--- title: "The Ad Campaign That Empties the Bank Account Is Usually the Good One" url: "https://cfodrive.com/insight/the-ad-campaign-that-empties-the-bank-account-is-usually-the-good-one/" author: "Dr. Igor Ivitskiy PhD" published: "2026-09-25" updated: "2026-09-25" --- # The Ad Campaign That Empties the Bank Account Is Usually the Good One I spent two thousand dollars on my first advertising campaign, and I took the money from a credit card. My salary at that time was about four hundred dollars a month, so I had spent five of my salaries at once. After 35 days the card payment was due, the last day to pay without interest. I had orders. The affiliate program had registered them. But all of them were on hold, and I had not received a single dollar. The campaign worked. What failed was the calendar. My card gave me 35 days. The payouts needed more. I have run more than 2,000 audits of advertising accounts since 2006. I still see this calendar problem, now in companies with finance teams much more careful than I was. It almost never appears in the marketing report. It appears in the bank account. ### Ad money moves in one order Ad money moves in one order only. The platform is paid first. You are paid last. A US advertiser on Google Ads is [charged automatically after the ads run](https://support.google.com/google-ads/answer/2375432), on the first day of the month or earlier, when spend reaches a payment threshold. Paying later on a monthly invoice is possible only for businesses that spend at least $5,000 a month in three of the last twelve months. Google has other conditions too. The money from customers takes longer. How much longer depends on the business: the sales cycle, the payment terms you give customers, refund windows, marketplace payouts. Amazon, for example, [settles seller accounts every two weeks](https://sell.amazon.com/blog/amazon-seller-payments) and typically holds funds for seven days after delivery. Many companies wait even longer. Their first order only pays for the advertising, and the profit comes with the repeat orders. Now look at the cash a typical company holds. The JPMorganChase Institute tracks 2.1 million US small businesses. In its [2026 small business snapshots](https://www.jpmorganchase.com/institute/all-topics/business-growth-and-entrepreneurship/denver-small-business-snapshot), the typical firm in that nationwide sample held 17.6 cash buffer days in 2025. That cushion has to cover everything, payroll and suppliers included. The advertising gap has to fit inside it together with all the rest. ### The ad report and the bank use different calendars The marketing report does not show this order. In Google Ads, [the main conversion columns are calculated by the time of the click](https://support.google.com/google-ads/answer/6270625), not the time of the conversion. The revenue in last month's return on ad spend is dated by the click, not by the day the money reached the bank. And conversion value is the order value the company sends to the platform, not the money received after refunds and payment terms. So when a finance director reads that ROAS held at 400 percent last month, two things are true at the same time. The revenue in that number is real. A part of it is still somewhere between the customer and your bank account. Return on ad spend has no time unit. A ROAS of 4 that comes back in ten days and a ROAS of 4 that comes back in ten months are the same number in the report and two different businesses in the bank. ### Why the good campaign is the dangerous one The cash gap is easy to estimate from above. Multiply the daily ad spend by the number of days until that spend comes back to the bank as gross profit. I call the second number cash payback days. A company that starts spending $2,000 a day with a 45-day payback can be up to $90,000 short before the money returns. If the money comes back gradually, the real gap is smaller, and your own receipts will show by how much. The upper number is the one to fund. A bad campaign does not stay long. Someone sees the numbers and cuts it. The good campaign is the one that gets more money, and every increase opens a new gap. Say the budget goes from $2,000 to $3,000 a day. In most accounts I see, the extra money buys more expensive clicks and buyers who are less ready. Its payback is longer, for example 60 days instead of 45. The extra $1,000 a day can put the company up to $60,000 short again. The first $2,000 a day needed 45 days of its own spend in cash. The next $1,000 needs 60. Every new dollar needs more cash than the dollars before it. This is why the most dangerous moment is a good start. I see the same mistake again and again. The first two weeks bring many leads. The owner decides it will always be like this and raises the budget sharply. After that, advertising becomes more expensive with no result, and the cash is already committed. The report looked perfect on the day of the decision. For an established campaign I keep each budget step at about 18 percent and never above 20. I do it not more often than every three or four days, and only when the current budget is fully spent. That rule protects the campaign. It does not protect the cash, because in three or four days nobody can see a 45-day payback. The cash rule is different. Add up all the increases you plan before the first one pays back. That sum is the cash you commit before you have any proof, and it has to fit in the bank. ### The number I ask for before ROAS In the audits I run with Profit Forensics, the first thing I want to know is how long the money is out. Most companies do not have this number. They have ROAS and lifetime value. Lifetime value without time has the same problem. Take a customer who brings $12,000 of gross profit over their lifetime, and an acquisition cost of $6,000. If that value arrives in three months, the customer brings about $4,000 a month. The $6,000 comes back in six weeks, and you start earning from the middle of the second month. If the same value arrives over ten years, the question is different. Are you comfortable spending five years only getting the acquisition cost back, and making profit from the sixth? I think the answer is no. Some companies do this on purpose. They pay $300 in advertising for a customer whose first purchase brings $100 of profit before advertising costs. They stay loss-making for one, two or three years and wait for lifetime value to pay it back. That can be a valid strategy. But it is a financing decision. How much to spend on advertising is a question about the business model. It is a decision for the founder and the finance director, and marketing should not make it alone. When cash is the constraint, the ranking of channels changes. Say channel A returns $1.30 of gross profit for every dollar within two weeks, and channel B returns $2.00 but needs four months. In the ad report, B wins. In the bank, over the same four months, a dollar can go through A about eight times. That brings roughly $2.40 of profit, if there is enough demand to spend it again. Through B the dollar goes once and brings $1.00. ### When this is not your problem Billing terms decide who funds customer acquisition. Some companies have the order reversed. Take a large advertiser approved for a monthly Google invoice, whose customers pay by card at checkout. It gets the customer's money within a few business days and pays for the clicks weeks later. Amazon sellers who pay for [Amazon ads from their account balance](https://advertising.amazon.com/library/news/update-on-advertiser-payments), which Amazon says most of its advertisers do, pay after the customer has already paid. Annual prepaid subscriptions can work the same way. For these businesses a bigger budget does not open a cash gap. It can still lose money, but that is a different problem. Two dates give you the answer: the date your ad platform takes the money and the date the money from those clicks arrives. If the ad bill is paid later, your constraint is somewhere else. If it is paid earlier, the rest of this article is about you. ### What I would put in place on Monday In my own business, advertising is planned as working capital. The amount for each month sits in the budget in advance, including a fixed sum for testing a new channel. It is not money taken from somewhere else when a campaign looks promising. It is decided before the month starts, together with how long it may stay out. For a finance team this is five steps. - Measure cash payback days for each channel, from the ad charge to gross profit received in the bank. Use the platform's conversion-time columns and your own receipts, not the default click-date view. - Before approving a budget increase, multiply the extra daily spend by its payback days, and assume a longer payback for the new money. That is the most cash the increase can take. If the company cannot fund that amount, the increase is too big, whatever the ROAS. - Fund the largest cash gap before you raise the budget, not after. The next big increase waits until the previous one has shown its payback in the bank. - Add an expected-cash line for advertising next to the revenue line in the forecast. - If you pay for ads by card, compare the card's grace period with your payback days. When the payback is longer, the card is financing your customers. My card gave me 35 days. The affiliate program needed more. Every company that advertises has the same two numbers, and the finance team should know both before the budget grows. --- Igor Ivitskiy is a PhD in mathematical modeling and the founder and chief scientist of [Doctor Ads](https://ivitskiy.com), a UK practice focused on the economics of advertising. He has applied mathematical models to advertising budgets since 2006 and has run more than 2,000 audits of advertising accounts. He developed Profit Forensics, a method for finding where advertising spend loses money for the business that pays for it. He was ranked #6 in The PPC Survey's Top 50 Most Influential PPC Experts of 2026.