Meta Ads Optimization Event: Shallow vs. Real Conversions

Setting up a campaign in Meta Ads always involves a choice that looks simple and isn’t: which Meta Ads optimization event to point the campaign at. Picking the wrong one doesn’t stop the campaign or throw a visible error. It just makes the algorithm chase the wrong audience, or never properly exit the learning phase. This guide explains the difference between optimizing for a shallow event and for a real conversion, and how to decide which one to use at each stage of your Telegram operation.

What changes when you switch the optimization event

Meta Ads doesn’t deliver ads to “everyone similar to your audience.” It delivers to people who look like the ones who completed the event chosen as the optimization goal. Changing that event changes the audience the algorithm pursues, even with the same creative, the same budget and the same initial targeting. Optimizing for link clicks makes the algorithm look for people likely to click; optimizing for FTD makes it look for people likely to deposit. Those are different audiences, even coming from the same ad.

The event hierarchy in a Telegram funnel

A typical Telegram funnel has a chain of events, each rarer and more valuable than the one before:

Event	Volume	Signal quality
Ad link click	High	Low
Bridge page view (ViewContent)	High	Low to medium
Click on the button that leads to Telegram (Lead)	Medium to high	Low to medium
enter_channel (channel join)	Medium	Medium
Platform registration (CompleteRegistration), in iGaming/binary options	Medium to low	Medium to high
FTD / deposit / purchase	Low	High

See the guide to S2S postbacks in iGaming for how the deposit event reaches Meta via CAPI, and the FTD glossary entry if the term is still new to you. The further down the table, the more the event represents real intent, and the less often it happens.

The trade-off: volume vs. signal quality

Optimizing for a shallow event (click, page view) gives you high volume, so the campaign exits the learning phase quickly. But the audience the algorithm pursues is the people who click, not the people who actually convert, which usually means cheaper, weaker leads. Optimizing for a real conversion (FTD, purchase) gives you a more valuable signal. But if that event’s weekly volume is low, the algorithm doesn’t gather enough data to learn the pattern, and the campaign may never exit the learning phase in a stable way, delivering erratically and expensively in the meantime.

The most widely used reference in practice for considering a campaign out of the learning phase is around 50 optimization events per week, per ad set. As an illustration: a campaign optimizing for FTD in an operation that generates 10 deposits a week will struggle to get there; the same campaign optimizing for enter_channel, with hundreds of joins a week, exits the learning phase without trouble.

How to choose your Meta Ads optimization event at each stage

• New account or no history: optimizing for a shallower event (enter_channel) tends to work better at first. It gives you enough volume to exit the learning phase and start building account data.
• Mature account, with reasonable FTD volume: now it makes sense to test optimizing for FTD directly, or to run both campaigns in parallel. See the guide to optimizing campaigns for FTD for why that combination tends to beat picking just one.
• FTD volume still low even on a mature account: it may be worth optimizing for an intermediate event, such as platform registration (CompleteRegistration), and moving up to FTD as volume grows, instead of jumping straight to the rarest event of all.

Common mistake: switching events too often

Every change of optimization event restarts the ad set’s learning phase. The delivery history built up until then is discarded, and the algorithm starts its audience search from scratch. It’s common to see someone, after one weak week, switch the event trying to “fix” it, and unknowingly restart a process that was close to stabilizing. Before switching, confirm that event volume really supports the change, instead of reacting to a single bad week.

How Track4you helps with this decision

The decision depends on knowing the real volume of each event, and Track4you shows it at both ends of the funnel:

• From the funnel: the page visit, the button click and the channel join reach the pixel as ViewContent, Lead and enter_channel (plus left_channel on exit). The Dashboard shows “Entradas” (joins) by period, and “Métricas” (Metrics) shows each lead with the campaign’s UTMs, so you can compare which campaigns bring in joins that stay in the channel.
• From the platform: with Integrations (Pro and Enterprise plans), a registration reaches the pixel as CompleteRegistration, the FTD as the custom event FTD (confirmed once in Events Manager) and a deposit as Purchase, via the platform’s S2S postback. Every conversion the platform sends back goes to the pixel, with no time limit on Track4you’s side, even weeks after the click, and becomes signal for the algorithm; whether it appears credited to the campaign in Ads Manager is decided by Meta’s attribution window (default: 7 days after the click / 1 day after the view). The Registrations, FTDs and Deposits cards on the Dashboard show the volume of each.

Instead of guessing whether FTD volume can already sustain the learning phase, you can look at the real weekly volume and decide with data, not only with what Ads Manager shows, which counts only what Meta’s window credited. For join campaigns, Track4you’s recommended starting point is the Sales objective, Website as the conversion location and a custom conversion created from enter_channel. For the event names themselves and how to confirm them, see the guide to Meta Ads custom events; plans are at track4you.app.

Frequently asked questions

Terms like FTD and CAPI have quick definitions in the glossary. To understand why Meta’s attribution window also distorts this reading, see the guide to the Meta Ads attribution window.

Conclusion

There’s no universally “right” optimization event. There’s the right event for each stage and each data volume of your operation. An event that’s too shallow wastes the chance to target the people who actually convert; an event that’s too deep, without the volume behind it, keeps the campaign stuck in the learning phase. Understanding that hierarchy, and revisiting the choice as the account matures, avoids both mistakes: never moving the goal forward, and switching events too early and resetting learning for nothing.