We recently helped a client solve a long-standing issue with lead quality and ad optimisation. Although lead volume increased, many of those leads had lower revenue potential. As a result, Google Ads began optimising toward the wrong kind of lead.
We tested several solutions. But the breakthrough came in the form of a Zapier AI agent !
Optimisation for conversion value in Google Ads
Usually, when you are doing Lead Generation campaigns via Google Ads, the goal is to have as many leads as possible while reducing the cost of each conversion.
When you have only one product, the difference between one or another conversion does not matter. But if a certain product you sell has a much higher profit margin or generates significantly more revenue, you probably want to optimise for that one.
One way to do it is with offline conversions tracking.
Offline conversions? What is this?
In short, you track each conversion online (when the user submits the form), assign a value to it (manually, offline), and send it back to Google Ads.
With this data, Google Ads can now distinguish between a CHF 100 lead and a CHF 10’000 one.
This can then be used to optimise campaigns toward leads that bring the most value. But assigning value to form submissions isn’t always easy.
How to automatically assign value to leads ?
Depending on the client and their setup, it might not be as straightforward as it sounds.
In the case of e-commerce, the value of each conversion is clear: it’s the amount the customer paid. But for B2B, that value often varies widely and can only be defined precisely after a full project assessment.
But in this case, we didn’t need the exact value. We just needed to assign an estimated value to incoming leads.
Are we looking at a CHF 1000 lead, or one worth CHF 40’000 ?
What we tested
We tested various approaches:
- Different lead forms, per product range on the website => Too complicated & annoying to maintain
- Hidden fields with estimated values => this also required different forms
- Manual updates in Google Sheets => Too complex. Too manual. The client would always forget.
The solution – tadaaa – Artificial Intelligence!
Now enters the “Zapier AI agent”. Think of it like a tiny AI assistant living inside your automations.
(Zapier is super useful, for tiny automation)
What we did:
- Send the information from the forms on the website to Zapier
- Let the AI agent read the submitted form content
- Then estimate the value based on a pre-trained product/price list
- Send it to Google Ads as an offline conversion
While the results aren’t perfect, they’re accurate enough to make a meaningful difference.
And no manual work needed. The technical aspect is very easy to maintain.
The actual Zapier workflow involved more steps, but we’ve simplified the diagram below for clarity.
This is the “prompt” for the AI Agent. Very simple. The blurred out part is just a list of projects with their estimated price.
Other use cases
This system could be used for other cases.
For example:
- Estimate the chances of a deal closing based on certain criteria, then use the score to modify the conversion value
- Automatically assign a value of 0 to unqualified leads
- Estimate lifetime value (LTV) of a new customer.
Each of these use cases adapts the conversion value to teach the algorithm how to optimise for the highest-value leads.
That’s it for today
Curious how Zapier AI agents can enhance your ad performance? Let’s talk.