The AI that reads
every lead.
Every phone call. Every form submission. TrolleyShield uses AI to read them in real time, understand what the lead actually wants, and feed only the real ones to your ad platforms, so your marketing optimizes toward revenue instead of noise.
The Innovation
One AI engine. Two surfaces. Zero junk.
Ad platforms like Google reward you for conversions, but they can't tell a real customer from a spam bot, a job seeker, or a wrong number. So they optimize toward whatever fires a conversion tag, including the junk. The result is wasted budget and a marketing engine pointed at noise.
TrolleyShield puts an AI classification engine in front of that. Whether a lead arrives as a recorded phone call or a web form submission, the same engine transcribes or parses it, classifies the intent with a confidence score, and decides whether it's a real lead worth passing on as a conversion.
The breakthrough isn't spam-blocking. It's understanding every lead well enough to feed ad algorithms clean signal, across voice and text, in real time, automatically.
Reads voice
Transcribes and classifies live phone calls into 8 intent categories.
Reads text
Analyzes form submissions for legitimacy, spam, and relevance.
Feeds platforms
Fires offline conversions to Google Ads, GA4, and Meta — for real leads only.
How It Works
From raw lead to clean signal, in seconds
Capture
A call hits a tracking number or a form is submitted. TrolleyShield captures it in real time, along with attribution data, behavioral signals, and source.
Transcribe & Parse
Calls are transcribed to text; forms are parsed. Both are normalized into the same structured input for the AI engine.
Classify Intent
The AI assigns an intent — qualified lead, general inquiry, existing customer, vendor, job seeker, spam, and more — each with a confidence score and a recommended action.
Gate the Conversion
Only leads that clear the intent and confidence thresholds are fired as conversions. Junk is filtered out before it ever pollutes Smart Bidding.
Proof · Call Shield
Reading 388 phone calls a home remodeler couldn't
A Pennsylvania home-remodeling company runs every inbound call through Call Shield. Over ten weeks, the AI analyzed 388 phone calls, turning a black box of ringing phones into a clear map of demand, lead quality, and a costly blind spot the owner had never quantified.
What the callers actually wanted
AI intent classification across every transcribed call.
Which sources drove real leads
Share of calls classified as qualified leads, by channel.
Percentages show the qualified-lead rate within each source.
The discovery
56% of calls were going unanswered.
By scoring every call, the AI surfaced something no spreadsheet had: 219 of 388 calls — more than half of every lead the business paid to generate — were missed. With 67% first-time callers, that's new demand ringing into the void. The owner couldn't fix a leak nobody had measured. TrolleyShield measured it.
Calls analyzed
388
Qualified leads found
38%
of analyzed calls
Non-customer noise flagged
18%
vendors, spam, wrong numbers
Missed-call leak surfaced
56%
Proof · Form Shield
Cutting one company's ad spend by 65%
A national pharmacy-services firm spent over $7,000 a month on Google Ads. TrolleyShield's AI proved that 74% of their paid clicks were never real leads — spam, irrelevant, and suspicious submissions Google's Smart Bidding was optimizing to get more of. Form Shield gave the algorithm clean signal, and the spend fell month after month.
Monthly Google Ads spend
As AI-filtered conversions retrained Smart Bidding, spend dropped from $7,320 to $2,530 — without losing real leads.
What the AI found in the forms
Intent classification across 317 form submissions.
74% of paid clicks were never real leads — and every one would have counted as a conversion.
Ad spend reduction
65%
$7,320 → $2,530 / mo
Paid clicks that were junk
74%
filtered from the signal
Submissions analyzed
317
and counting
AI confidence
88%
average classification
The Platform
Two products, one intelligence layer
Call Shield
Dynamic number insertion, call tracking, and AI intent analysis on every recorded call — with attribution that tells you which campaigns drive real phone leads.
- → 8-category intent classification
- → Automatic transcription
- → gclid / gbraid attribution
Form Shield
Real-time AI analysis of every form submission, filtering spam, bots, and irrelevant inquiries before they reach your CRM or pollute your ad platform's bidding signals.
- → Intent + behavioral scoring
- → IP reputation intelligence
- → WordPress + API integration
Both products share one AI classification engine and one conversion pipeline to Google Ads, GA4, and Meta. Whether your leads call or click, TrolleyShield makes sure your marketing only ever optimizes toward the real ones.
Capabilities
Everything TrolleyShield does
A complete picture of the platform — from call tracking and AI intent analysis to spam filtering, attribution, and conversion delivery across every major ad platform.
Call Shield
- →Local & toll-free tracking numbers with area-code selection
- →Dual-channel call recording with transparent forwarding
- →Automatic speech-to-text transcription of every call
- →AI intent classification across 8 categories with confidence scores
- →Dynamic Number Insertion (DNI): a unique number per website visitor
- →Caller location enrichment and first-time-caller detection
- →Voicemail detection and per-number call context rules
- →Manual action overrides with an audit trail
Form Shield
- →Real-time AI spam, bot, and phishing detection — no CAPTCHA
- →Behavioral scoring from mouse, typing, scroll, and paste signals
- →IP intelligence: VPN, proxy, Tor, and hosting detection
- →Threat-level scoring (low → critical) on every submission
- →WordPress / Gravity Forms plugin with usage widget
- →Per-intent configurable actions: allow, flag, or block
- →Public analyze API for any form or framework
Custom Forms
- →No-code drag-and-drop form builder
- →Full styling controls or CSS-only "no style" output
- →TrolleyShield-hosted page, script embed, or iframe
- →Built-in behavioral fraud detection on every submission
- →Automatic UTM, click-ID, and referrer capture
- →Success message or redirect on submit, with API failsafe
AI Intelligence
- →Unified engine reads both voice transcripts and form text
- →Intent + 0–100 confidence + written reasoning on every lead
- →Recommended action returned for each classification
- →Account-level business context tunes accuracy per client
- →Up to 25 custom service tags auto-assigned from content
- →Service-tag conversion gating (e.g. fire only on "roofing")
Integrations
- →Google Ads offline conversions (separate call & form actions)
- →GA4 server-side events: leads, conversions, and blocked spam
- →Meta Conversions API with SHA-256 hashed matching
- →CallRail two-webhook sync with late-transcription backfill
- →GoHighLevel "Analyze Call" & "Analyze Form" workflow actions
- →Conversion retries, status tracking, and error logging
Attribution & Platform
- →Captures gclid, gbraid, wbraid, fbclid, fbp & msclkid
- →Full UTM, landing-page, and referrer tracking on every lead
- →GA4 client-ID stitching for session-level attribution
- →Developer REST API with bearer-token keys (up to 10/account)
- →Multi-organization accounts with owner / admin / viewer roles
- →Usage tracking, overage billing, and CSV / branded-PDF export
The AI's intent vocabulary
Call intents
Form intents
Each classification returns an intent, a 0–100 confidence score, a written reasoning, a recommended action, and any matching service tags — used to decide which leads become conversions.
Ad platforms got smarter.
The signal they run on didn't.
TrolleyShield is the intelligence layer between your leads and your ad platforms — reading voice and text alike, understanding intent, and making sure every dollar of ad spend chases a real customer. Not a bot. Not a job seeker. A lead that can actually become revenue.
See TrolleyShield