Marketero · Entrevistas
How to MAXIMIZE your Facebook Ads in Q4
31 ideas de marketing sacadas de «How to MAXIMIZE your Facebook Ads in Q4», de Professor Charley T. Sobre todo de canales de adquisición, medición y…
- 31 ideas despiezadas
- Fuente Professor Charley T
- Dura 1 h 01 min
- Publicada en 2024
Ver la entrevista entera en YouTube
De qué habla
Las 31 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 31 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.
Estructura de campaña broad con un control y dos variables de prueba
Medición y experimentosPara aplicar el método científico y escalar cuentas publicitarias, la campaña broad principal debe organizarse en tres conjuntos de anuncios: uno de control probado y dos conjuntos de variables que prueben variaciones de un mismo elemento contra el control.
Is there a recommended number of ad sets we have in our main broad campaign? Three. One control. Two variables. Those variables should be two variations of the same thing. So it's an A-B split test against a single element of your control ad set. That's how you set up the scientific method. That's how you scale out accounts.
Escúchalo en el minuto 29:58
Oferta y precioLa publicidad de adquisición no debe repartirse entre todo el catálogo de productos o servicios, sino concentrarse en la oferta de entrada que produce la mayor tasa de retención y compras recurrentes. Dado que la primera transacción suele ser la menos rentable, el objetivo de la captación pagada debe ser adquirir clientes que generen flujo de caja futuro con segundas y terceras compras.
But what if one of those products brings people to back come back and buy three more times? And all of the other ones make sure that when somebody buys that product, what if of 12 products, 11 of them, if somebody buys that product, they never buy again. And on the 12th, if somebody buys that product, they buy three more times. Why are you promoting the other 11 products? Is the profit on those other 11 products so good that it's better than getting future cash flow? Remember, the first transaction is the least profitable. The first transaction should be set up to acquire future transactions…
Escúchalo en el minuto 35:52
Estructura de tres campañas por objetivos de negocio en Meta Ads
Canales de adquisiciónPara escalar marcas de alto volumen, la estructura recomendada se divide en tres campañas con objetivos comerciales independientes: 1) Campaña Hero para generar volumen de clientes recurrentes, 2) Campaña de catálogo para conseguir el mayor volumen al menor costo, y 3) Campaña de ofertas de bundles para atraer clientes de mayor valor. No se deben mezclar objetivos de negocio distintos dentro de una misma campaña.
We have our One Hero campaign specifically designed to bring in a volume of customers that buy more than once. We have a catalog campaign designed to bring in the most amount of customers to the lowest cost. And then we have a bundle offer style campaign, kind of things we were talking about early in this video, specifically to attract the highest quality customers. And ultimately, based on our business objectives and our business needs, our investment between those three variations tends to shift so we can maximize our opportunity. No, Marvin, those are all in three different campaigns. Each…
Escúchalo en el minuto 32:51
Canales de adquisiciónEvita crear reglas para pausar anuncios durante las horas de la madrugada en que no hay compras directas. Aunque apagarlos aumente la eficiencia aparente a corto plazo, reduce el volumen de impresiones y prospectos tempranos, limitando la capacidad del algoritmo para escalar a largo plazo.
Should you set up rules to turn them off during that time? Short answer is no. I would doubt that the bad hours are really spending any significant amount of money. And I'll bet you if you were to turn those off, yes, your campaigns will get more efficient, but they would lose efficiency over time because all of the people that you're showing ads to and people who are not ready to buy, all of those people, well, that count of them is going to go down. So if you're really optimizing to get better, but it ultimately means you just have fewer people to show your ads to, at some point, that…
Escúchalo en el minuto 57:11
Creación de bundles para campañas estacionales basados en el historial de clientes recurrentes
Oferta y precioEn lugar de lanzar descuentos porcentuales generalizados que atraen clientes de baja calidad durante temporadas de alta competencia publicitaria, analiza el historial de compras para empaquetar los productos que los clientes de mayor valor suelen comprar en su primera y segunda transacción, estructurando una oferta escalonada (bueno, mejor, superior) para elevar el AOV y el LTV.
When it comes to Q4, what I want you to do right now is scrap your idea of a 20% site-wide discount. No, I want you to get rid of the idea of we're going to define success as paying more money to get a lower quality customer who's less likely to ever spend money with us again. Stop running discounts when the marketing efforts are more expensive. Well, the simplest way is what about if you just looked at everybody that bought from you twice? What if you put together an offer that was the first and second thing they bought? That's going to raise your AOV. And if anybody buys that offer, odds…
Escúchalo en el minuto 11:30
Alinear el evento de conversión en Meta Ads con el volumen necesario para salir de la fase de aprendizaje
Medición y experimentosPara optimizar una campaña por un evento profundo del embudo (como lead calificado o suscriptor), el presupuesto y la tasa de conversión deben permitir generar el volumen suficiente para salir de la fase de aprendizaje. Si el volumen semanal es bajo, es preferible optimizar por el evento previo (como lead general) y mejorar la conversión en el embudo propio antes de cambiar el objetivo.
And so when trying to ultimately determine what we're doing, we have to remember that we need to get the campaign out of learning phase. So for instance, if one out of 10 leads becomes an engaged lead, and we can only afford to buy 100 leads a week, it's very unlikely that we'll be able to optimize for engaged leads because we don't have the unit economics to get that ad set out of the learning phase, in which case we're kind of wasting our money. It would be a lot better in that case to take the leads route and make the customer journey effective enough that we go from a 10% conversion rate…
Escúchalo en el minuto 22:31
Aislamiento de concepto y formato dentro de pruebas creativas dinámicas (DCT)
Medición y experimentosCada conjunto de anuncios de creativos dinámicos (DCT) debe probar iteraciones de un único concepto y mantener homogeneidad de formato, evitando mezclar video con imágenes estáticas para poder aislar la variable y comprender con claridad el motivo de rendimiento de cada variante.
All right, so anyway, Robert, creative diversity should happen with different concepts. Each DCT should be its own concept. It's which of these iterations of this concept did better or worse and why. I would not pair video with static. I agree with that. Just make sure that you understand that the reason you run any test is ultimately to answer a question. So if your test is designed to answer a question, then you're sitting in a great place.
Escúchalo en el minuto 28:17
Reemplazar ofertas en campañas activas en lugar de duplicar para conservar el aprendizaje
Canales de adquisiciónAl cambiar un producto o tripwire en un embudo publicitario, es preferible sustituirlo dentro de la campaña activa en lugar de crear un duplicado desde cero, aprovechando así la fase de aprendizaje y el historial de datos ya acumulados.
And the doubt is if we change the Tripwire to another offer, which is more expensive, it will affect the learning of the campaign. I would just swap it out. Yes, it will impact the learning, but at least you're starting from somewhere. I'd rather start from somewhere than from nowhere. Let's lean into what we've already built.
Escúchalo en el minuto 47:50
Estructura de prueba DCT y consolidación en conjuntos ganadores
Canales de adquisiciónEl método estándar de escalado basado en machine learning consiste en probar diferentes componentes creativos en conjuntos de anuncios dinámicos (DCT) separados y luego consolidar los anuncios ganadores en un solo conjunto de anuncios de escalado.
Do you structure the campaign different with different ad sets and test different components in each ad set using dynamic creatives and then bundle the winning ads from each dynamic creatives into one ad set using the mentioned? Yeah. This is exactly what we do. Is this the way to scale further? Absolutely. Because you're leveraging machine learning instead of luck.
Escúchalo en el minuto 59:54
Desactivar Advantage+ Creative durante las pruebas creativas si se cuenta con diseñadores
Canales de adquisiciónDurante la fase de creative testing es preferible evitar las optimizaciones dinámicas automatizadas de la plataforma si se dispone de diseñadores que generen diversidad creativa intencional, reservando la automatización solo cuando no existan recursos de diseño.
I generally avoid using the advantage plus creative optimization while I'm doing creative testing. I think it can help for creative diversity, but I'd rather lean on designers to do that than lying on Facebook. But that being said, if I don't have designers, 100% use the advantage creative optimization.
Escúchalo en el minuto 50:29
Medición y experimentosLas pruebas creativas en Facebook Ads deben enfocarse en resolver un problema específico mediante variaciones dinámicas (como formatos 3:2:2). Estas deben lanzarse en CBO compitiendo directamente contra el conjunto de anuncios ganadores con el fin de superar y reemplazar el anuncio más débil.
So the point here is I am not shooting for top and bottom of funnel efforts in the same three to two ad. The point of a three to two ad is to ultimately focus on solving a specific problem. Every time you launch a creative test, it should be done with multiple variations. So in this case, we might have two three to two ads launched in a CBO against an ad set of our proven winners where the specific purpose is to ultimately improve the weakest ad of our proven winners.
Escúchalo en el minuto 3:55
Medición y experimentosOptimizar la atribución en publicidad digital recortando las conversiones view-through para centrarse solo en clics puede hacer que las métricas del canal parezcan mejores, pero reduce el impacto incremental real y deteriora el crecimiento del negocio en su conjunto. La atribución debe tratarse como una herramienta de datos y no como una carrera por adjudicarse crédito comercial.
It's not like I was in a conversation with somebody earlier today and they said, well, when we're looking at seven-day click, one-day view, look at all these view-through conversions. We've got to get rid of view-through. We're going to optimize just for click. And somebody else said, oh, yeah, that makes total sense. When we did it, our meta performance improved dramatically. Our Facebook ads got better and better and better. And I asked them, well, what happened to the business? They said, well, the client ultimately didn't really work out and their business kind of shrunk away. Yeah…
Escúchalo en el minuto 6:28
Consolidar presupuesto en un solo anuncio dinámico para evitar canibalización y fragmentación de datos
Canales de adquisiciónEs preferible concentrar el aprendizaje en un único activo dinámico potente (metodología 3:2:2) dentro del conjunto de anuncios en lugar de fragmentar el presupuesto en múltiples anuncios flexibles paralelos. Dividir en varios anuncios provoca canibalización entre ellos y mantiene a cada uno con datos insuficientes para optimizar.
And so really the primary difference of having it all in one ad versus three is do you want to have three things that are kind of smart or one really smart thing? Do you want to have three cannibalistic things to keep each other dumb, or do you want to invest having one really, really strong asset? For me, let me have one really, really strong asset with actionable data instead of three lottery tickets that keep each other held back.
Escúchalo en el minuto 55:43
No pausar anuncios basándose exclusivamente en fluctuaciones a nivel de anuncio individual
Medición y experimentosEl descenso en el rendimiento de un anuncio específico no implica necesariamente que la campaña o el negocio global estén empeorando; evaluar el rendimiento requiere mirar el impacto comercial total y no depender ciegamente de la atribución individual.
Do you turn off winning ads when they're decreased to performance? No, I don't always turn off ads when they decrease in their performance because one ad getting better or worse does not actually mean that the campaign is getting better or worse or that our business is getting better or worse. And one of the beauties of running our simple campaign architectures, we're going to be able to see like, oh yes, in fact, this ad getting worse did detrimentally impact everything else. But for those of you that are running a lot of things, you have no idea if an ad doing
Escúchalo en el minuto 48:41
Estructuración de campañas separadas por ciudad en negocios regionales
Canales de adquisiciónCuando se gestiona publicidad para múltiples ciudades o anunciantes locales independientes, se debe crear una campaña dedicada por cada ciudad en lugar de segmentar a nivel de conjunto de anuncios, garantizando una asignación justa del presupuesto y evitando que el algoritmo concentre el gasto en una sola plaza.
My concern about doing it at the ad set level could mean one city would earn more spend leading to one advertiser unfairly getting more leads in other cities. Ryan, yes, you need one campaign per city. You're going to have a one campaign, but it's going to operate for each city. The point is, if you're each campaign has a business objective. In this scenario, each campaign is supporting each advertiser of each city. Ryan, your business model has set you up to require a one campaign architecture for each individual city.
Escúchalo en el minuto 26:20
Validar ofertas de rebote/upsell en retargeting antes de escalarlas a la oferta principal
Oferta y precioPrueba y valida las ofertas de rebote o upsell ('rebupsell') primero con audiencias de retargeting de alta intención. Una vez validadas en el fondo del embudo, desplázalas hacia arriba para convertirlas en la oferta principal (hero offer) para tráfico frío y adquisición amplia.
This is a rebuttal upsell campaigns. Start that with your retargeting audiences, right? Start it there. Validate it. Move it further up the funnel. And, hey, if you do it right, it can actually be your to-market broad offer. Instead of being the rebuttal upsell offer, it became the hero offer.
Escúchalo en el minuto 16:39
El aprendizaje del algoritmo depende del volumen y calidad de datos, no del tamaño de audiencia
Medición y experimentosLa velocidad y efectividad de aprendizaje del algoritmo publicitario no dependen del tamaño del público objetivo, sino del volumen y la calidad de los datos generados. Aunque una audiencia pequeña pueda resultar más costosa en CPM, si los datos de comportamiento son consistentes, la estructura puede escalar sin problemas a otros mercados.
The honest truth is that the size of the audience doesn't matter. It's around the volume and the quality of the data. You can do a small city. That's fine. It might be more expensive because the audience is smaller. But if the data is congruent to other markets, you should have no problem scaling.
Escúchalo en el minuto 53:33
Descuidar la actividad orgánica eleva el CPM de las campañas pagadas
Canales de adquisiciónIgnorar la presencia y el contenido orgánico en plataformas como Facebook e Instagram encarece los costos de adquisición y eleva el CPM al inicio. El algoritmo penaliza con costes más altos a las cuentas que solo buscan extraer dinero mediante anuncios sin contribuir positivamente a la experiencia del usuario dentro de la red.
Is it normal to have a high CPM at the beginning? I've tried multiple times to create when my CPM is high. $50 in UAE. Yeah, it can be normal, especially if you're not running any organic. This is, again, getting back to the creative diversity video. If you're not running any organic, if you're ignoring the way that the platform sees you as a business partner, other than just trying to spend money on it to extract money by giving its users a bad experience, you're going to pay more money. If you are not taking care of your organic social, your Facebook, your Instagram, if you're not worried…
Escúchalo en el minuto 9:31
Estrategia y decisiones de negocioCopiar lo que hacen las marcas gigantes (de 50M$ o más) es un error para negocios en etapas tempranas (<1M$); las grandes empresas sobreviven a anuncios mediocres gracias a la fortaleza de su modelo de negocio, distribución física y servicio al cliente, ventajas que un negocio pequeño no posee.
There's no way you're going to 20x your business copying what they're doing now to get where they are today. And the reason I say that is if you're doing less than 1 million, copying what somebody who does 50 million a year isn't going to be what gets you there. Don't look to what people are doing now as a path to get where they are. Also fully understand that, well, to be fair, really top brands are there because of the customer service and the business model and a lot of things that have very little to do with their Facebook ads. You might be good at running ads, but the reason your…
Escúchalo en el minuto 18:06
Canales de adquisiciónCuando una cuenta publicitaria cuenta con anuncios ganadores y la solidez necesaria para escalar el presupuesto, se debe detener el testeo incesante de nuevos creativos y conceptos. Continuar buscando alternativas cuando ya se encontró lo que funciona genera costos innecesarios y distrae del objetivo de crecer el negocio invirtiendo más capital.
If you're winning, stop trying to continue to see what works. I'm not going to spend more money and time continuing to look for it. But people do that on Facebook ads all the time. They have an ad account strong enough to scale their budget. And they say, great, this is working. What I'm going to do now is continue to look for things that would put me in a situation I'm currently in. If the point of running your ads is to be successful enough to invest more so you can grow your company, if you're in a position where you can invest more so you can grow your company, why would you continue…
Escúchalo en el minuto 31:42
Canales de adquisiciónCuando una cuenta publicitaria cuenta con anuncios ganadores y la solidez necesaria para escalar el presupuesto, se debe detener el testeo incesante de nuevos creativos y conceptos. Continuar buscando alternativas cuando ya se encontró lo que funciona genera costos innecesarios y distrae del objetivo de crecer el negocio invirtiendo más capital.
If you're winning, stop trying to continue to see what works. I'm not going to spend more money and time continuing to look for it. But people do that on Facebook ads all the time. They have an ad account strong enough to scale their budget. And they say, great, this is working. What I'm going to do now is continue to look for things that would put me in a situation I'm currently in. If the point of running your ads is to be successful enough to invest more so you can grow your company, if you're in a position where you can invest more so you can grow your company, why would you continue…
Escúchalo en el minuto 31:42
Desventaja del ensamblaje manual de anuncios frente a la optimización algorítmica
Canales de adquisiciónConfigurar anuncios individuales asumiendo que se conoce mejor que el algoritmo el emparejamiento óptimo de texto y creatividad anula la ventaja del machine learning; las pruebas multivariadas permiten al sistema identificar las combinaciones más eficaces.
If you are still building ads one by one by one thinking that you are smarter than the machine at pairing copy with creative and trying your best to hack the machine. You are just you are misusing the world's smartest machine by undermining the thing that is its unfair advantage to ultimately play the lottery while keeping the machine dumb. You could be leveraging the scientific method in the world's strongest AI ever invented.
Escúchalo en el minuto 4:28
Priorizar creativos duraderos frente a la rotación constante de anuncios
Canales de adquisiciónLanzar creativos nuevos a diario fragmenta los datos, destruye la optimización del algoritmo de machine learning y satura la cuenta con anuncios de bajo volumen. Resulta más rentable y estable construir activos creativos sólidos y duraderos que acumulen volumen de compradores en el tiempo.
If you launch a new ad every day for a year, that's just 27 buyers per post. But my average ad in that machine is over 190 new customers. And it goes on to talk about why instead of launching new ads often enough that you create ad fatigue and ultimately destroy machine learning, you can instead focus on building strong and smart assets to deliver you stability.
Escúchalo en el minuto 24:17
Estrategia y decisiones de negocioEnfocar el negocio exclusivamente en obtener rentabilidad en la primera transacción mediante descuentos agresivos contamina la base de datos con clientes de baja calidad que no volverán a comprar, descuidando el desarrollo de un producto o servicio capaz de generar retención y valor de vida sostenible.
If you're still trying to focus on getting the highest volume of customer who will never buy from you again, do not complain when in mid-February everything goes to shit. Because you've invested in polluting the data set of your business to acquire the lowest quality customers. You need to make a mental shift of trying to be good at selling anything you can and starting to be good at buying the right type of customers. It is not about contribution margin on a single transaction. If you have ever said the thing of, I need to be first order profitable. What you're saying is, I don't have a…
Escúchalo en el minuto 12:53
Estrategia y decisiones de negocioEnfocar el negocio exclusivamente en obtener rentabilidad en la primera transacción mediante descuentos agresivos contamina la base de datos con clientes de baja calidad que no volverán a comprar, descuidando el desarrollo de un producto o servicio capaz de generar retención y valor de vida sostenible.
If you're still trying to focus on getting the highest volume of customer who will never buy from you again, do not complain when in mid-February everything goes to shit. Because you've invested in polluting the data set of your business to acquire the lowest quality customers. You need to make a mental shift of trying to be good at selling anything you can and starting to be good at buying the right type of customers. It is not about contribution margin on a single transaction. If you have ever said the thing of, I need to be first order profitable. What you're saying is, I don't have a…
Escúchalo en el minuto 12:53
Estrategia y decisiones de negocioEnfocar el negocio exclusivamente en obtener rentabilidad en la primera transacción mediante descuentos agresivos contamina la base de datos con clientes de baja calidad que no volverán a comprar, descuidando el desarrollo de un producto o servicio capaz de generar retención y valor de vida sostenible.
If you're still trying to focus on getting the highest volume of customer who will never buy from you again, do not complain when in mid-February everything goes to shit. Because you've invested in polluting the data set of your business to acquire the lowest quality customers. You need to make a mental shift of trying to be good at selling anything you can and starting to be good at buying the right type of customers. It is not about contribution margin on a single transaction. If you have ever said the thing of, I need to be first order profitable. What you're saying is, I don't have a…
Escúchalo en el minuto 12:53
Impacto del contenido orgánico y la experiencia de usuario en los CPMs publicitarios
Canales de adquisiciónTener un contenido orgánico deficiente o inexistente incrementa los costos publicitarios (CPM) porque priva al algoritmo de datos sobre qué usuarios interactúan positivamente con la marca y porque los anuncios puramente agresivos de respuesta directa atentan contra el objetivo de la plataforma de retener la atención de los usuarios.
Ultimately, one of the biggest predetermining factors for what you pay in a CPM is your business relationship with the platform. And if your organic content sucks, then you're going to be paying a lot more money because of three reasons. One, the platform doesn't know who to show content to because you've given it no data. Number two, if your organic sucks or isn't there, odds are your paid media has nothing to do with making people happy to see your content. So you are probably pushing aggressive DR-style content that is ultimately just a liability to Facebook's business model. Remember, the…
Escúchalo en el minuto 10:16
Retorno de inversión en captación local para estudio de pilates
Canales de adquisiciónCon una inversión de aproximadamente 500 dólares en anuncios locales de Instagram integrados con ManyChat, un negocio local de pilates generó más de 30.000 dólares en un periodo de 6 semanas dentro de un radio de 5 millas.
How do you scale ads for local businesses in the serving area for that business is a max of five miles. I actually have a great video on ads for local businesses with real world case studies around my wife's Pilates business and leveraging Instagram ads and many chat and all sorts of stuff. There's a wonderful case study in there. I think we spent about 500 bucks and we made over 30,000 in a six week period.
Escúchalo en el minuto 1:00:29
Oferta y precioOptimizar la estrategia publicitaria persiguiendo únicamente un CPA bajo a costa de ofrecer descuentos continuos erosiona la rentabilidad y la integridad comercial del negocio en lugar de construir un crecimiento sostenible.
And quick little note, if you think the path to success is discounts and getting a low CPA, you've completely missed the boat. And hopefully today we can get you out of that Kool-Aid of just destroying your business integrity and instead get you to a place where you're going to be making a lot more money and going to be a lot more successful. And again, if you think the path is low CPAs, you've completely and fundamentally misunderstood why you're running ads on Facebook to begin with.
Escúchalo en el minuto 0:21
La fatiga publicitaria es un error del operador y no una fatalidad del algoritmo
Canales de adquisiciónUn anuncio bien estructurado y con una demostración sólida de autoridad puede escalar cientos de miles o millones de dólares y mantenerse rentable durante años en un único Post ID. Atribuir la caída de rendimiento a la fatiga del anuncio suele enmascarar errores en la gestión y configuración del operador.
But if you watch that video, you can see just how beautifully written and how great of an exhibition of expertise that third ad is. And I can assure you that the number is not close to what it will be before that ad dies. That ad could easily run for years. And I think the upside for that ad is easily hundreds of thousands, if not millions of dollars. And that's coming from somebody that has regularly spent millions of dollars on single post IDs. And anybody that tells you that that's impossible is that ad fatigue is something that you need to deal with. Just go run in the other direction…
Escúchalo en el minuto 8:39
Oferta y precioEn la marca Undrafted, transformar la oferta 'compra 3 y llévate 1 gratis' en la oferta principal en todo el catálogo multiplicó por 10 el LTV y elevó la facturación de 50.000 $ al mes a más de 1.000.000 $ por semana.
I did that with – for instance, a great example of that is Undrafted, where now the buy through, get when free is on every product everywhere. Instead of being the rebuttal upsell offer, it became the hero offer. And as a result, the LTV has gone like 10x. The business went from 50,000 a month to over a million a week.
Escúchalo en el minuto 16:55
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.