Guides for everything, an assistant that's read all of it, and humans when you want them.
New here? Three ways to start — check a list you already have, connect your form tool or CRM with no code, or protect your website.
Quick start guide →Platform installs (WordPress, Shopify, WooCommerce, Webflow…), the full REST API reference, CRM recipes and troubleshooting.
Read the docs →The assistant in the corner has read every guide and FAQ on this site — ask it anything, any hour. Humans reply within one business day.
We use high-precision structural validation. Hard blocks are reserved for high-confidence junk (disposable domains, email domains that can't receive mail, honeypot hits), borderline cases get a soft "suspicious" verdict you control, and every decision ships with plain-English reasons so you can audit any of them.
Any website that can carry one script tag: WordPress (Contact Form 7, WPForms, Gravity Forms, Elementor…), WooCommerce, Shopify, Webflow, Wix, Squarespace and custom sites (website setup guide). CRMs (GoHighLevel, HubSpot, Salesforce, Pipedrive), Facebook lead ads and form tools (Typeform, JotForm, Webflow…) connect with no code — how to connect. GoHighLevel also has a one-click app on its Marketplace.
No. Upload a list of leads you already have, or have your form tool, CRM, Zapier or Make send LeadShield a copy of each new lead from your private lead-intake URL — nothing changes on your website (how to connect). For blocking fakes in real time on your own forms, use the WordPress plugin or paste one line before </body> — about 60 seconds, with step-by-step guides in the docs.
The script is tiny and loads asynchronously; visitors see nothing until they submit. Validation typically completes well under a second, and there's a hard 5-second timeout that lets the form through if we're ever slow — fail-open, always.
valid (0–39): safe — send to sales. suspicious (40–69): borderline — nurture rather than call; blocked by default, but you can allow them through flagged. invalid (70+): high-confidence junk, blocked. Full routing playbook in the docs.
Check the reason in your dashboard first. If it was "suspicious", set data-block-suspicious="false" to only hard-block outright junk, or use the onValidation callback for full control. Flat-out wrong verdict? Send it to us — we tune the engine on real cases.
Check for stray spaces, and that it's in the X-API-Key header (API) or data-api-key attribute (snippet). You can regenerate a fresh key in the dashboard — the old one stops working immediately, so update your site after.
We email you at 80% and 100%. At the limit, submissions pass through unvalidated until the 1st — or protection resumes instantly when you upgrade from your dashboard.
On a paid plan, Dashboard → Manage billing opens the secure Stripe portal: switch plans, update your card, download invoices, or cancel — all self-serve, effective immediately. Cancelled plans run to the end of the paid period, then move to the Free plan (100 checks a month, with your setup kept). On the Free plan or a trial there's nothing to cancel — pick a plan from your dashboard whenever you want to upgrade.
That's us — LeadShield.ai bills through its parent company, Autonoma. The amount always matches your plan, and every invoice is in your billing portal.
Screened lead data is processed in real time and kept on a rolling 90-day window for your dashboard history, then deleted automatically. Agency plans can shorten this — down to zero retention. Payments are handled entirely by Stripe; we never see card numbers. Details: privacy · security.
Yes — free forever, no card, no time limit: 100 validations a month with email, disposable-domain and phone checks, behavioural bot detection, IP screening, connections to any form tool or CRM, and the full dashboard. AI contextual analysis, device fingerprinting, webhooks and the ad-platform exports come with Pro. Website forms on the Free plan carry a small "Protected by LeadShield" badge. Start free.
Every account starts with 14 days of Pro — every feature — and no card. When they end your account moves to the Free plan rather than switching off: everything you connected keeps working at 100 validations a month, and your data and API key are preserved. Upgrade whenever you're ready.
Yes — webhooks are included on Pro and Agency (and in every account's first 14 days). On your dashboard's Settings tab, paste an HTTPS endpoint into the Webhook notifications card, choose every lead or only flagged leads, and hit Send test to confirm it works. We POST each verdict the moment it happens, signed so you can verify it came from us. Full payload and verification sample in the API docs.
Yes. On your website, a blocked lead never reaches your thank-you page, so a conversion there never counts it, and the verified lead signal lets button-click tags count only leads that pass the checks. To stop your ads reaching the same junk again, Pro and Agency add two downloads: an IP exclusion list for Google Ads, and a SHA-256 hashed email list for Google Customer Match or Meta Custom Audiences — upload it as an exclusion audience. Hashing happens before the download, so raw addresses never leave your account.
A switch on your dashboard's Connect tab, on every plan. LeadShield checks every website lead as normal but blocks nothing, and you get a report every seven days showing what it would have stopped and why. When you trust it, press Turn on blocking: nothing to change on your site. How it works
No. When a visitor has started your form and goes to leave without sending it, one short note says it hasn't been sent and takes them back to it. It runs only in their browser: it checks whether fields are filled in, never what they say, and nothing is sent or stored unless they press your form's button. Switch it on under Connect → Block fakes on your website, or in the WordPress plugin's settings. How it works
Your Overview tab groups junk leads by IP address, device, email and phone, so you can see who keeps coming back, even when they change their email address every time. On Pro and Agency you can download their IP addresses for Google Ads IP exclusions; any address that also sent you a good lead is left out, so a shared office or mobile network is never excluded.
Yes, on the Agency plan (and in your first 14 days): your dashboard's Tools tab makes a report for each client from their real leads, with your name and logo, as a link or a one-page PDF. It shows totals and reasons, never the leads themselves. On the paid Agency plan your clients' forms can also say "Protected by" your agency. How it works
Yes. The LeadShield.ai app is on the GoHighLevel Marketplace and installs in one click. It checks every new contact as it arrives and tags the fake ones leadshield-flagged, with a note saying why, so your automations can skip them. It works on every plan, and agencies can install it across their client accounts. If you already send contacts with a webhook workflow, you don't need it once the app is installed. GoHighLevel guide
Yes. In WordPress, go to Plugins → Add New, search for "LeadShield", install and activate it, then paste your API key under Settings → LeadShield. WordPress guide
The lead's text includes a known prompt-injection pattern: words written to make an AI assistant do something, sometimes hidden in invisible characters. It doesn't change the lead's score; it's a warning for anyone passing leads to an AI tool. On Pro and Agency your webhook carries message_safe, a cleaned copy to use instead. More
On Pro, Agency and your first 14 days, each real lead with a message gets a brief: what they want, the words from their message that show it, a question to ask first and a reply draft you can copy. Hot means the message asks for a quote, booking or call and states a timeline. It describes the words, not whether they'll buy. Drafts are never sent for you. More
On our public status page, where UptimeRobot checks the API around the clock and shows its uptime history. If LeadShield is ever unreachable, your forms still submit normally: submissions go through unchecked rather than being lost.
Scan it with Form Check: enter the page address and it shows which form tool the page uses and which spam protection can be seen from outside (reCAPTCHA, hCaptcha, Turnstile, a hidden trap field, Akismet, CleanTalk or LeadShield). It reads the page and never submits your form. Filters that run on your server can't be seen from outside, so it may under-report.
Yes, all without signing up: a spam message checker (paste a message and the AI says whether it's spam or a sales pitch, or a real enquiry), a disposable email checker and a fake lead checker. They're all on the free tools page. Nothing you paste is stored.
Each has a side-by-side comparison with facts from their own pages, dated, including the rows where they come out ahead. In short: CAPTCHAs check whether the visitor is a bot; LeadShield checks the lead itself (email, phone, IP, how the form was filled in, and on Pro what the message says), and works alongside a CAPTCHA.
No. Any email address works, and the company name is optional. You can also sign up with your Google account.
Yes — the free lead audit needs no account, no card and no code. Upload a CSV of up to 250 leads you already have and you'll see immediately how many are disposable addresses, email domains that can't receive mail, typos or duplicates, with a full breakdown of what's wrong. Three audits per day. Your file isn't stored. If you want the annotated file back with a verdict on every row, or every new lead checked automatically, that's what starting free adds.
Yes — that's the fastest way to see what LeadShield does. Export a CSV from your CRM, then on your dashboard's Tools tab use Clean a list you already have. You get the same file back with a verdict, risk score and plain-English reasons added to every row. Up to 100 rows per upload on Free, 1,000 on Pro and 5,000 on Agency (your first 14 days get the Pro allowance). Note that bulk checks run the deterministic layers only — email, disposable domains, MX, typos, phone and duplicates — because behavioural, IP and AI analysis need a live form submission. Anything flagged is genuinely junk, but a clean-looking row can still be caught in real time.
It's the junk we blocked that arrived from a paid ad click, multiplied by your cost per lead. We identify paid clicks from the gclid, fbclid and similar parameters on your landing-page URL — no cookies or tracking storage involved. Set your own cost per lead in the dashboard; leave it blank and we use a $45 industry estimate, which is exactly that — an estimate, not a measurement of your business.
On Agency (and during your first 14 days), yes — dashboard → Settings → Advanced controls. Verdicts sit on a 0–100 risk score, with invalid at 70 and suspicious at 40 by default. Lower the invalid cut-off to block more aggressively, or raise it to let borderline leads through. The same panel controls how long lead data is kept.
Every message gets a reply within one business day — from a person, not a queue.
Contact supportChecking whether we're up? See the live status page.