Most of the Google Ads budgets I am asked to review were set by feel. The owner picked a number that sounded survivable, divided it by thirty, and typed it into the daily budget field. Then they judged the campaign against a cost per lead they had never actually calculated. That is the real problem in a lot of underperforming accounts. Nobody knows what a customer is worth, so nobody can say whether the account is winning or losing. The bidding is not the issue. The absence of a target is.
An advertising budget is not a marketing question. It is an arithmetic question that happens to be answered by the marketing department. The auction does not care what you can afford. It sets a price, and either that price sits inside your margin or it does not. Your job is to know where the line is before you start bidding, not after three months of spend.
This is why I ask for gross margin before I ask for anything else. An account selling a service with 70 percent margin and an account selling hardware with 12 percent margin are playing completely different games in the same auction. The same cost per click is cheap for one and fatal for the other.
You can do this on paper in twenty minutes. You need four things, and you need them from your own books, not from an industry benchmark.
If you cannot produce these, that is your first project, not the ad account. I have seen accounts paused for months over a cost per lead that turned out to be perfectly profitable once someone finally worked out the close rate.
Work through this in order with your own numbers.
Notice that the budget falls out at the end. Almost everyone starts at step five and never does steps one to four.
The simple version is right most of the time. Here is when it is not.
Repeat purchase businesses. If a customer reliably buys again, first-order economics will tell you to stop advertising when you should be pushing harder. Use lifetime gross profit, but discount it. A customer who might spend again in eighteen months is not worth the same as one who spends today, especially if your cash is tight.
Capacity limits. If you can only take twelve jobs a month, the correct question is not how cheaply you can buy a lead. It is how expensive a lead you can tolerate to fill the twelfth slot. Businesses with fixed capacity should usually pay more per lead than they think and be much fussier about which leads they buy.
Referral loops. In trades and professional services, a happy customer often brings another one for free. If you can measure that, it belongs in the number. If you cannot measure it, do not invent a multiplier to make the ads look better. Leave it out and treat it as headroom.
Seasonality. Some markets have four good weeks a year. Averaging annual margin across twelve months will make you underbid in the weeks that matter and overbid in the dead months.
Pull your last hundred inquiries and count how many became customers. Pull the invoices and work out the true gross margin, including the hours you do not usually charge for. Then run the five steps above and write the resulting cost per lead on a sticky note next to your monitor. That single number turns every future ads conversation from an argument about opinions into a comparison against a target.
If the number comes out lower than what the auction currently charges, you do not have an ads problem. You have a pricing, conversion, or offer problem, and no amount of bid tuning will fix it. If you want a second pair of eyes on the arithmetic before you commit budget, you can book a free consultation.