Marketero · Entrevistas
We Tested Meta Ads with $50M: Here’s Everything You Need to Know
21 ideas de marketing sacadas de «We Tested Meta Ads with $50M: Here’s Everything You Need to Know», de Perpetual Traffic. Sobre todo de medición y…
- 21 ideas despiezadas
- Fuente Perpetual Traffic
- Dura 38 min
- Publicada en 2026
Ver la entrevista entera en YouTube
De qué habla
Las 21 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 21 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.
Definición y benchmarks del Hold Rate en anuncios de video
Medición y experimentosEl Hold Rate se calcula dividiendo las reproducciones de 15 segundos entre las reproducciones de 3 segundos para medir la retención del contenido; un Hold Rate fuerte se sitúa entre el 40% y 50%, mientras que el umbral mínimo aceptable es de aproximadamente 30%.
which is hold rate. So this is 15 second views divided by three second views. So this measures how well the video retains attention after the initial hook. So a strong hold, super strong hold rate is 40% to 50%. You know, low hold rate is anywhere between 20 and 30%. We've got right around like 30% thereabouts is sort of our line of demarcation here. So anything that's below that Mendoza line, so to speak.
Escúchalo en el minuto 22:21
Benchmark de Hook Rate para anuncios en video en Meta Ads
Medición y experimentosEl Hook Rate mide el porcentaje de usuarios que ven al menos los primeros 3 segundos de un anuncio en video; ubicarse entre un 30% y un 40% indica un gancho visual efectivo y favorece la entrega de impresiones y frecuencia por parte del algoritmo de Meta.
Anything less than that, probably something that Meta will not give a whole lot of impressions to and a whole lot of frequency to. But if you're in the 30, 40% range on hook rate, you're doing pretty well. That means your video is engaging. And that hook rate just on the video, because it's the first three seconds, like that's the So the big thing here is hook rate is three second video views and you can just do this
Escúchalo en el minuto 20:24
Umbrales de hook rate para optimizar anuncios en Meta
Medición y experimentosEn la optimización de anuncios de Meta para captación de clientes, el algoritmo prioriza indicadores tempranos como el hook rate: tasas por debajo del 25% suelen relegar el anuncio a menor entrega ('ligas menores'), mientras que un 30% o superior representa un rendimiento óptimo para escalar.
In terms of success, Meta is not focusing on the sales. It's these leading indicators. Like if you are below a 25% hook rate, your ad is probably being sent down to the minor league. 30% and higher is great. And I'll show you an example of one that's hitting both these metrics. And we're using these hook rates to scale up to get new customers.
Escúchalo en el minuto 0:00
Canales de adquisiciónDesactivar creativos en Meta únicamente porque no registran conversiones directas dentro de la plataforma (a pesar de tener impresiones y gasto) perjudica el rendimiento general. El rol del anuncio no es cerrar la venta de forma aislada, sino captar la atención de compradores cualificados, generar confianza y alimentar el embudo, mientras que el cierre depende de factores como la landing page, las reseñas y el…
I was just actually on a client call just within the last hour and we were talking about the exact sort of stuff that we're going to talk about right now, which is if you are optimizing all of your ads and you're just looking for sales inside the platform and you're turning stuff off that doesn't have sales but has impressions and has spend, you are going to be in a world of hurt. And we tested this with over 50 million dollars of our own money. I think the biggest shift that we've been pushing on this subject matter is that like, yeah, Meta is not just optimizing for your sales. And on Meta…
Escúchalo en el minuto 5:15
Configurar hook rate y hold rate como columnas predeterminadas en Meta Ads
Medición y experimentosLos equipos de marketing deben guardar como vista predeterminada en el Administrador de Anuncios las columnas nativas de porcentaje de 'hook rate' y 'hold rate' para evaluar sistemáticamente la calidad del consumo creativo sin necesidad de cálculos manuales ni conversiones personalizadas.
So I actually have here on our column, like book hold rate. And you can, what we always do is we will customize our columns. And this one here is one that we actually saved as a default setting. You can simply search for either hold or for hook right here. There's a lot of different ones. Hold rate percentage is what you want to find. And then here's your hook rate, hook rate percentage right there. And obviously we just saved that as a preset. So anyways, just for all those that have been doing it for more than a year, you do not have to do the calculations. If you're running a marketing…
Escúchalo en el minuto 27:02
Uso de hook rate y hold rate para decidir qué creativos iterar
Medición y experimentosPara determinar qué creativos de video (como UGC o contenido de creadores) deben escalarse o iterarse, es más efectivo evaluar métricas de retención de atención (hook rate y hold rate) que guiarse únicamente por el CPA directo atribuido en la plataforma.
It's hook rate. The hold rate. Like I said, on this client call, we're like we're looking at hook rate and hold rate as the two metrics to determine, OK, which creatives do we need to make more of? Which ones do we need to do more iterations of? So the vast majority of the creatives that we're finding working right now and with very few exceptions are video content. The point is this, is that video right now is absolutely moving the needle, but we're not looking at the CPA that's associated with it to determine which creatives we should iterate upon, which is really interesting.
Escúchalo en el minuto 8:43
No apagar anuncios basándose solo en ventas directas si sus métricas de retención son altas
Canales de adquisiciónUn anuncio de parte alta del embudo (TOFU) no tiene como único rol cerrar la venta directa, sino captar la atención de prospectos cualificados y generar confianza. Apagar creativos únicamente porque no se les atribuye una conversión directa es un error si mantienen indicadores tempranos sólidos (como hook rate y hold rate), ya que Meta premia el consumo de contenido en la subasta y contribuyen al CPA combinado.
And that's where like you and I have both been like screaming, like stop turning off ads, because people are turning off ads that don't have credit to the acquisition, but they could have really strong leading indicators, the hook rate, the retention rate, like all of that to tell you that this is well consumed. And that's where I'm like really bullish on your ad is not responsible for the sale. Their ad is responsible for buying attention of qualified potential prospects, and then earning their trust. Your ad has to earn their trust into why you're going to earn their dollars. And then…
Escúchalo en el minuto 25:24
Monitoreo de métricas tempranas de consumo para diagnosticar anuncios
Medición y experimentosPara determinar la efectividad creativa y la viabilidad de una campaña antes del cierre de ventas, se deben analizar indicadores de consumo temprano como la tasa de gancho (hook rate), la tasa de retención (hold rate) y la tasa de apertura de texto (see more rate).
Like I know what we talk about, like MPIs and the metrics that matter, but there's other leading indicators, like hook rates. We have retention rate, ramp rates, all those numbers. But even so like the see more rate, which is the button to expand any copy, like all of that comes into play. And that is evaluating all of those additional metrics to determine whether or not the campaign at large and the success of your sales is going to work or not. And then you have, like you said, the retention rate that you guys are looking at. It's important to know are people consuming the content they…
Escúchalo en el minuto 11:05
Medición y experimentosLos anuncios no deben evaluarse de forma aislada midiendo únicamente el costo por adquisición (CPA) o la conversión inmediata. Los anuncios en video educan y generan confianza en la parte alta del embudo, mientras que los formatos estáticos suelen llevarse el crédito del último clic; evaluar la cuenta sin una visión agregada lleva a descartar erróneamente los formatos que impulsan la venta global.
But when you're looking at on a creative level, like individual ad, you can't look at the acquisition to determine in isolation. Then you have to like pull a bird's eye view out, look at the campaign and understand how they're all working in aggregate, because there's that concept where these ads work in tandem with each other. It might not be the reason why they take action, but it's adding more and more knowledge so that the user who has survived a trust recession, who is going through an AI blitz right now, and is just needing to know more before they invest their dollars. You have to know…
Escúchalo en el minuto 10:19
Fórmula y benchmarks del hook rate en anuncios de video para Meta
Medición y experimentosEl hook rate mide la capacidad del anuncio de detener el scroll y se calcula dividiendo las reproducciones de al menos 3 segundos entre las impresiones totales. Para anuncios de video en Meta Ads, el umbral mínimo aceptable de rendimiento es del 25% (por debajo de este valor el anuncio pierde tracción en la subasta), mientras que un desempeño saludable u óptimo se sitúa entre el 30% y el 40%.
First off, hook rate is three second views over impressions. So it measures how well the ad stops the scroll. And we aim for 30%. That's our target. What you can see on your screen is 30 to 40%, which is amazing. So we use as a benchmark 25%, like if it's below 25%, which is exactly in line with what, So 25% or greater is sort of our benchmark. Anyway, the point is, is like if you are below a 25% hook rate, your ad is probably being 30 to 40% is great. Average hook rate, you should be no less than 25. So what we're finding is an average hook rate of at least 25% is sort of that line.
Escúchalo en el minuto 17:07
Iteración de creativos de video a partir del punto de caída en las gráficas de retención
Estrategia de contenidosAl analizar anuncios de video en Meta, revisa la curva de retención para identificar el segundo exacto donde ocurre la caída de audiencia (drop-off); si el hook funciona pero la retención cae rápido, adelanta la demostración de producto, ajusta el ritmo del cuerpo del video o regraba con otro creador manteniendo el gancho inicial.
So that type of component, if you were to see this, like looking at this again in the charts, what you can see is like, you have the information, the video play, the average play time, the hook rate, the hold rate. Like you can take this and see that like at there's like the steep happens at like second three or four, where the average play is eight seconds. Like what someone could do with an iteration is, okay, maybe this video gets reshot with a different person and then they tighten up how long it takes for this. Or the, uh, the meat of what happens in that slope decline. What I would love…
Escúchalo en el minuto 31:28
Métricas de referencia de Hook Rate y Hold Rate en anuncios de video para Meta
Medición y experimentosUn hook rate (tasa de enganche en los primeros 3 segundos) de entre 30% y 45% se considera un benchmark sólido en Meta Ads (con cifras excepcionales superando el 46-54%), mientras que un hold rate saludable suele rondar el 20% a 30%.
Your 45% benchmark, mine's 30. I mean, I would like to see this on the hold rate, maybe a little bit higher. Like it's at 18%. Like I said, we're, we're around 20, 30% hold rates, but the hook rate is so good at 46%. 46% rate phenomenal. And then you can find what this next one you're showing is now a 21.45% hold rate, which is The last seven days was 21% hook rate of 54%.
Escúchalo en el minuto 30:25
Ajuste de anuncios con alto gancho pero baja tasa de retención y conversión
Mensaje y copywritingCuando un creativo logra un gancho inicial muy alto (por ejemplo, mediante técnicas sensoriales o ASMR) pero decae en tiempo de visualización (hold rate) y ventas, se debe introducir el producto más temprano en el video y añadir capas de audio explicativas.
It's the people that are actually providing the service, creating like this ASMR kind of feel to it without any pitch. So it's like we're getting tremendous hook rates, but the hold rate and the sales conversions of the entire campaign are not where we want to be. So what we decided, we said, okay, we're getting great on the ASMR sort of hook rate side of it. How can we introduce the product a little bit earlier and then do some audio overlay?
Escúchalo en el minuto 35:26
Optimización mediante indicadores adelantados creativos hacia conversiones
Medición y experimentosEn la compra de medios moderna, los anunciantes deben actuar como estrategas de crecimiento usando indicadores adelantados (como hook rate y hold rate) combinados con el rendimiento del conjunto de anuncios para predecir y empujar el indicador rezagado final, que son las conversiones.
So what I'm saying is that if you can use hook rate and hold rate in combination, plus looking at the entire ad set as to how you're doing from a new customer acquisition standpoint, that's how you're going to be able to read this data here. You're a growth strategist. And that means utilizing the data that we've seen here, leading indicators that ultimately lead to lagging indicator, which is conversions. And this is the sort of thing that's really moving the needle with creative diversification, especially.
Escúchalo en el minuto 36:25
La distribución publicitaria en Meta depende de indicadores tempranos de retención
Canales de adquisiciónEl algoritmo de entrega de Meta prioriza el tiempo de permanencia y la retención del usuario en la plataforma por encima del objetivo final de venta del anunciante. Como consecuencia, para maximizar el alcance y la escala en adquisición de nuevos clientes mediante video ads, se deben optimizar métricas predictivas (leading indicators) como el hook rate y hold rate, y no evaluar la creatividad únicamente por…
And we're seeing this with more impressions, more reach, because hook rate and hold rate, especially on video, are the two things that we really find that alongside with outbound click-through rates to a certain degree. These are the two different metrics here that we're finding that meta is really optimizing for. Yeah, they want you to make money, but they need people to stay on platform for them to make money. And in order for them to stay on platform, you have to have quality content. You have to understand what does meta want for you to get what you want. What meta wants is retention…
Escúchalo en el minuto 15:22
Evitar ganchos desconectados (bait-and-switch) para no desplomar la retención
Estrategia de contenidosAunque el gancho inicial debe captar la atención desde el 'segundo cero', debe mantener coherencia con el hilo narrativo del resto del anuncio para no generar un efecto de engaño que destruya la tasa de retención (Hold Rate) antes del mensaje de conversión.
designing for second zero. As soon as someone gets it, it's second zero. It's showing up before anything even plays. You have to capture the attention. And then you don't want to have that bait and switch where I'm like, oh, hey, like, boom, I popped a balloon. Whoa, that got my attention. What's happening here? And then if you don't have that thread through, if the storyline isn't going, you don't have an engaging enough ad for someone to get to the answer.
Escúchalo en el minuto 24:31
Canales de adquisiciónNo apagues anuncios de video de alto enganche guiándote únicamente por su CPA directo; anuncios con alto consumo y retención suelen generar reconocimiento que reduce el costo global de adquisición de nuevos clientes (NCAC) dentro de una estrategia de diversificación creativa.
So you can't neglect what is happening with the ad itself. It's just like, you can't look at only the sales performance because Meta is not looking at only the sales performance. The point is, is this is actually pulling in a lot of awareness and leading to a lowering of an overall NCAC. I mean, creative diversification. Creative diversification. But the point is, is like the overall how this is working in aggregate with all of the other creatives in this creative diversification strategy, it's working. And those are the types of things that you really want to look at and not just pure CPA…
Escúchalo en el minuto 31:42
Medición y experimentosPara anuncios estáticos con texto, una tasa de 'ver más' (See More Rate) de referencia es de aproximadamente 1.05%, lo que equivale a que al menos 1 de cada 100 personas se interese con los primeros 140 caracteres y haga clic para desplegar el resto del texto.
It's the see more rate where that's you're getting people to click that see more. We use that as a 1.05% as a bunch. I want at least every one in a hundred to be engaged with the first 140 characters and click see more to read it.
Escúchalo en el minuto 20:53
Meta evalúa los anuncios de forma individual en el motor de subasta
Canales de adquisiciónAunque Meta optimiza las campañas a nivel macro para generar conversiones y ventas, su motor de subasta y recuperación califica y selecciona cada anuncio de forma individual según su eficacia creativa y señales de consumo.
So I think the bottom line is this, is that, yes, Meta is, in fact, optimizing your ads for sales, but not the way it's going. They're optimizing your campaigns for sales. But the way they're evaluating the efficacy of your ads, the way they're saying, how am I going to choose you in the ad retrieval engine? They're not choosing you based off of sales and of a campaign level. This is an individual race. And the ads are being chosen and the ads are being qualified. Meta is evaluating your ads as if they're dogs and best in show. And so they're trying to put the best dog forward so that you can…
Escúchalo en el minuto 36:37
Canales de adquisiciónAproximadamente el 70 % de todo el tiempo de permanencia de los usuarios en Instagram se consume en formato de Reels y contenido de video.
But the thing is, the matter is that you look at it, I think like what 70 percent of all time on Instagram is reels. Their inventory, what's available, what people are consuming is definitely video. Video are reels in the meta ecosystem.
Escúchalo en el minuto 12:24
Medición y experimentosEn una campaña de prospección con segmentación amplia (broad) para e-commerce, un anuncio TOFU registró un costo por adquisición de nuevo cliente (purchase new customer) de $34 con una tasa de gancho (hook rate) del 46%.
This is an e-commerce brand. Broad targeting. Once again, like no targeting. The creative creates the targeting. This ad, keep in mind, is the top of funnel ad. This is new customer. This is a KPI import with Edge Tag. The purchase NC or purchase new client or new customer is 34 bucks, which is good. So this one actually is just one that I randomly put and it's got a hook rate of 46%.
Escúchalo en el minuto 28:54
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.