Financing ad spend, and why lenders treat it differently from inventory

Borrow 40,000 for inventory and the lender has something to look at. If the business stops paying, the goods exist, they have a wholesale value, and a liquidator can convert them. Borrow 40,000 for advertising and there is nothing behind the loan at all. The money is consumed on contact.

That difference drives the terms, the underwriting, and the structures on offer — and it should also drive how you decide whether to do it.

What the lender sees

The consequences are consistent:

• Facilities for marketing are cash-flow underwritten, which means the whole business stands behind them.
• The structures that reach this use most readily are the general-purpose ones — a line of credit, a short-term loan, revenue-based products, a business card — rather than anything asset-secured.
• Where a funder is comfortable, it is usually because they can see the revenue that results. Platform-integrated and processor-linked funding sits here, and the price of that comfort is that the funder sees your sales data and often sits in the payment flow. See funding offered by the platform that processes your sales.

None of this makes financing advertising wrong. It makes it a decision that must be justified by the payback arithmetic alone, because there is no collateral to fall back on.

The payback arithmetic

At a cost per acquired customer of 165, that buys 218 customers. First-order contribution is 74 a customer, so 16,145. If 34 per cent buy once more within twelve months, add 5,489. Total twelve-month contribution: 21,635 against 36,000 of spend. 0.60 times. The campaign does not repay inside a year, financed or not.

At a cost per acquisition of 92, the same 36,000 buys 391 customers, producing 28,957 of first-order contribution and 9,845 of repeat — 38,802, or 1.08 times. It repays in roughly eleven months.

Now add the financing. At an illustrative total cost of 1.22 times the amount advanced, the 36,000 costs 43,920. At the 165 acquisition cost, the effective cost per customer becomes 201. The borrowing did not just add interest — it raised the acquisition cost by 22 per cent, and acquisition cost is the variable the whole case rests on.

The rule that falls out: borrow for advertising only when the measured payback period is shorter than the term of the money. If customers repay their acquisition cost in four months and the facility runs twelve, the campaign funds its own repayments. If payback is eleven months and the facility is a six-month product with daily remittance, the repayments come from the rest of the business, which is a different transaction from the one you thought you were doing.

The measurement problem behind the terms

Lenders discount marketing partly because they know the figures are usually reported by the party selling the advertising. Platform-attributed conversions count customers who would have bought anyway. Last-click attribution credits the final touch. Multi-platform reporting double-counts the same sale.

So before borrowing, establish what you can measure independently: total new customers by month from your own records, contribution per new customer from your own margins, and repeat behaviour by cohort. If the only evidence for the campaign working is the advertising platform's own dashboard, you are borrowing against a number produced by the vendor.

One more number belongs in the decision: the share of total spend this represents. A campaign at 36,000 against a business doing 2,400,000 of revenue is a test. The same 36,000 against 400,000 of revenue is 9 per cent of turnover committed to an unproven acquisition cost, financed. The size relative to the business, not the size in absolute terms, determines whether a failed campaign is a bad quarter or a solvency event.

Structuring it so a miss is survivable

1. Spend a tranche, measure, then spend the next. Borrow the full amount if the facility requires it, but release it in stages against results. With a revolving line, draw in stages.
2. Set the kill criterion in advance. A cost per acquisition above a stated level for two consecutive weeks stops the spend. Written down beforehand, it gets acted on.
3. Keep the repayment inside existing cash flow. Test whether you could service the payment with zero incremental revenue. If not, the campaign has to work, and campaigns that have to work get extended past the point where the data says stop.
4. Prefer facilities you can repay early without penalty, so a campaign that works fast lets you clear the debt.
5. Do not stack this on top of an existing advance. Marketing is the use of funds most likely to be financed by whatever is available rather than by what fits, and that is how a second position gets added to a business that cannot carry one.

What to have ready before you ask

• Twelve months of new-customer counts from your own system, by month.
• Contribution per new customer, computed from your own margins after delivery costs.
• A cohort table showing repeat purchase rates over twelve months.
• Your current cost per acquisition by channel, measured your way.
• The payback period in months, and the facility term next to it.

If those five numbers are not available, the first thing to fund is the measurement, not the campaign.