Should a campaign always optimize toward the deepest possible event?

Generally yes once enough volume supports it, but a very deep, rare event can sometimes lack enough data for reliable bidding, in that case starting with a slightly shallower event and moving deeper as data accumulates is a common approach.

Can a campaign have more than one optimization event?

A campaign typically optimizes toward one primary event at a time, though some platforms support secondary or value-based optimization that considers multiple actions together.

Does the optimization event affect CPC or just conversions?

It affects the whole bidding strategy, since the system adjusts bids per auction based on predicted likelihood of that specific event, which indirectly shapes CPC, audience reach, and overall delivery.

How much data is needed before switching optimization events?

Enough conversion volume on the new, typically deeper event to give the algorithm reliable signal, switching too early with too little data can produce unstable, unpredictable results.

What happens if the optimization event is tracked incorrectly?

The algorithm will optimize just as efficiently toward the flawed data, bidding to maximize whatever is actually being recorded, accurate tracking is a prerequisite for the whole system to work as intended.

What Are Optimization Events?

An optimization event is the specific conversion action an ad platform's algorithm is told to pursue and bid toward, a purchase, a lead form submission, an app install, a page view. It's the target outcome that shapes every bidding and delivery decision the system makes across a campaign.

How Optimization Events Works

When setting up a campaign, an advertiser selects which tracked action counts as the optimization event, and from that point forward, the platform's predictive bidding system evaluates every auction specifically based on its predicted likelihood of producing that event, not any other action a user might take. A campaign optimized for "Add to Cart" will behave very differently from one optimized for "Purchase," even running on identical creative and audience settings.

Choosing the wrong optimization event is one of the most common, and most consequential, setup mistakes, a campaign optimizing toward a shallow event (page view) when the business actually cares about a deeper one (purchase) will find plenty of cheap page views that never convert further, since that's literally what it was told to maximize.

Optimization Events in Action

An ecommerce brand sets Purchase as its optimization event rather than Add to Cart, ensuring the algorithm bids toward users likely to actually complete a transaction, not just add an item and abandon it.
A lead-gen business optimizes toward Lead Form Submission specifically, rather than Landing Page View, so the system prioritizes auctions likely to produce a genuine contact, not just a visit.
An app-based company initially optimizes toward App Install, then shifts to a deeper in-app purchase event once enough volume accumulates, since installs alone weren't reliably predicting paying users.

The Formula Behind Them

Optimization events aren't calculated, they're a configuration choice, but the metric that results from that choice follows the standard formula for Cost Per Optimization Event = Total Spend ÷ Total Events, essentially the same calculation as CPA, applied to whichever specific event was selected as the target.

Optimization events are closely tied to, but distinct from, conversion tracking generally. Conversion tracking records that various actions happened; the optimization event is the specific one out of all tracked actions that the bidding algorithm is actively told to pursue and optimize delivery toward.

Choosing the Right One

A campaign's results should always be read in light of what its optimization event actually was, a campaign showing a great cost per result when optimizing for Page View isn't necessarily a great campaign if the business needed Purchases, since a cheap shallow event doesn't guarantee the deeper outcome that actually matters follows from it.

Why the Choice Matters So Much

The optimization event is one of the single highest-leverage settings in a campaign, since it determines exactly what the entire predictive bidding system is working to maximize. Getting it wrong means the algorithm will very efficiently deliver the wrong outcome, technically working as designed, just optimized toward the wrong goal.

Frequently Asked Questions

Should a campaign always optimize toward the deepest possible event?: Generally yes once enough volume supports it, but a very deep, rare event can sometimes lack enough data for reliable bidding, in that case starting with a slightly shallower event and moving deeper as data accumulates is a common approach.
Can a campaign have more than one optimization event?: A campaign typically optimizes toward one primary event at a time, though some platforms support secondary or value-based optimization that considers multiple actions together.
Does the optimization event affect CPC or just conversions?: It affects the whole bidding strategy, since the system adjusts bids per auction based on predicted likelihood of that specific event, which indirectly shapes CPC, audience reach, and overall delivery.
How much data is needed before switching optimization events?: Enough conversion volume on the new, typically deeper event to give the algorithm reliable signal, switching too early with too little data can produce unstable, unpredictable results.
What happens if the optimization event is tracked incorrectly?: The algorithm will optimize just as efficiently toward the flawed data, bidding to maximize whatever is actually being recorded, accurate tracking is a prerequisite for the whole system to work as intended.

See NYX in action.

Book a demo