GHL Prospecting Audit Accuracy SOP
Purpose
This SOP defines quality control procedures for GHL Prospecting audit reports before they are used in outreach. Audit reports are auto-generated from public data and can contain inaccuracies. Sending outreach based on incorrect findings damages credibility. This SOP ensures every report used for outreach has been verified.
Scope
Covers:
- Which audit sections to include or exclude per prospect
- Section-by-section accuracy verification
- Report refresh rules
- Red flags that should pause outreach
- Quality checklist before adding a prospect to the pipeline
Does Not Cover:
- How to generate audit reports (see KB-015: GHL Prospecting Tool)
- Email outreach content (see Prospecting Outreach Email Templates)
- Call content (see Prospecting Cold Call Scripts)
- Cadence and pipeline state management (see Outreach Cadence SOP)
Section Selection Rules
Not every audit section applies to every prospect. Including irrelevant sections (e.g., Website Performance for a business with no website) weakens your outreach.
A worked example of every section is filed in Drive. Symphony Core's own audit, run 2026-09-23, at
sc-docs-storage/03-sales/enablement/2026-09-ref-symphony-core-own-marketing-audit-report.pdf— the tool's own filename wasreport_2026-09-23_en-US.pdf. Open it before verifying your first report: the section names in this SOP are the summary-page labels, and the section headings the report prints are different. KB-015 carries both, side by side.
| Section | Include? | Condition |
|---|---|---|
| Business Details | Always | Core identity -- always relevant |
| Technology Stack | Always | Reveals what tools they use (or don't) |
| Google Business Profile | Always | Essential for any local business |
| Listings | Always | Directory presence matters for all local businesses |
| Online Reputation | Always | Reviews are universal to local businesses |
| Website Performance | Conditional | Only if the business has a real website (not just a Facebook page or Yelp listing) |
| SEO Analysis | Conditional | Only if the business has a real website with indexable pages |
| AI Search Visibility | Conditional | Local presence always applies. Crawler access, AI readability and agent readiness only if the business has a real website |
Rule: If a business has no website, do not reference Website Performance, SEO Analysis or the website-based AI Search Visibility areas in your outreach. Focus on GBP, Listings, Reputation and AI local presence instead.
Section-by-Section Accuracy Guide
1. Business Details
| Item | How to Verify | Common False Negatives |
|---|---|---|
| Business name | Google search the business | May flag a DBA vs. legal name difference as an error |
| Address | Check GBP listing directly | Suite/unit numbers often mismatch across sources |
| Phone number | Call the number or check the website | Tracking numbers may differ from main line |
| Website URL | Visit the URL | May flag a redirect (e.g., www to non-www) as missing |
| Hours | Check GBP or website | Seasonal hours cause false discrepancies |
Accuracy confidence: High -- this section is usually reliable.
2. Technology Stack
| Item | How to Verify | Common False Negatives |
|---|---|---|
| CMS detection | View page source or use Wappalyzer | Custom-built sites may show "No CMS detected" |
| Analytics | Check for GA4/GTM tags in page source | Tag Manager containers may hide analytics from detection |
| Chat widgets | Visit the site and wait 5-10 seconds | Some widgets load on delay or only on specific pages |
| Scheduling tools | Look for booking links in the nav/footer | Embedded iframes may not be detected |
Accuracy confidence: Medium -- detection depends on how the site is built. False negatives are common for SPAs and heavily cached sites.
3. Google Business Profile (GBP)
| Item | How to Verify | Common False Negatives |
|---|---|---|
| Profile exists | Search "[Business Name] [City]" on Google | Multi-location businesses may have a different GBP than expected |
| Photo count | Open the GBP listing and check Photos tab | Count may lag behind actual uploads by a few days |
| Recent posts | Check the Updates/Posts section on GBP | Posts expire after 7 days -- a "no recent posts" flag may be accurate |
| Attributes | Check the "About" section on GBP | Some attributes are Google-sourced, not owner-set |
Accuracy confidence: High -- GBP data is pulled directly from Google's API.
4. Listings
| Item | How to Verify | Common False Negatives |
|---|---|---|
| Directory presence | Spot-check 2-3 directories (Yelp, BBB, Facebook) | Some directories block automated checks |
| NAP consistency | Compare name/address/phone across 3+ sources | Minor formatting differences (St vs. Street) may be flagged |
| Listing count | Compare against industry benchmarks | New businesses legitimately have few listings |
Accuracy confidence: Medium -- directory data varies by scraping access. Verify at least 2-3 listings manually.
5. Online Reputation
| Item | How to Verify | Common False Negatives |
|---|---|---|
| Review count | Check Google Maps listing directly | May not include reviews from other platforms |
| Average rating | Check Google Maps listing directly | Rating may have changed since report generation |
| Response rate | Scroll through reviews to check owner replies | Recent responses may not be reflected yet |
| Review recency | Check dates of last 5 reviews | Seasonal businesses may have legitimate gaps |
Accuracy confidence: High for Google reviews. Medium for cross-platform aggregation.
6. Website Performance
| Item | How to Verify | Common False Negatives |
|---|---|---|
| Page speed score | Run Google PageSpeed Insights manually | Scores fluctuate by 5-10 points between runs |
| Mobile responsiveness | Open site on mobile or use Chrome DevTools | Some sites use separate mobile URLs (m.example.com) |
| SSL certificate | Check for padlock icon in browser | Recently renewed certs may show as missing briefly |
| Core Web Vitals | Check Google Search Console (if accessible) | Lab data vs. field data can differ significantly |
Accuracy confidence: Medium -- scores vary between testing tools and time of day. Use as directional, not exact.
7. SEO Analysis
| Item | How to Verify | Common False Negatives |
|---|---|---|
| Title tags | View page source, check <title> | Dynamic titles (React/JS apps) may not be detected |
| Meta descriptions | View page source, check <meta name="description"> | Same dynamic rendering issue as title tags |
| Schema markup | Use Google's Rich Results Test | Schema may exist but not be detected by the audit tool |
| Heading structure | View page source, check H1-H6 tags | Visual headings using styled <div> won't be detected |
Accuracy confidence: Low-Medium -- SEO audits are the most prone to false negatives, especially for JavaScript-rendered sites. Always cross-check with a manual page source review if you plan to reference SEO findings in outreach.
8. AI Search Visibility
The section is a readiness check. It scores whether AI assistants can find, read and act on the business. It does not ask any AI assistant anything. Scoring detail is in KB-017.
| Item | How to Verify | Common False Negatives |
|---|---|---|
| Crawler access (GPTBot, ClaudeBot, PerplexityBot, Google-Extended) | Open <domain>/robots.txt and look for a Disallow under those user agents | A block set at the CDN or firewall (for example Cloudflare bot rules) does not show in robots.txt, so "Allowed" can still be wrong |
| Phone number found / call action | Open the home page on a phone and tap the number | A number inside an image or a JS-rendered header can be missed |
| Inquiry form / booking button | Check the home page, then the contact and booking pages | Agent readiness tests one page only. A business with a contact form or booking page elsewhere still shows "Not ready" |
| Structured data fields | Google's Rich Results Test on the home page | Schema injected by JavaScript or on another page may not be counted |
| Content dependent on JavaScript (%) | View page source and compare with the rendered page | Noisy on Elementor, Wix and Squarespace sites, which load most content through scripts even when crawlers read it fine |
| Business directories needing attention | Spot-check one listed directory before quoting the count | Vendor directory data; same scraping limits as Listings |
Accuracy confidence: High for crawler access and the phone check (binary, and the prospect can confirm them in about twenty seconds). Medium for structured data and directories. Low for the JavaScript percentage and for agent readiness "Not ready" until other pages are checked.
Rules for outreach:
- Prefer crawler access and the phone check as the AI finding to quote. They pass the twenty-second test in the Pre-Call Gate.
- Check for a contact or booking page before quoting "not ready" for inquiries or booking.
- Do not lead with the JavaScript percentage.
- Spot-check one directory before quoting a directory count.
- Never tell a prospect the report shows what ChatGPT, or any assistant, says about them. It does not. Say it shows whether AI assistants can find, read and act on the business.
Report Refresh Rules
| Scenario | Refresh Required? | Reason |
|---|---|---|
| Report is less than 7 days old, first outreach | No | Data is fresh enough |
| Report is 7-14 days old, first outreach | Yes | Data may have changed |
| Report is older than 14 days | Yes | Stale data risk is high |
| Before sending Follow-Up 1 (Day 3) | No | Too soon for data to change |
| Before sending Follow-Up 2 (Day 7) | No | Still within original report window |
| Before re-engagement (Day 60-90) | Yes | Significant time has passed |
| Prospect mentions they made changes | Yes | Verify their claims before responding |
Red Flags: When to Pause Before Outreach
Stop and investigate before proceeding with outreach if any of these conditions exist:
- Business appears to be closed: GBP shows "Permanently Closed" or website is down
- Business is already an SC client: Check the CRM contact list before reaching out
- Business is a direct competitor: Marketing agencies, web design firms, CRM consultants
- Audit data is clearly wrong: NAP data doesn't match any online source
- Business is outside SC's service area or ICP: Non-local businesses, enterprise companies, government entities
- Recent negative press: Google the business name -- if there's active controversy, skip
- No contact information found: If you can't find a real person's name and email, the outreach won't land
Action on red flag: Do not add to the pipeline. Note the reason in the prospect record and move on.
Pre-Call Gate
The checklist below gates every first touch, spoken or written. It applies to a call exactly as it applies to an email, and on a call it matters more: a finding quoted aloud cannot be retracted, and an inaccurate one confirms the prospect's suspicion that the call is a scam.
Two call-specific additions to the standard checks:
- The finding you plan to lead with must be verifiable by the prospect in about twenty seconds on their own phone. If they cannot check it while you are talking, choose a different finding.
- Record which finding you used in
Audit Finding Usedon the opportunity, so the call and the paired email tell the same story and so opener performance can be compared later.
If the audit resolved to the wrong business, do not call. Log wrong-business and flag the report for refresh. See the Prospect Phone and Call Execution SOP.
Quality Checklist
Complete this checklist before a prospect is cleared for outreach. Passing it is what the prospecting-audit-verified tag attests to:
- Business is confirmed to be open and operating
- Business is not an existing SC client or active lead
- Business fits SC's ideal client profile (local service/professional business, 1-50 employees)
- At least 3 audit sections have been spot-checked for accuracy
- Conditional sections (Website Performance, SEO) are excluded if no real website exists
- At least 1 clear, verifiable finding identified for outreach personalization
- A real contact person has been identified (name + email or LinkedIn). See Prospect Email Discovery SOP for step-by-step lookup methods.
- Report is less than 7 days old (or has been refreshed)
- No red flags present (see list above)
On pass: record the verification in GHL
The checklist result must be visible to the system, not just to the person who ran it:
- Record on the contact: the shareable report link in
audit_pdf_url(open it in a private browser window to confirm it loads without login),Audit Finding Used, andOpener Used. There is no audit-date field — the report's own generation date plus checklist item "less than 7 days old" is the freshness record - Correct the contact's name fields. The prospecting agent writes the business name into
First name— replace it with the verified contact person's first name (andLast namewhere known), or clear it. No outreach mergesfirst_nameuntil this is done - Apply the tag
prospecting-audit-verified. It also triggersC-021.Prospecting - 01.generate audit, which generates the report and writes its link. Since 2026-11-04 it sends no email: the first email is Email 1A, sent by the caller before the call (Execution SOP Step 5) - When contact verification is also complete (
prospecting-contact-verified, per the Call Execution SOP), applyprospecting-cadence-start— the deliberate start signal for outreach. Apply it last, after both verification tags are on: it is the one tag that starts the cadence, whether run by hand or by the C-024 workflows
A prospect failing the checklist gets no tags. Note the reason on the record and either refresh the report or discard.
Re-activation gate
The single permitted re-activation touch reuses this checklist. Before anyone applies prospecting-reactivation-approved to a dormant prospect, refresh the audit and re-run the checklist — the "less than 7 days old" item is the freshness gate the C-024 reactivation workflow relies on; the workflow itself does not check dates. A prospect already tagged prospecting-reactivated has used its one re-activation: do not apply the approval tag again.
Related Documents
- Marketing Audit Production PRD -- The end-to-end process this SOP feeds: prospecting audit input, correction, on-brand SC audit output
- Marketing Audit Correction Worksheet Template -- Per-prospect worksheet that applies this SOP's section confidence levels
- KB-015: GHL Prospecting Tool -- How the prospecting tool works
- Prospecting Pipeline SOP -- Pipeline stage definitions and lifecycle
- Prospect Email Discovery SOP -- Finding a decision-maker's email before outreach
- Prospecting Outreach Email Templates -- Email templates for each outreach stage
- Prospecting Email Outreach Strategy -- Strategic rationale for outreach approach
- Sales Pipeline SOP -- The downstream sales pipeline
Document History
| Version | Date | Changes |
|---|---|---|
| 1.6 | 2026-11-12 | Adds the AI Search Visibility section (seen on SC's report 2026-11-11): selection rule, a verification table, confidence levels and five outreach rules, chief among them that the section never shows what an assistant says about the business |
| 1.5 | 2026-11-04 | The verification tag's side effect restated: prospecting-audit-verified triggers 01.generate audit, which generates the report and no longer sends an email |
| 1.4 | 2026-09-23 | Points at a filed worked example: SC's own audit, every section rendered. Notes that this SOP's section names are the report's summary-page labels and that the printed section headings differ, which is how a blog post came to be drafted from the wrong artifact |
| 1.3 | 2026-08-27 | Field-model simplification (PRD v2.2): verification records audit_pdf_url, Audit Finding Used, Opener Used on the contact only (opportunity fields and Audit Date retired — checklist freshness item is the record). Added the Re-activation gate section: checklist re-run + audit refresh before prospecting-reactivation-approved; prospecting-reactivated marks the one permitted use |
| 1.2 | 2026-08-22 | Quality-checklist pass now writes system state: audit fields recorded, prospecting-audit-verified applied, and prospecting-cadence-start applied last as the deliberate outreach start signal. Removed the stale "Report Ready" pipeline-stage reference (the four-stage pipeline was never built) |
| 1.1 | 2026-08-18 | Added the pre-call gate: the checklist now explicitly gates spoken touches, with a verifiability test on the finding led with and a wrong-business stop |
| 1.0 | 2026-03-02 | Initial version -- section selection rules, accuracy guide, refresh rules, quality checklist |
Document Information
- Version: 1.3
- Created: 2026-03-02
- Updated: 2026-08-27
- Location:
04-operations/sops/ghl-prospecting-audit-accuracy-sop.md