On August 1, 2026, a directly-contacted group of Amazon advertisers lost the ability to put ad spend on a credit card as their primary payment method. Ad costs now come out of seller or vendor account balance by default, or Pay by Invoice if you chose it. If you were in that group, you did not just change a setting — you handed back your working capital float and your card rewards. It is September. This post is about what to check now, not what to do before a deadline that has already gone.
Table of Contents
First, the correction most coverage got wrong
For about three weeks in April, half the Amazon ecosystem reported this as “Amazon bans credit cards for ads.” It never was that.
Amazon’s own announcement is narrow and specific: the change applies to a small group of advertisers who were contacted directly, by letter, in their Seller Central accounts. The overwhelming majority of Amazon advertisers were already paying via account balance deduction and noticed nothing at all. If no letter arrived in your account, nothing happened to you.
That distinction matters, and not just for accuracy. It tells you what this actually was: Amazon retiring an exception, not launching a policy. A pocket of advertisers had been grandfathered onto card-primary billing, and Amazon closed the pocket.
But if you were in that pocket, the “small group” framing is cold comfort. The money is real, it is ongoing, and almost nothing written in April told you how to measure it after the fact.
What actually changed
Affected advertisers moved from credit card as the primary payment method to one of two options.
| Option | How it works | Timing | Best suited to |
|---|---|---|---|
| Account balance deduction (the default) |
Ad costs are netted against your available seller or vendor account balance. Debits and credits are handled inside Amazon automatically. | Continuous. Ad spend draws against proceeds as they accrue. | Sellers with steady daily sales comfortably exceeding daily ad spend. |
| Pay by Invoice | Amazon issues an invoice at the end of each month. You pay it separately. | Net 30 from issuance, so roughly 30–60 days from the spend itself. | Anyone who needs a payment window, a cleaner paper trail, or separation between ad cost and marketplace proceeds. |
The three things this actually costs you
The setting change takes ten seconds. The consequences do not.
1. You gave back the float
A credit card is not just a payment method, it is an interest-free loan. Spend on day one of a statement cycle, and you might not pay for it for another 50 or 55 days. Across a full month of ad spend, the average sits somewhere around 45 days of free working capital, permanently revolving.
Balance deduction collapses that to roughly zero. Your ad spend is now settled against proceeds as they accrue. Pay by Invoice gives some of it back — net 30 from month-end is genuinely useful — but it is still shorter than a well-timed card cycle.
That released working capital was funding something. Inventory buys, most likely. If you have not consciously replaced it, you have quietly tightened your own reorder cycle without deciding to.
2. You gave back the rewards
At 2% cash back, a seller spending $50,000 a month on ads was earning $12,000 a year for payments they were making anyway. That is not a rounding error, it is a headcount.
Worth understanding why Amazon wanted this: card processing costs Amazon roughly 3% in merchant fees on every dollar of ad spend routed through a card. On a large advertiser, Amazon was paying meaningfully more to collect the money than you were earning in rewards. Netting against proceeds costs Amazon nothing. This change was never really about you.
3. Your ad spend is now coupled to your daily sales
This is the one that will actually hurt someone this Q4, and it gets the least attention.
Under card billing, ad spend and sales velocity were independent. You could spend aggressively through a slow patch and settle up later. Under balance deduction they are mechanically linked: a soft sales week is now also a week where your balance is thinner against the same ad bill.
That coupling is dangerous in exactly the moment you least want fragility — a launch, a rank push, a stockout recovery, or the run-up to peak. These are all situations where you deliberately spend ahead of revenue. If you are planning that kind of push, read our notes on the mistakes sellers make during peak season before you set budgets, and build the forecast properly using a spend and revenue model built before you spend a dollar.
| Monthly ad spend | Working capital you lose | Card rewards lost per year | Click credits offered |
|---|---|---|---|
| $10,000 | $15,000 | $2,400 | $12,500 |
| $25,000 | $37,500 | $6,000 | $12,500 |
| $50,000 | $75,000 | $12,000 | $12,500 |
| $100,000 | $150,000 | $24,000 | $12,500 |
| $250,000 | $375,000 | $60,000 | $12,500 |
Look at the last two columns together. The credit fully covers a smaller advertiser’s annual rewards loss several times over. At $250,000 a month it covers about ten weeks of it, once, and never again. The compensation is flat while the cost scales — which tells you precisely who Amazon was worried about losing, and it was not the big accounts.
About those click credits
Yellow callout box (#FDE8AF, yellow left border):
Attribution matters here. The $2,500 per month for five consecutive months, August through December, totalling $12,500, comes from reporting on the letters sent to affected advertisers. Amazon’s public announcement does not mention click credits, incentives, or compensation of any kind. If you were in the affected group, trust your own letter and your own billing console over any blog post, this one included.
Assuming the credits are real and widely distributed, there is a second-order effect worth thinking about now rather than in December.
Credits are not cash. They can only be spent inside the auction. When a cohort of advertisers is holding budget that has no value unless it is spent on clicks, the rational move for each of them individually is to bid more aggressively than they otherwise would. PPC analyst Alexander Swade put the consequence bluntly, predicting “a massive spike in CPCs” as a result.
Whether that materialised at scale depends on how large the affected cohort really was, and “small group” cuts against it. But the window runs to the end of December, which overlaps exactly with Q4. If your CPCs ran hot in Q4 and your usual explanations do not cover it, subsidised competition is at least a hypothesis worth holding — and a reminder that a CPC rise is not automatically a signal about your own account. Our guide to reading suspicious PPC metrics covers how to tell an account problem from a market problem.
Balance deduction or Pay by Invoice?
If you were migrated by default and never actively chose, this is still worth revisiting. The setting is not frozen.
Stay on account balance deduction if your daily sales comfortably exceed your daily ad spend, your TACoS is stable, and you value the operational simplicity of never touching a bill. For most healthy, steady-state sellers this is genuinely the lower-friction option and the noise around it was overblown.
Move to Pay by Invoice if any of the following are true: your sales are seasonal or spiky, you are planning a launch or an aggressive rank push, your margins are thin enough that a 30-day window materially changes your inventory cycle, or your finance function needs ad cost separated from marketplace proceeds for clean reporting.
That last one is underrated. When ad spend is netted invisibly against proceeds, it stops feeling like a bill and starts feeling like weather. Costs you cannot see are costs you stop managing. If you have never separated true product profitability from headline revenue, this is the moment — start with why high revenue doesn’t mean high contribution margin, then pair it with TACoS as the metric that actually reflects total ad load.
Your September audit
The deadline is gone. What is left is verification. Work through these in order.
- Confirm which method you are actually on. Ads Console → Billing. Do not assume the migration did what you expected. Verify the primary method and confirm which card is sitting in the backup slot.
- Check whether that backup card is still valid. This is the quiet failure mode. A balance shortfall falls through to a card that expired in July, and delivery stops mid-Q4. Check the expiry date now.
- Pull your August invoice or balance statement and reconcile it against your reported ad spend. First cycles under a new billing rail are exactly when discrepancies surface.
- Look for a click credit line item. If you were promised credits, confirm they are actually landing and being applied, month by month. Do not assume.
- Recalculate your cash conversion cycle without the float. If ad spend was silently financing inventory, find out by how much before your next reorder, not after.
- Rebuild your true cost per unit including the lost rewards. If you were banking 2% on ad spend and treating it as margin, that margin is gone and your break-even shifted.
- Set a balance floor and a monitoring habit. Decide the balance level below which you throttle spend, and check it weekly rather than discovering it through a delivery interruption.
- Cut the waste that no longer has a float to hide behind. Run a bulk file audit for wasted ad spend and work through the 9-check PPC audit. Inefficiency you could previously carry for 45 days interest-free now hits your balance the same week.
That last point is the real operational shift. Card billing was a shock absorber. Every dollar of wasted spend was a dollar you did not have to fund for a month and a half. Without it, waste is immediate, and the case for fixing it got materially stronger. If your account has structural problems you have been tolerating, our writeups on common optimization mistakes and diagnosing account issues are the place to start. And if you want to offset some of what you lost, there are legitimate ways to get free ad spend out of Amazon that most sellers never claim.
What we still don’t know
Being straight about the gaps, because several of these matter more than the settled facts:
- How large “a small group” actually is. Amazon has never quantified it. Every estimate in circulation, including the CPC-inflation argument, rests on a number nobody outside Amazon has.
- Whether the click credits were uniform or scaled. A flat $2,500 per month reads like a retention gesture aimed at smaller advertisers. Whether larger accounts received proportionally more is unreported.
- Whether this expands. Amazon framed it as tidying up one group. Nothing in the announcement says card-primary billing survives indefinitely for everyone else. The economics that motivated this change apply just as well to the rest of the base.
- What Pay by Invoice actually requires. No published detail on credit checks, spend thresholds, or whether approval can be withdrawn.
- The exact behaviour when balance and backup card both fail. Whether delivery throttles, pauses, or stops outright — and how fast — is not documented anywhere we can find.
What this actually signals
Step back from the billing mechanics and this fits a pattern that has run all year.
In June, an off-Amazon ad serving column appeared in bulksheets with no announcement. In August, Sponsored Products started running inside creator reviews with six days’ notice, on by default. Reporting surfaces were consolidated through the unified reporting rollout. And here, a payment rail changed via individually-mailed letters and a single support-page update.
The common thread is not malice, it is direction. Amazon is steadily converting advertiser-held optionality into Amazon-held defaults, and announcing it through the narrowest channel that technically counts as disclosure. Float was optionality. Placement control was optionality. Both went the same way, in the same year, with the same “no action required” framing.
The practical response is not outrage, it is instrumentation. Know which defaults you are sitting on, check them on a schedule rather than when something breaks, and treat “no action required” as the phrase that most reliably means action is required.
Summary
One thing to do today: open Ads Console → Billing and confirm three things — your primary payment method, the expiry date on your backup card, and whether any promised click credits are actually posting. Ten minutes. It is the cheapest insurance available on the run-up to Q4.
