Marketero · Entrevistas
The Future of Media Buying: John Moran on cAPI Imports + Edge Tagging (Part 2)
15 ideas de marketing sacadas de «The Future of Media Buying: John Moran on cAPI Imports + Edge Tagging (Part 2)», de Perpetual Traffic. Sobre todo de…
- 15 ideas despiezadas
- Fuente Perpetual Traffic
- Dura 27 min
- Publicada en 2025
Ver la entrevista entera en YouTube
De qué habla
Las 15 ideas de esta entrevista se reparten así entre las etapas del marketing. El reparto es lo que distingue una entrevista de otra: dice en qué se detuvo de verdad quien hablaba.
Las 15 ideas, con su cita y su minuto
Cada una está sacada de un punto concreto de la entrevista. El enlace del titular lleva a
la página de esa idea, donde se ve quién más la dice por separado; el del pie salta al
segundo exacto del vídeo para que la compruebes tú.
Consolidar campañas en Meta Ads para evitar sobrecostos derivados del motor Andromeda
Canales de adquisiciónCon el motor Andromeda y el sistema de recuperación creativa de Meta, el propio contenido del anuncio define la segmentación; colocar el mismo creativo en diferentes campañas compite por la misma persona y duplica el gasto, por lo que es necesario consolidar la estructura en una sola campaña y conjunto de anuncios donde cada creativo sea único.
actually more important because of the Andromeda update that happened for everyone back in May. It started in December of 2024, finalized in May of 2025. And what the Andromeda engine does is it uses what they call creative retrieval. So a creative retrieval is, and this is actually just Google which is Andromeda basically says, I need to tell a story to these users because the functionality of Andromeda means that when the creative does the targeting, you have to consolidate your campaigns because campaign a with ad a and campaign B with also ad a is now one campaign because they're going to…
Escúchalo en el minuto 11:39
Validación e iteración de anuncios mediante asignación de presupuesto por hook
Medición y experimentosPara probar creativos en Meta, utiliza una nomenclatura estricta por concepto, gancho y oferta (ej. concepto 1, gancho 1, oferta 1) y evalúa el éxito del gancho observando qué variación absorbe mayor presupuesto publicitario de forma natural. Cuando identifiques el gancho ganador, genera 2 o 3 variaciones adicionales para comprobar si consolidan y mejoran la tracción.
this statement. I think we were, we're making originally, which is if you have good concepts and you're testing those concepts with good, with good nomenclature. So concept one, hook one, offer one concept, one hook to offer one concept, one hook three, offer one, you can identify. Okay. This one seems to have more ad spend. This is a better hook. Okay. Does it, is it working well? Can you make two or three other variations? Does it also perform as good, maybe a little bit worse or even better than that was happening? Yes. Do we see the other hooks now earning less ad spend? Yes. Okay. That's…
Escúchalo en el minuto 21:35
Distinción entre atribución de plataforma y contribución incremental
Medición y experimentosAumentar el presupuesto publicitario suele hacer que las plataformas como Meta o Google se apropien de una mayor porción de las ventas basales mediante datos modelados, inflando los resultados dentro del administrador de anuncios sin generar un crecimiento incremental real en el negocio.
advantage of having more accurate attribution. That's not their job. Right. But if meta just gets better at attributing them to themselves, that's good for meta. Now, if Google is better at attributing sales to themselves, it's, it's better for Google. Now, what you have is two hungry platforms that are getting really, really good at attributing sales to themselves through the guise of, well, we probably missed it. And that's, that's where it is both good and bad. So that's where it's hard more and more sales. Now snap back to your backend of your Shopify. And you're like, well, I still only…
Escúchalo en el minuto 3:41
Iterar variaciones creativas sobre los anuncios que absorben la mayor parte del presupuesto algorítmico
Canales de adquisiciónCuando el algoritmo de Meta asigna la mayor parte del presupuesto a un anuncio específico debido a su rendimiento de CPA y retención de gancho, no se debe pausar por alto gasto; la estrategia correcta es producir múltiples versiones e iteraciones con nuevos ganchos alrededor de ese mismo concepto ganador.
And it's actually fairly simplistic. Meta will do a lot of that for you initially. What I mean by that is if you have, let's say 10 ads and you always have Pareto rule where 80% of your sales And when you start to see ads that have a higher and higher CPA, and then you'll see the ads, those same ads that have higher CPA also start to earn less ad spend. So meta feeds more ad spend to or I have other ads that are resonating better with. Okay. And so when you start to see what ads have the highest amount of ad spend or what ad spend is increasing or growing, especially as you're scaling, if…
Escúchalo en el minuto 17:23
Diversificar ángulos y formatos creativos en un solo conjunto para facilitar la secuenciación algorítmica
Estrategia de contenidosPara maximizar el rendimiento con algoritmos basados en recuperación creativa, se debe alimentar un mismo conjunto de anuncios con múltiples variantes temáticas (historia del fundador, desglose de beneficios/características individuales y testimonios UGC), permitiendo al sistema reenganchar al usuario con un ángulo nuevo cuando la frecuencia de impacto llega a 2 o 3.
What we're going to do is what Meta wants is how do you capture the attention, build interest, get them to also want to take an action and be happy with that action. And you have to do it all inside of one ad set. So they have something called like the founder story. People like buying products from people. They don't really necessarily want to buy from brand brands. They want to buy from, from users and from people. That's why UGC is so popular and, and social media is so social. So the founder story is like telling a person, Hey, this is I'm John Moran. The reason why I understand why…
Escúchalo en el minuto 12:53
Explicar el valor y funcionamiento del producto en los anuncios en lugar de limitarse a promocionar descuentos
Mensaje y copywritingLos anuncios no deben depender únicamente de ganchos de descuento sin contexto, sino emular una venta directa tradicional explicando con claridad qué es el producto, por qué es único y cuáles son sus beneficios a través de diversos ángulos creativos (historias de fundadores, comparativas, características).
at a booth and I have the remarkable tab and I work at a remarkable and someone comes up and says, what's that? I'm like 20% off. I said, what is that? 20% off. Buy it. What is it? I don't know. So that's what it is. That's what we do. That's not how humans buy. No, no, but that's how we build ads. Right. So it's a really cool ad. We'll make some mega starburst come out the back of it and say 20% off sale. It's like, well, it's right. So you have to tell them, well, what is it? It's remarkable. What's that? It's a drawing tablet. Okay. Pen and paper does that buy. Yeah. So you have to think…
Escúchalo en el minuto 15:16
La sobre-atribución de ventas surge porque el retargeting absorbe tráfico de todos los canales
Medición y experimentosLas plataformas de anuncios reclaman ventas duplicadas porque sus píxeles de remarketing reimpactan a cualquier usuario que visite el sitio web, sin importar qué canal originó la visita inicial. Esto genera que múltiples canales (Google, Meta, TikTok) se atribuyan individualmente las mismas transacciones del backend.
site, they are going to remarket them and try to bring them back to your site. But the area where they dive into in order to tag that person to remarket has all users through every other channel Did you buy something? No. Okay. You left. I have to have my pixel tag you to remarket. So that one click turns into technically now potentially a 10,000 user remarketing campaign because that's that day is because Meta again, does it, it's not going to remarket only its own traffic. It has to that day or maybe that week or whatever, but yep. Yes. So that's why the platforms over attribute because due…
Escúchalo en el minuto 8:38
Medición y experimentosEl modelo de atribución de último clic de Meta asigna todo el crédito de la compra al anuncio final, ignorando los impactos publicitarios previos que educaron y convencieron al usuario a lo largo del ciclo de venta. Guiar la toma de decisiones únicamente por el CPA dentro de la plataforma puede provocar que se apaguen anuncios esenciales en el proceso de decisión.
There's a sales process. Now, if you have 10 ads that all had an actual impact on the user in order to purchase something, but the attribution of Meta says, well, you get to have one point and it's last click. Remember Meta's attribution is last click attribution. So if I have 10 ads and, but I can only give credit to one of those, but they all are equal. What do you do? Like, where do you put that? Where do you put that? So it has to go to last click. So what I've been using as is CPA is good. I understand it, but just because it wasn't the last thing doesn't mean that it wasn't the best…
Escúchalo en el minuto 20:17
Medición y experimentosEn plataformas como Meta y Google, entre el 40% (en anuncios de imagen basados en clics) y hasta el 60% (en campañas centradas en video por ventanas de visualización de un día) de los datos de conversión mostrados no son registros directos reales, sino atribuciones modeladas algorítmicamente.
if you look at Meta and or Google, like what percentage do you think is modeled? Oh, 40%. It's 40% as much as I can identify it. About 40%. Okay. Now if you have videos, it's a little bit more like 60% because you got your one day view and then your one day engage view. One day view is imagery, one day engage views video. So if you have all image ads, it's 40% of what you're seeing inside And so they're true. So it's 40% if it's image ads, which require more of a click. But then if you You heard it here for sure. 40 to 60% of what you see inside the ad platforms is modeled. It's not actual…
Escúchalo en el minuto 6:26
Mensaje y copywritingLos creativos publicitarios deben construirse de forma explícita y clara para que la inteligencia artificial de Meta entienda con precisión qué producto u oferta se está vendiendo antes de que intente convencer al usuario. Si el anuncio es vago o confuso, la IA no podrá mapear correctamente la audiencia idónea para esa creatividad.
that is actually really interesting. We don't create ads for people anymore. We have to create ads for AI first. Yeah. We have to convince Meta that this is what product we have. And then Meta says, okay, I understand it. I will go find the user now. Yeah. So if, again, if your ad is, Hey guys, what's that thing on your table over there? And I say 20% off. Yeah. Imagine telling that to a robot. Yeah. You're going to be even more confused than the person. Absolutely. Oh my God.
Escúchalo en el minuto 22:17
Comprender la causa del rendimiento publicitario antes de que falle
Medición y experimentosConformarse con ver conversiones sin entender el motivo exacto por el que ocurren funciona hasta que las campañas dejan de rendir. Si no se comprende la causa del éxito mediante datos precisos y trazables, resulta imposible diagnosticar y corregir el problema cuando el rendimiento cae.
you could, you could take the approach where it's like, well, as long as it's counting a conversion from somewhere, I'm happy, but it always will work until it doesn't. And then I always say, if you're not understanding why it's doing well, you won't know how to fix it when it's not doing well. So that's, that's what this does is it allows you to identify, okay, that is actually doing well. And I know why, because I can see it, you know, it's good data in good data out.
Escúchalo en el minuto 25:29
Proporción de datos modelados en los reportes de Meta Ads
Medición y experimentosEntre el 40% y el 60% de los datos y métricas reportados dentro de la plataforma de Meta corresponden a conversiones estimadas estadísticamente mediante modelado y no a mediciones directas exactas.
obviously don't necessarily believe everything that you see inside of the platform itself, knowing full well that we're talking about anywhere between 40 to 60% of it is modeled. You know, whether you take a
Escúchalo en el minuto 22:44
Las plataformas publicitarias destinan entre 30 y 40 centavos por cada dólar a remarketing automático
Canales de adquisiciónMeta y Google asignan aproximadamente entre el 30% y el 40% del presupuesto de pauta en reimpactar a usuarios que ya llegaron al sitio web, lo que explica por qué la frecuencia de campaña sube de 1 incluso en campañas de prospección.
all of the users that Meta and Google bring to your site, they are going to use approximately 30 to 40 cents of every dollar in ad spend remarketing them to bring them back. So that's why you see those. That's frequency that exists because as long as frequency is above one, you have remarketing. Right. Right. Well, re-engagement also.
Escúchalo en el minuto 7:50
Nivel de precisión de la atribución nativa de Shopify
Medición y experimentosLa atribución nativa en Shopify alcanza aproximadamente un 55% de precisión en Estados Unidos y un 45% en territorios asiáticos debido a limitaciones de seguimiento y bloqueadores de anuncios.
well, they could have had ad blockers. I mean, even Shopify says that their attribution is only 55% accurate in US territories, 45% accurate in Asian territories because of those issues.
Escúchalo en el minuto 2:54
Medición y experimentosEn un negocio con 12.000 ventas diarias, una inversión inicial de solo 120 dólares en Meta resultó en 4.000 ventas atribuidas dentro de la plataforma, ilustrando cómo los algoritmos pueden canibalizar y adjudicarse crédito de demanda orgánica o de otros canales.
a very, large promotional company, a lot of TV commercials, and they wanted to try Meta. And I said, I'm going to predict what is going to happen. And I told them, I said, this is a few months ago. I said, you're going to start your campaigns on Meta, and you're going to see some unbelievable results. Now, I asked them, how many sales do they have a day? They have about 12,000 sales a day. And I said, just wait to see that. And they said, yeah, John, you were right. We launched a Meta campaign. We spent $120. We had 4,000 sales. And I was like, wow, that is amazing. He's like, we have like a…
Escúchalo en el minuto 2:03
Seguir leyendo
Esta es una de las 1872 entrevistas despiezadas. Puedes
ver todas, leer las ideas por tema o
empezar por la ruta guiada si no sabes por dónde.