80% of RFQ submissions were competitors and students — Google bid on all of them

The situation

A German industrial equipment manufacturer was running Google Ads targeting procurement managers and plant engineers. Their RFQ (Request for Quote) form was getting 85 submissions per month and Google reported a strong CPA of "280. The sales team said fewer than 5 per month were worth pursuing.

The problem

The RFQ form was indexed publicly and receiving submissions from: competitors researching pricing, engineering students doing coursework, job applicants using the form as a contact, and procurement managers with no budget authority. Google was optimizing to find more RFQ form submitters — and getting very efficient at finding non-buyers who happened to use the contact form.

What we did

1. Implemented a lead qualification webhook: RFQ submissions trigger a Zapier workflow that checks company size (via Clearbit), marks qualified/unqualified in CRM
2. Built offline conversion import to Google Ads: only qualified leads (company size 50+, industry match, not a competitor domain) count as conversions
3. Added form questions: "Annual equipment budget" and "Decision timeline" — immediate disqualifiers filter spam before submission reaches Google
4. Suppressed student and competitor email domains from Smart Bidding signal via audience exclusions

Why it worked

B2B form conversion quality is invisible without manual qualification. A contact form that gets 85 submissions sounds like a success story until you look at who is submitting. Google Smart Bidding's strength — finding efficient conversion pathways — becomes a liability when the conversion signal includes noise. Offline conversion import of CRM-qualified leads is the only way to train Google on signal that maps to actual revenue.