Fake Leads from Facebook and Google Ads: Why B2B Forms Fill with Junk

Fake leads are the quiet tax on B2B ad budgets: the form fills, the number rings out, and your sales team learns to ignore the next one. Here is where fake leads from Facebook and Google ads really come from, how to tell fake from junk from wrong-fit, and the fixes that cut them without starving your pipeline.

Why do you get fake leads from Facebook and Google ads?

Because ad platforms optimise for form submissions, and a submission is cheap to produce. Pre-filled forms, broad audiences, free-offer hooks, bots and vendors fishing for contact details all create submissions that look like leads. The platform counts each one as a win and goes looking for more people like them.

Most ‘fake leads’ are not bots at all. In B2B, the bigger share is real people who tapped a form that asked for nothing: a student after a price, a trader with a 50-piece need pointed at a 5,000-piece MOQ, a buyer who never meant to submit. Sort what you receive into four piles before you change a single setting.

• Fake: invalid phone or email, gibberish names, throwaway domains. Fix the form and the traffic quality.
• Spam: bots, or vendors pitching SEO and web design through your enquiry form. Fix with form protection.
• Junk: real people with no buying intent, such as students, job seekers and brochure collectors. Fix the offer, the audience and the questions.
• Wrong-fit: real intent, but below your MOQ, outside your markets or outside your grades. Fix with qualifying questions and clearer ads.

How do you tell fake leads, spam leads and junk leads apart?

Read the leads, not the dashboard. Open the last 30 from each campaign and call every number. Within an hour you will know whether the problem is bad data, bots, low intent or mismatch. The table maps the symptom you see to its usual cause and fix.

Why do fake leads from Facebook ads happen, and how do you stop them?

Instant forms make submitting almost effortless. Meta pre-fills name, phone and email from the person’s profile, one thumb-tap sends it, and broad audiences with Advantage+ expansion let the algorithm chase whoever submits cheapest. The result is wrong numbers, accidental taps and people who forgot they filled anything in.

Work through these fixes in order, and measure after each one.

1. Change the form type. In Meta’s instant form builder, pick Higher intent instead of More volume. It adds a review screen so people confirm their details before submitting. Expect fewer leads and usually a higher cost per lead; that is the trade you want.
2. Add custom questions with dropdown answers (quantity band, company, requirement) and conditional questions that skip what does not apply.
3. Ask for a WhatsApp number or work email instead of trusting the profile pre-fill.
4. Rewrite the offer. ‘Free catalogue’ attracts collectors. ‘Get a quote for 500+ pieces’ attracts buyers.
5. Check placements where your setup still lets you choose them (Meta has been removing that control) and drop Audience Network first if junk concentrates there.

Why do fake leads from Google Ads happen?

Google leads go wrong in three places: loose targeting (search partners, Display Network expansion and broad match pulling in ‘jobs’ and ‘pdf’ queries), automated campaigns that chase cheap form fills, and bots or vendors hitting the landing-page form. Google filters much invalid traffic automatically, but not all of it.

The fixes are mostly housekeeping. Read the search terms report every week and add negatives such as job, salary, course, free, pdf, project report and how to make. Use phrase and exact match on product-plus-supplier terms like ‘stainless steel flange manufacturer’. In the campaign settings, switch off search partners and the Display Network option for lead campaigns. Run plain Search first and add Performance Max only once you can import qualified outcomes, because Performance Max optimises toward whatever conversion you give it, including a junk form fill. Over a few weeks the search terms report shows you the buyer’s real vocabulary, which is useful beyond ads.

How do you improve lead quality with qualifying questions?

Three questions do most of the work: how many you need, who you are, and when or at what budget. Use dropdowns so answers can be filtered. Each extra field loses some submissions, and the ones you lose are mostly the ones you did not want. Stop at four or five fields.

Pick questions that only a real buyer finds easy to answer:

• Quantity per order or per month, in bands relative to your MOQ.
• Company name or website. This alone removes most students and hobbyists.
• Application or grade, or ‘Do you have a drawing?’ with yes, no or not yet. Drawings are a strong signal in machined and fabricated parts.
• Timeline: this month, this quarter, just exploring.
• Budget band, where your offer has a real floor.

Most ‘fake leads’ are real people we made it too easy to be. The fix is not another traffic filter. It is one honest question the wrong buyer will not answer.— Murtaza Udaypurwala, DESENO

How do you stop spam leads and bots on website forms?

Add friction that bots cannot pass and real buyers barely notice: a hidden honeypot field, reCAPTCHA or Cloudflare Turnstile, a minimum time-to-submit, rate limits per IP, and email checks that reject throwaway domains. For serious enquiries, send paid traffic to a landing page instead of an instant form.

Phone verification catches fake numbers at the source. SMS OTP in India needs a DLT-registered sender and template, so many teams verify by WhatsApp or email instead. Verify before the lead reaches sales, not after. A landing page also lets you state the MOQ, price floor, certifications and proof, so buyers who do not fit leave before they submit. Our guide to landing pages that convert covers the page itself.

On our own enquiry forms we ask a budget band. If the answer falls under what we can serve, the form closes with a polite message instead of passing a lead to sales. We lose submissions and gain hours. If your product has a real MOQ or price floor, say it, ask it, and let the under-floor buyer leave kindly.

How do you teach Facebook and Google ads what a good lead is?

Send outcomes back. On Meta that is the Conversions API for CRM integration with the Conversion leads performance goal. On Google it is offline conversion import, now upgraded to enhanced conversions for leads. Both tell the platform which leads your team qualified or won, so it stops optimising for raw form fills.

Meta’s Conversions API for CRM integration guide says the goal works with instant form leads, expects the Meta lead ID stored in your CRM, recommends at least 200 leads a month and daily uploads, and wants events within 28 days of the lead. On Google, capture the gclid in a hidden form field and store it in the CRM, or pass hashed email or phone through the tag, then upload qualified and won outcomes. Google’s enhanced conversions for leads help page covers setup, and Google has been moving these uploads onto its Data Manager API, so ask your developer or CRM vendor which route they use.

Most manufacturers will not reach 200 form leads a month. Then optimise for an event you can count, a qualified RFQ rather than a closed order, which is too rare for any algorithm to learn from. Our guides to qualified leads, MQL and SQL and lead tracking and attribution cover the definition and the hidden fields. Then watch cost per qualified lead: cost per lead will rise while cost per qualified lead falls.

The bottom line

Fake leads are a symptom of easy forms and blind algorithms. Sort your leads into fake, spam, junk and wrong-fit. Add friction in the right place: Higher intent forms, qualifying questions, a landing page for serious enquiries, bot protection on your own forms. Then close the loop by sending qualified outcomes back to Meta and Google. DESENO’s performance marketing team can run this audit on your account.