Stop optimising Google Ads to cost per lead

Google’s automated bidding learns from the conversions you report. If ‘form submitted’ is the only conversion, the system has no way to tell a ₹50 lakh project enquiry from a fake number. It optimises the number you gave it, and the cheapest way to get more form fills is usually lower-intent traffic.

This shows up in a familiar pattern: cost per lead falls month after month, the agency report looks healthy, and the sales team says lead quality has collapsed. Both are telling the truth. The campaign is succeeding at the wrong goal.

Broad match, Performance Max and Display placements make this worse, because they give the algorithm more room to find cheap conversions. They can work well — but only when the conversion signal reflects real value to your business.

The classic mechanism is simple. When someone clicks an ad, Google adds a click identifier (the GCLID) to the landing page URL. Your form captures it in a hidden field and stores it with the lead in your CRM. When that lead later becomes qualified or closes, you upload the GCLID with the conversion name, time and value.

Enhanced conversions for leads does the same job using hashed first-party data, such as the email address or phone number the lead submitted, instead of relying only on the click ID. It holds up better when click IDs go missing, and Google now steers lead advertisers toward it. Check the current Google Ads documentation for the method that suits your setup.

Uploads can be manual (a CSV on a schedule), through Google Sheets, or automatic through a CRM integration. HubSpot, Salesforce and Zoho all offer connections, but a disciplined weekly manual upload is enough to start.

• Capture the click ID in a hidden form field
• Store it on the lead record in your CRM
• Define stages: qualified lead, opportunity, closed won
• Upload outcomes with a time and value, at least weekly

Bidding algorithms need volume and speed. A conversion that happens a handful of times a month, months after the click, gives the system very little to learn from. A qualified-lead stage — sales has spoken to the person and confirmed need, budget and fit — usually happens within days and often enough to be useful.

Set the qualified stage as the primary conversion for bidding, and keep raw form fills as a secondary conversion so you can still see them. Add closed sales as a secondary or value-based conversion if you can assign real values. Google’s guidance on conversion volumes for each bid strategy changes, so check the current recommendation before switching.

If you sell services with very different values, give each qualified lead a value — a rough expected deal size is enough — and move toward value-based bidding. That tells the system a ₹5 lakh enquiry matters more than a ₹20,000 one.

Deeper metrics change the conversation. A campaign with a high cost per lead and a strong qualification rate may be your best performer; a campaign with a low cost per lead and almost no qualified leads is quietly wasting budget. You cannot see either without joining ad data to CRM data.

The join can be as simple as a spreadsheet: campaign, spend, leads, qualified leads, opportunities, wins, revenue. Update it monthly. Once the numbers exist, decisions about pausing keywords, shifting budget and judging agencies get much easier.

A form that only asks for name and phone collects anyone. Adding a question such as budget range, project timeline or company size costs some volume but gives you a quality signal and lets you route leads properly. Dropdowns are quicker to answer than free text.

Ad copy filters too. ‘Packages from ₹25,000’ or ‘For companies with 50+ staff’ puts off people who were never a fit, so you stop paying for their clicks. Discouraging clicks feels counter-intuitive, but you pay for every one.

Turn on spam protection — reCAPTCHA or a honeypot field — and validate phone numbers. Fake submissions do not just waste sales time; if they count as conversions, they actively train bidding toward more of the same.

Start by importing offline conversions as secondary, so they are visible but not used for bidding. After a few weeks you will see which campaigns and keywords produce qualified leads. That alone often justifies budget changes before any bidding switch.

When you do change the primary conversion, do it on one campaign with stable volume, keep the budget steady, and avoid other big changes in the same period. Compare cost per qualified lead before and after — not cost per lead, which will probably rise, and that is fine.