and research assistant, not proof that a subcontract opportunity exists. A permit can suggest activity; it does not prove that 8A Pro’s scope is available. Verify the project, company, timing, and contact before outreach.
3 • Qualify

Only good opportunities enter parsed.csv.

  1. Name the project / award / permit.
  2. Identify who 8A Pro would contact.
  3. Write the likely 8A Pro scope in plain English.
  4. Score fit from 1–5 and explain the score.
  5. Add outreach status and a specific next action.
4 • Rank

Work the best leads first.

Rank in this order:service fit → geography → repeat relationship potential → contactability → contract value.

A giant contract with no matching scope is not a better lead than a smaller project that repeatedly needs exactly what Juan does.
5 • Contact & update

Every lead must have a next action.

  1. Contact or research the highest-ranked lead with an unfinished next action.
  2. After the action, update outreach_status.
  3. Enter last_contact_date.
  4. Add short notes.
  5. Replace the old next action with the new one.
A lead without a next action is a dead row.
Simple routine

One hour is enough.

15 min:check 3 sources.15 min:collect projects.15 min:qualify the best ones.15 min:work the current top five and update next actions.