How Meta Actually Works in 2026

[00:00:00] Tony: VP Paid Media at Common Thread Collective, and I'm here today to talk to you about a component of the CTC Canon, specifically how we think about modern Meta advertising. Meta is the most powerful advertising platform that's ever been created, and as of this moment in time, continues to represent the vast majority of share of wallet for e-commerce advertisers.

[00:00:29] the power of Meta's advertising engine is explicitly a function of two core systems that are relatively new in the ecosystem. Number one, Andromeda, a creative-first ad retrieval engine, and GEM, which is short for the Generative Evaluation Model, which is Meta's ranking intelligence layer operating at LLM scale capacity.

[00:01:03] And together, these two systems process billions of interactions daily making probabilistic allocation decisions that no human operator can match. And it's, it's in the interaction of these two systems that the CTC Canon, how we think about structuring and operating in a Meta Ads account is generated from So I wanna talk first a little bit about how Meta optimizes and, and targets in the new world of Andromeda and Gem.

[00:01:44] And before you can build an effective Meta advertising strategy, you have to understand this fundamental shift that has occurred in how the platform operates. And Meta moved from an audience-first world to a creative-first world. The system no longer relies primarily on your targeting selections. Instead, it evaluates creative elements to determine who sees your ads The infrastructure is Andromeda and Gem.

[00:02:12] And Andromeda operates on a creative-first paradigm using creative attributes user context, and behavioral signals to build a candidate set of el-eligible ads for each impression opportunity. Meta's engineering team describes Andromeda as enabling 10,000 10,000X increase in model capacity, allowing the system to evaluate far more ad candidates per impression than was previously possible.

[00:02:37] Paired with Gem, the ranking intelligence layer it determines what should be shown from that candidate set. And these two things are the underpinning of what is often colloquially referred to in our in our industry around the need for more creative into the, into the Meta Ads ecosystem. And, and these two technological engines are, are the reason for that.

[00:03:03] Another super important concept to understand to take into consideration for modern Meta advertising is something called the breakdown effect, and it's one of the most important things for us to internalize. Historical ROAS does not predict future ROAS. This always reminds me of some of you may be familiar with the gambler's fallacy.

[00:03:23] This, this concept is always related to, to me, to the gambler, gambler's fallacy. So if you flip a coin 10 times in a row and it lands on heads 10 times, it is not more likely to land on tails. So th-this is it's not a perfect analogy because the the Meta probabilistic engine is actually taking into consideration signals that are are actually real as opposed to a coin flip.

[00:03:48] It, it sort of helps me detach from the idea that historical ROAS is pred-is predictive of future ROAS. And ultimately, this is why making decisions based on yesterday's ROAS is fundamentally flawed. You're operating on incomplete information using a small sample set of historical outcomes to predict a future that a system models far more accurately than you can.

[00:04:11] And Meta's own documentation states this explicitly: "The relationship between historical per-performance and future delivery is not deterministic." And so the reason why this matters for account management is because understanding these systems changes everything about how you manage a Manage- Meta account.

[00:04:28] If the algorithm can evaluate 10, 10,000X more ad candidates per impression, your job is to give it more options. If the ranking intelligence operates on signals you cannot see, your job is to trust that that allocation decision it's making within your defined cost controls If historical performance does not predict future performance, your job is to stop making ad-level optimization decisions based on backward-looking dashboards.

[00:04:54] And trust me, as a lifetime career media buyer, I feel intimately how challenging these ideas are. But the power is in the system. And our job at CTC is to set the constraints, feed the machine, and let it work. And as a result of that, the Canon CTC's Meta Canon uses a, a very simple framework for our campaign structure.

[00:05:17] Basically, we have a three-tier hierarchy. At the campaign level, we're using sales objective in 95% of the cases. Value optimized with minimum ROAS is our starting point. ASC or CBO to leverage Meta's budget allocation and inflated budgets with cost control. At the ad set level, seven-day click optimization window is our default with broad targeting with ACQ and retention exclusions.

[00:05:41] And ROAS targets are set here, not at the campaign level. And at the ad level, we're using AI enhancements where the brand is-- where our brands are willing to test specifically dynamic tests and creative enhancements and multiple URL testing. We're leaning into machi-machine learning optimizations as, as much as possible.

[00:05:58] I wanna talk a little bit about the, the idea of bidding and cost controls and liquidity because this is another important aspect of modern Meta advertising. Inflated budgets with cost controls allow for fluidity of market access. And this is one of the most misunderstood concepts in Meta advertising.

[00:06:17] Meta calls this liquidity, the ability for the system to take advantage of moments of high usage. Meta's budget optimization documentation describes how budget liquidity improves delivery efficiency, and this leads to day-to-day spend variance because market price varies relative to, to usage. And we, we actually want this vary-variation.

[00:06:37] We wanna take advantage of cheap moments and pull back during expensive ones and force the constraint of profitability via minimum ROAS or a cost cap. Allow liquidity to move across the time spectrum. Another extremely important part of modern Meta architecture is signal quality and setup. The highest performing Meta accounts are built on a foundation of high-quality signal.

[00:07:00] This is not optional. What is included in high-quality signal is CAPI integration. First and foremost CAPI API sends server-side events directly from your back end to Meta which bypasses browser-based limitations like ad blockers and cookie restrictions and incomplete pixel fires. Event match quality score measures how well your server events can be matched to Meta user profiles.

[00:07:22] High EMQ scores seven out of 10 or greater ensure the system has reliable signal to optimize against. Shopify is CDC's preferred native integration for CAPI. The Shopify Meta integration provides an automatic server-side transmission with high-quality match rate right out of the box. It's, it's really easy to use.

[00:07:42] For custom builds or headless e-commerce platforms, you need to ensure that your engineering team implements CAPI with proper customer information parameters. And It's a critical check that server events are tracked and server events and tracked purchase signals match the total number of orders for Shopify.

[00:08:02] This is a data integrity requirement. If Meta is receiving eighty percent of your actual purchase events, the algorithm is optimizing on incomplete data, your performance will suffer, and measurement will be un-unreliable. Auto-audit this monthly at a minimum. And properly set up the distinction between audience definitions and new versus engaged versus existing customers.

[00:08:24] This is required to appropriately track these audience segments and acquisition versus retention or targeting campaigns. This is becoming even more important as the platform continues to evolve. Meta has brought back in optional life cycle settings and configurations at the campaign level, similar to the, the previous existing customer cap in the in vintage ASC campaigns.

[00:08:47] But making sure these, these audiences are configured at the account level is paramount. Finally, I just wanna talk briefly about the, the basic outline for campaign structure. If you come and work with CTC, this is how you're gonna hear us talk about setting up your, your campaigns, and this is the way that we believe we can take the most advantage of the algorithm the algorithmic engine.

[00:09:06] And essentially, there are two types of campaigns. There's evergreen campaigns, there's marketing moments campaigns. In an evergreen campaign, there is a split between acquisition and retention, okay? And contained within each of those is ad sets that are that we bring new creative into until they are full.

[00:09:29] Meta keeps changing the limits on us. Now the current limit is 150 ads. So acquisition campaign ad set number one, creative goes into it until it's full at 150, then ad set number two. Also, in this acquisition, evergreen campaign is a DABA-specific ad set, okay? So we have two arms. Ac- evergreen campaigns, marketing moment campaigns.

[00:09:53] In the evergreen arm, we have acquisition and retention, and in each of those, we have ad sets that get filled up sequentially, and we also have a DABA. The marketing moment campaigns have ad sets that are related to marketing moments. Pretty straightforward. You have a new sale, a product launch.

[00:10:09] Anything that is time-bound starts here, ends here, goes in marketing moment. Okay? Now the-- That's the cornerstone of the structure. We think about signal quality, we think about clean account structure, we think about budget liquidity. Let's talk a little bit about some of the ideas for resolving underspend or exploring opportunities Everything that orbits around, I'm not spending as much media dollars as I want or at the efficiency that I need.

[00:10:47] Phase number one, if this is the problem, creative expansion. Deploy additional creative into evergreen acquisition retention and active marketing moments campaigns, and prioritize the diversity of concepts. So not minor variations big swings, big differences in the creative. Creative variety is what moves the, is what moves media delivery on Meta.

[00:11:10] Okay? So that's number one. It is almost certainly true that the pathway to solve the media delivery problem that you are facing on Meta is through creative diversification and creative volume.

[00:11:23] Number two, bid surface expansion. Allow Meta to access additional conversion opportunities using different optimization mechanisms. For example, duplicate your evergreen evergreen campaigns into lowest cost or bid cap or even cost cap. using the same creative pool with a different bidding optimization is an opportunity to access a different audience.

[00:11:47] There's no additional segmentation needed. It's gonna capture a different type of audience that's gonna respond to a different bidding profile

[00:11:54] another area of opportunity for expansion is taking a look at underspend ads. So what we're talking about here is a delivery environment for ads that have not received sufficient spend in the core system, which is a, a problem or a challenge that is, I think, a very difficult part of this whole exercise.

[00:12:16] This whole exercise of the, the practical truth of the vast majority of the creative assets that we put into Meta are not going to achieve the results that we, that we want. The problem is we all make a lot of investment and time and energy to create the things, to build the things in the campaigns, and it can be challenging to see those creative assets not get delivery We believe that there is an opportunity to launch a dedicated underspend expansion ASC campaign with a single ad set that's value optimized, same T-ROAS target as evergreen, and include ads where spend in the evergreen is less than 3X the CPA target.

[00:12:55] This threshold defines ads that have not had sufficient opportunity to scale. It's important that you pull from evergreen only and do not introduce new creative do not include ads that have already had a m-- that have already meaningfully scaled. It's, it's a the core idea here is to grab anything that isn't scaling into the, the primary setup and give it another shot.

[00:13:19] And then phase four, the very last stage of this process is constraint relaxation. So expand total spend beyond the, the current efficiency limits only when justified by the business performance, AKA changing your ROAS target or your cost cap to something different than what the math says you need it to be.

[00:13:38] Be very careful. The day-to-day management of this entire system is designed for execution by, by one expert operator, a profit engineer, and the core pris- principle that enables this efficiency is simple but counterintuitive, and that is to never actually turn off an ad. The principle challenges decades of media buying orthodoxy, but it's fundamental to how modern meta advertising works, and this-- it's the single most common mistake we see brands make.

[00:14:06] The platform Meta themselves state historical ROAS does not equal future ROAS. The breakdown effect means that the probabilistic model Meta uses for future delivery is fundamentally disconnected from backward-looking summary statistics you see in your dashboard. And if you're running a CBO with cost controls and an ad has, has poor historical ROAS, the system has already reduced allocation to it.

[00:14:31] Turning it off removes optionality without improving the outcomes. So the primary job for media at CTC and profit engineers at CTC has three main responsibilities. Number one, ensure correct budget allocation with no limitations at the campaign level. Number two, ensure the bid matches the business objective.

[00:14:55] So we wanna make sure that we understand how we make money our incremental marginal return on the media, and make sure we have the bids set accordingly. And from there, we wanna identify and launch new creative. Feed the machine. The system performs best when high creative volume and diversity meet.

[00:15:11] That's it. That's modern meta advertising. That's the CTC canon