Feed CRM outcomes back or you are bidding on form fills

If your Google Ads account optimises to form submissions, you are paying Google to produce form submissions. It is very good at that. Whether those submissions turn into revenue is a question the bidder has never been asked.

That gap is the single largest source of wasted spend in lead generation accounts, and closing it does not require new budget. It requires sending your sales outcomes back into the platform.

What the bidder is actually learning

Automated bidding optimises toward whatever you told it counts. Tell it a form fill counts and it will find the people most likely to fill in a form. Those are not the same people most likely to sign a contract, and in some categories they are close to the opposite.

The tyre-kicker who submits five enquiries on a Sunday night is a cheap conversion. The procurement manager who fills in one form on a Tuesday afternoon is expensive. Give the system a form-fill target and it will steer toward the first group, month after month, getting better at it.

Feed back which enquiries became customers and the target changes. Now it is finding people who look like buyers.

Data Manager is the path now

Google consolidated data ingestion into Google Ads Data Manager. Offline conversion imports, customer data, first-party connections all run through it. If you have an old scheduled upload or a legacy integration, plan the migration rather than waiting for it to break.

The mechanics come in two flavours.

GCLID-based. Your site captures the GCLID from the ad click into a hidden form field, that value writes into your CRM against the lead record, and when the deal changes stage you upload the GCLID with a conversion name, a timestamp and a value. This is the accurate method and it needs your CRM to hold the field.

Enhanced conversions for leads. You upload hashed customer data, usually email, instead of a click identifier. Google matches it back to the click. Slightly less precise, considerably easier, and the right choice if you cannot get the GCLID into your CRM.

Use enhanced conversions for leads if the alternative is doing nothing. A slightly lossy signal beats no signal by an enormous margin.

Send stages, not just closed deals

The mistake most teams make on their first build is uploading only closed-won deals. In a business with a 90-day sales cycle, that means the bidder gets its first useful feedback three months late, which is far too slow to steer on.

Send the intermediate stages too, with different values.

A qualified enquiry might be worth 1. A booked meeting 5. A quote issued 20. A closed deal the actual contract value. The bidder gets a signal within days rather than months, and the values encode what you actually care about.

This connects to journey-aware bidding, which lets Search campaigns on Target CPA learn from goals you have not marked as biddable, such as calls and form fills. Secondary conversion actions you set to observation only are no longer inert. Go and look at what you have sitting in that bucket, because it is feeding the system now.

The build, in order

1. 1.Capture the GCLID on every form and store it against the lead in your CRM. If you use a phone system, capture the call identifier too.
2. 2.Define your stages and assign a value to each. Use real numbers from your own sales data. Average contract value by service line is fine to start.
3. 3.Create the corresponding conversion actions in Google Ads. Mark one primary, the rest secondary.
4. 4.Build the upload through Data Manager. Daily is better than weekly. Google accepts imports with a lag, but shorter is better for learning.
5. 5.Wait. Do not judge it for six weeks. The bidder needs volume before the new signal changes behaviour.

What it costs and what it will not do

The build is a week of work for a competent team, mostly in the CRM rather than the ad account. The ongoing cost is discipline: if your sales team stops updating deal stages, your bidding signal degrades silently and nobody notices for a quarter.

It will not fix a campaign with too little conversion volume. Automated bidding needs a reasonable flow of conversions to learn from, and if your qualified enquiries run at five a month, the stage-based values are what give you enough signal to work with. That is the real reason to send intermediate stages, not just the speed.

And it will not make bad leads good. It will just stop you buying more of them, which over a year is usually worth more than any bid strategy change you will ever make.