Run Your Own Google Ads Analysis in 5 Minutes
Most owners can't tell you two numbers: what a lead costs them, and what a customer is worth. This guide gets you both, using your real Google Ads data and Claude or ChatGPT. No agency, no dashboard headache.
What you'll find out
Your cost per lead, per campaign and overall
Which campaigns are making you money
Which ones to scale and which to cut
What you need
A Google Ads account with conversion tracking switched on. If you are not tracking conversions, the report shows spend and clicks but no leads, and the math will not work. Fixing that is step one for another day.
Claude or ChatGPT.
Step 1: Export your report from Google Ads
Open Google Ads and click Campaigns in the left menu.
Set the date range, top right, to the last 90 days. Do not leave it on All time. Old data will throw the numbers off.
Check your columns. You want Cost, Clicks, Conversions, and Cost per conversion showing. If they are missing, click Columns, then Modify columns, and switch them on.
Click the Download button just above the table (not the one up by the chart) and choose .csv.
That file has everything Claude needs.
Step 2: Paste it into Claude with this prompt
Open Claude or ChatGPT and paste this, then paste your report under it:
I'm pasting my Google Ads campaign report. In plain English, no jargon, tell me:
1. For each campaign: how much I spent, how many leads (conversions) it brought in, and my cost per lead.
2. Rank the campaigns from best to worst cost per lead.
3. My overall cost per lead across everything.
4. Flag any campaign that spent money but brought in no leads.
Here is my data:
[paste your CSV here]
You'll get a clean breakdown in seconds. That is the math most owners avoid for months.
Step 3: Add the second number, what a customer is worth
Cost per lead only means something next to what a customer is worth. Run this follow-up:
A customer is worth about $[YOUR NUMBER] to me over the time they stay with me, and roughly [X]% of my leads turn into customers.
Based on that, tell me which campaigns are actually making me money and which are losing money. Then tell me which to scale and which to cut.
Not sure what a customer is worth? Ask first:
Help me work out what one customer is worth. My average sale is $[amount], a typical customer buys [how often] and stays with me about [how long]. Give me a simple lifetime value number.
Step 4: Act on it
Scale the campaigns with a low cost per lead that are clearly profitable. Raise their budget slowly, about 20 to 30% at a time.
Cut or fix the ones costing more per lead than a customer is worth.
Park anything that spent money and got zero leads until you know why.
Step 5: Do it for every platform
Run the same thing for Meta, and any other platform you advertise on. Once you can see your cost per lead side by side, it is obvious where your money should go.
One honest note: each platform reports its own results and tends to flatter itself. Use these numbers to spot the trend and compare, not as the gospel truth to the cent.
The 5-minute recap
Google Ads, Campaigns, last 90 days, Download the table as .csv.
Paste it into Claude with the prompt above.
Add what a customer is worth.
Scale what works, cut what doesn't.
Repeat for every platform.
I've spent over 15 years helping local service businesses grow, and knowing these two numbers is where the smart ones start. This is the stuff that actually works.
Comments