How is a target cost initially decided?

Usually based on recent historical average cost for that event, adjusted to what the business can profitably sustain, rather than picked arbitrarily.

What happens if the target is set too aggressively?

Volume typically drops sharply, since the algorithm becomes highly selective, bidding only on auctions it predicts can hit that tight a cost.

Can the target be changed after a campaign is running?

Yes, though frequent changes can disrupt the algorithm's learning process, gradual adjustments over time tend to work better than large, sudden shifts.

Does this approach need much historical data to work well?

Yes, generally, enough recent volume of the chosen optimization event is needed for the platform to predict conversion likelihood reliably before setting a tight target.

Is target cost per optimization event available on every platform?

The specific terminology varies, but the underlying goal-based automated bidding concept is available in some form across most major ad platforms.

What Is the Target Cost Per Optimization Event?

Target cost per optimization event is a bidding approach where an advertiser sets the average amount they're willing to pay for each occurrence of their chosen optimization event, purchase, lead, install, and the platform's algorithm bids across auctions to hit that average, rather than the advertiser setting a fixed bid for every individual auction manually.

How This Bidding Approach Works

Once a target cost is set, the platform's predictive bidding engine evaluates each auction's likelihood of producing the chosen optimization event and bids accordingly, higher for auctions it predicts are likely to convert cheaply enough to support the target, lower for ones it isn't confident about, working to land on the target average across the full campaign rather than enforcing it auction by auction.

This is functionally the same mechanism as Target CPA, expressed using "optimization event" terminology specifically because the target can be set against any configured event, not only a narrowly defined acquisition, giving it a slightly broader, more flexible framing across platforms that use this exact phrase.

Putting a Target Cost to Work

An app company sets a target cost of ₹80 per install, and the algorithm bids more aggressively on users predicted to install at or below that cost while pulling back on unlikely prospects.
A lead-gen business sets a target cost of ₹350 per form submission, adjusting it downward gradually each month as conversion data accumulates and confidence in a tighter target grows.
An ecommerce brand initially sets an aggressive target cost per purchase below its historical average, and sees volume drop sharply as the system struggles to find enough auctions it can win at that tight a cost.

Working Out the Number

Target Cost Per Optimization Event = Total Spend ÷ Total Optimization Events

A campaign that spends ₹40,000 and generates 200 leads, with leads as the chosen event, is averaging ₹200 per lead, the number a target would be benchmarked against.

Target Cost vs Its Close Relatives

Target cost per optimization event is essentially the same bidding logic as Target CPA and Target ROAS, all are goal-based automated bidding strategies, the difference is largely terminology, "optimization event" language emphasizes that the target can apply to any configured event, not exclusively a narrowly defined acquisition or revenue outcome.

Reading Performance Against the Target

Performance against a target cost should be read as a rolling average over a meaningful window, not judged day to day, since the system is optimizing toward that average across the whole campaign period, individual days will naturally land above or below the target even when the overall trend is on track.

Why Set One at All

Setting a target cost shifts the campaign from a volume-maximizing or manually bid-managed approach to one explicitly anchored to what a business can actually afford to pay per outcome, making budget planning more predictable and giving the bidding algorithm a clear, concrete goal to optimize toward rather than an open-ended one.

Frequently Asked Questions

How is a target cost initially decided?: Usually based on recent historical average cost for that event, adjusted to what the business can profitably sustain, rather than picked arbitrarily.
What happens if the target is set too aggressively?: Volume typically drops sharply, since the algorithm becomes highly selective, bidding only on auctions it predicts can hit that tight a cost.
Can the target be changed after a campaign is running?: Yes, though frequent changes can disrupt the algorithm's learning process, gradual adjustments over time tend to work better than large, sudden shifts.
Does this approach need much historical data to work well?: Yes, generally, enough recent volume of the chosen optimization event is needed for the platform to predict conversion likelihood reliably before setting a tight target.
Is target cost per optimization event available on every platform?: The specific terminology varies, but the underlying goal-based automated bidding concept is available in some form across most major ad platforms.

See NYX in action.

Book a demo