6 min read
In brief
- Your starting budget is arithmetic, not opinion: cost per order divided by contribution per order decides whether advertising can work before you choose a daily figure.
- There are two separate budgets. One buys enough clicks to find out whether the page converts; a much larger one buys enough orders to judge whether the campaign does.
- Ads are waste in four situations at any budget — thin margin, a page that does not convert, a delivery promise you cannot keep, and measurement you cannot trust. None of them is fixed by spending more.
The short answer, with the arithmetic
Start with three figures you already have: average order value, gross margin percentage, and the cost of a click in your category. Those three produce the only number that matters — what an order costs you to acquire, before you have spent anything.
Worked example, using invented figures to show the method rather than a benchmark: an average order of $60 at a 40% gross margin contributes $24 per order. If a click costs $0.80 and 1.5% of clicks convert, an order takes about 67 clicks and costs about $53. That is more than the $24 it contributes, so Google Ads loses $29 per order at every daily budget.
The conclusion is not to spend less. It is that no budget fixes arithmetic — and that the three levers, in order of power, are average order value, conversion rate and click cost.
The one-line version
Break-even cost per order equals average order value multiplied by gross margin. If your cost per click divided by your conversion rate is above that number, you are paying to lose money.
The three numbers, and where they come from
| Figure | Where to find it | What people get wrong |
|---|---|---|
| Average order value | Store reporting, net of discounts | Using the product price rather than the basket total |
| Gross margin | Cost of goods, including shipping to the customer and payment fees | Ignoring freight, returns and payment processing, which together often take 10–15 points |
| Cost per click | The keyword planner range, or a week of real data | Using the cheapest category average instead of your own product terms |
Two refinements matter more than they look. Returns reduce effective margin, so a category with a 20% return rate has a lower contribution per order than the same margin with 5%. And if shipping is free above a threshold, the contribution on a $49 order is materially different from a $60 one.
There are two budgets, not one
The first budget buys evidence about your landing page. The second buys evidence about your campaign. Confusing them is why merchants conclude Google Ads does not work after eleven clicks.
- A landing page read needs roughly a hundred clicks — enough for a conversion rate signal, not enough for a budget decision. At $0.80 a click that is about $80, and it answers one question: do people who arrive buy?
- A campaign read needs enough orders to distinguish a good result from luck, which realistically means tens of conversions rather than a handful. At a $53 cost per order, thirty orders is around $1,600.
- If your break-even cost per order is $24 and you cannot fund the second budget, you are not ready to judge the channel — you are ready to test the page, which is the cheaper and more useful test.
Why the first budget is the one to spend
Almost every store that fails with Google Ads fails at the landing page, not the bidding. Spending $80 to learn that is cheaper than spending $1,600 to learn it slowly.
Four situations where ads are waste at any budget
- Margin too thin for the category's click cost. If break-even cost per order is below what a click costs, the model cannot work — and this is the most common reason a dropshipping store should not run ads yet.
- A product page that does not convert. Ads amplify a conversion rate; they do not create one. A page converting at 0.6% will lose money at every budget.
- A delivery promise the operation cannot keep. Paid traffic multiplies complaints, chargebacks and policy risk, and on Google the same promise is what the Merchant Center review compares.
- Measurement you do not trust. If reported purchases do not match your order table, you cannot compute a cost per order, and every scaling decision is a guess dressed as a number.
There is a fifth, specific to Shopping: a disapproved feed. Ads cannot spend if nothing serves, which is what a wall of disapprovals looks like from the advertiser's side — zero impressions with a working campaign.
Budget by stage, with the rule for each
- 1
Validation: enough to read the page
A small daily budget held for two weeks, with one product and one landing page. The target is a hundred clicks and a conversion rate you can compare with break-even, not a profit.
- 2
Proving: enough to reach break-even on a product that converts
Increase only after the page clears a conversion rate that makes the arithmetic work at your current click cost. The benchmark is your own break-even, not an industry average.
- 3
Scaling: percentage increases, not doubling
Raise budget in steps small enough that you can see the effect — around a fifth at a time, held for several days. Doubling resets the learning and hides which change did what.
- 4
Cutting: when the numbers say stop
If a campaign has spent three times its break-even cost per order with no conversion, that is information. Pausing is cheaper than waiting for the algorithm to find a customer it was never going to find.
The only rule worth memorising
Never scale on a number you have not reconciled against your order table. Doubling budget on a doubled purchase count is how a store spends twice as much to lose the same amount.
What a bigger budget does not fix
- A click cost that is above your contribution per order. More budget buys more of the same loss.
- A single-product store with no bundle, no upsell and a $24 contribution. Average order value is the lever, and it is a store problem, not a budget one.
- A feed with disapprovals, or a Merchant Center with no shipping settings. Spend creates no impressions.
- A conversion rate built on a delivery promise you understated. The refunds arrive after the sales, which is why the loss looks like a cash-flow problem.
Where this connects to tracking
Every figure in this guide depends on knowing what an order actually cost to acquire. If the platform counts two purchases for one order, your cost per order reads at half its true value — and half your break-even is a number you will happily scale into.
That is why the honest order of operations is: reconcile the order count, then compute break-even, then decide the budget. Reversing it means making a spending decision on a number you cannot defend.
If you want the three figures for your own store before committing a budget, that is a two-hour job with access to your order data and your tracking — and it is the cheapest part of the decision.
