How the trade-off works
Every bidding target sets a price you are willing to pay for a result. Tighten it and the system becomes choosier: it enters fewer auctions, bids on fewer queries and reaches fewer people, keeping only the cheapest opportunities. Costs per result fall and total results fall with them. Loosen it and the opposite happens — more auctions are entered, more marginal demand is bought, volume climbs and the cost of each result rises.
The trade-off exists because demand is finite and unevenly priced. The easiest conversions are available cheaply and there are only so many of them; everything beyond that costs progressively more. A bid target is really a decision about how far down that queue you are prepared to go.
The same logic applies whether the control is a target cost per acquisition, a target return on ad spend, a bid cap or a cost cap, and whether the platform is search or social.
Why the trade-off matters
Most disagreements about ad performance are really this trade-off in disguise. A finance view wants a lower cost per lead; a sales view wants more leads; both are achievable, but not simultaneously, and no amount of optimisation removes the tension. Naming it turns an argument into a decision.
It also explains a common surprise. Tightening a target to improve reported efficiency often reduces total profit, because profit comes from the number of sales multiplied by the margin on each, not from the cost per lead alone. A very efficient campaign that produces few leads can be worth less than a looser one that keeps the sales team busy.
Where the trade-off goes wrong
The first mistake is moving the target sharply. Large changes throw automated bidding back into a learning period, and the results during that period describe the disruption rather than the new setting. Small, spaced adjustments give readable answers.
The second is setting the target from a wish rather than from the business. A target far below what the auction will bear does not produce cheap leads; it produces almost nothing, because the system simply stops bidding. The lever has a floor set by competition, not by your spreadsheet.
The third is ignoring capacity. Loosening a target to increase volume is only useful if someone can answer the enquiries. Leads that go unanswered cost the same as leads that convert, and in a small business this constraint binds sooner than the budget does.
What to do about it
Decide which side you are actually optimising for this quarter, and say so plainly before touching a setting. If capacity is full, tighten and take the cheaper results. If capacity is spare and the margin allows it, loosen and accept a higher cost per result in exchange for more of them.
Then move in small steps, wait for the learning to settle, and read the outcome as total profit rather than as cost per lead. The point where a further loosening stops paying is the same point marginal ROAS identifies, so the two ideas are best used together. Set the starting target from your own margin and close rate rather than from a benchmark, and revisit it when either changes. If the campaign uses Smart Bidding, remember that the target is the main lever you still hold.
