35 ideas de marketing sacadas de «A BETTER Way To Run Facebook Ads In 2025 (it's simple)», de Professor Charley T. Sobre todo de medición y experimentos…
Ver la entrevista entera en YouTube
Las 35 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.
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.
El beneficio por transacción (GPT) se calcula restando el costo por adquisición al valor promedio del pedido (GPT = AOV - CPA), permitiendo determinar si cada conversión individual genera ganancia o pérdida neta directa.
know is GPT. GPT is AOV minus CPA. This is how profitable are those transactions.
El costo por mil cuentas alcanzadas combina el CPM y la frecuencia para reflejar el costo real de impactar a usuarios únicos (Costo por 1.000 cuentas = CPM × Frecuencia). Si el CPM es de $20 y la frecuencia es 2, el costo real para alcanzar a mil personas únicas asciende a $40.
cost per,000 account centers reached. cost per thousand account centers reach is because this is essentially combining CPM and frequency. It's saying how much does it cost to actually reach a new person. This is essentially cost per If your frequency is a two, then you really only reach 500 people. if your CPM is 20 and your frequency is a two, your cost per thousand account centers will be 40.
El GPT (Gross Profit per Transaction) se calcula restando el Costo de Adquisición (CPA) al Valor Promedio del Pedido (AOV): GPT = AOV - CPA.
creating these metrics GPT is AOV minus CPA. Okay, average order value. Now you can
Si la tasa de conversión está por debajo del 1%, la página de destino tiene problemas graves; si supera el 3%, el presupuesto en anuncios es insuficiente para el potencial del producto; y si con una conversión alta no se puede costear más publicidad, el precio del producto está fijado demasiado bajo.
here's the metrics you need to understand. If your conversion rate is below 1%, your landing page is If your conversion rate is over 3%, you are not spending enough money on ads. If you can't afford to spend more money on ads, your price is too low. Done.
Optimiza el gasto publicitario desactivando los anuncios cuyo CPA esté por encima del promedio y cuyo beneficio por venta (GPT) esté por debajo, concentrando el presupuesto exclusivamente en creatividades con 'unit economics' saludables.
Turn off the ads where the cost per purchase is above average and the GPT is below. For instance, let me give you an by cutting out all the high CPA and low GPT ads. We only showed the ads that were actually good on a unit economic basis that the machine actually wanted to show. So again, basically half the ad
Métricas como hook rate, hold rate, CTR y CPC no indican rentabilidad directa; optimizar creativos para retener atención superficial entrena al algoritmo a buscar audiencias interesadas en el mensaje pero sin disposición de compra.
have any sales? No, they don't matter. Hook rate, hold rate, rorowaz, CTR, CPC, none of those things actually tell you if that ad is making you any money. And I'd argue that ads with a really good hook rate and a really good hold rate, with a really good CTR or really good C, the ads that are really, really good at keeping the attention of people that have no interest in buying from you are the reason the rest of your business is struggling because you're training the machine to go after people who are just really interested in what you have to say, but not interested in what you have to…
Una página de destino con alta tasa de rebote o baja conversión eleva automáticamente los CPMs en Meta, ya que la plataforma penaliza la experiencia negativa de usuario cobrando un inventario publicitario más caro.
Do you think CPM is affected by landing page? Absolutely. The landing page really matters. Um, and impact. It's do people bounce? If your landing page is if your landing page has a high bounce rate and or a low conversion rate, the machine will just charge you higher CPMs because getting some tricking somebody into clicking on your ad gives them a bad experience. but you're going to pay way more than everybody else to do it."
Un anuncio con un CPA aparentemente más bajo puede ser perjudicial si solo atrae a clientes existentes. En lugar de usar listas complejas de exclusión, conviene apagar los anuncios cuyo volumen de compradores nuevos sea mínimo y priorizar aquellos que realmente traen clientes netos, aunque tengan un CPA nominal más alto.
Why that matters is if some ads are really good at the ad level, we can see, oh, this ad is a CPA that's 10% less. Yeah, but only 5% of the people to buy are net new. Let's say an ad is coming in at 50 bucks where 90% of them are new. And we have another ad coming in at 45, but only 5% of them are new. Which one of those ads is going to get you more revenue? I'll give you a hint. it's not the one selling to repeat customers at a lower price. And you could say, well, I'm going to do exclusions to make sure that existing customers don't see my ads. Or you could just turn off the ads that appeal…
Evaluar la rentabilidad publicitaria únicamente mediante el ROAS porcentual oculta la escala real del negocio, ya que un porcentaje alto puede representar ingresos marginales, mientras que la métrica clave de decisión debe ser el beneficio neto absoluto (valor total de conversiones menos gasto publicitario).
That's how we're figuring this math. But your profit volume is purchase conversion value minus spend. Simple. If you don't know this number, how do you have any idea what's profitable or not? Well, you're looking at the row as of an ad. What's the scale of that? A percentage doesn't tell you an absolute value. Do you want the highest row as ads or do you want to make more money?
Evaluar la rentabilidad publicitaria únicamente mediante el ROAS porcentual oculta la escala real del negocio, ya que un porcentaje alto puede representar ingresos marginales, mientras que la métrica clave de decisión debe ser el beneficio neto absoluto (valor total de conversiones menos gasto publicitario).
That's how we're figuring this math. But your profit volume is purchase conversion value minus spend. Simple. If you don't know this number, how do you have any idea what's profitable or not? Well, you're looking at the row as of an ad. What's the scale of that? A percentage doesn't tell you an absolute value. Do you want the highest row as ads or do you want to make more money?
Para optimizar la rentabilidad de una cuenta publicitaria se debe calcular el beneficio bruto por transacción (AOV menos CPA), en lugar de confiar ciegamente en métricas fuera de contexto como el ROAS o el CPA aislado.
cost per purchase. GPT is a custom metric where you do AOV minus CPA. That is your profitability per sale. If you don't know that number, literally nothing else matters. If you want to optimize your profitability of your ad account, don't focus on things with a higher rorowaz or lower CPA. Those things are out of context data.
En plataformas de paid social, la frecuencia publicitaria indica la posición del usuario en el embudo: una mayor frecuencia significa que el algoritmo está impactando a audiencias de parte baja del embudo (inventario premium y más costoso), mientras que una menor frecuencia refleja impacto en audiencias de la parte superior.
Frequency is what is the audience that If the frequency goes up, you are reaching a lower funnel audience. You are hitting If the frequency goes down, you're hitting people further up the funnel. Ads with the higher the frequency, the frequency is higher, when you're reaching people who are lower in the funnel, that is premium inventory. the CPM there will be higher because it is premium inventory.
El presupuesto publicitario asignado a un anuncio es una métrica de rendimiento que el algoritmo otorga en función de su eficacia. Intentar forzar manualmente el gasto en anuncios específicos es microgestionar un algoritmo más inteligente y reduce los resultados potenciales de la cuenta.
Remember, ad spend is earned. Ad spend is a performance metric. If you are trying to control where your ad budget goes, you are trying to micromanage a machine that is smarter than you. And that is half the reason that you're never going to be nearly as successful as possible. spend is earned. If you're trying to
Optimizar creatividades basándose en hook rate o hold rate entrena al algoritmo publicitario a mostrar anuncios a personas que consumen contenido pero nunca compran, empeorando el rendimiento del embudo.
metrics like hook rate? No. Do I care how well a creative grabs the attention of somebody who doesn't buy? Not at all. Do I care about hold rate? Do I really care that people who are never going to buy like to see all of my ads and that teaches the machine to show my ads to people who are never going to buy? No. No, I don't.
El presupuesto publicitario asignado a un anuncio es una métrica de rendimiento que el algoritmo otorga en función de su eficacia. Intentar forzar manualmente el gasto en anuncios específicos es microgestionar un algoritmo más inteligente y reduce los resultados potenciales de la cuenta.
Remember, ad spend is earned. Ad spend is a performance metric. If you are trying to control where your ad budget goes, you are trying to micromanage a machine that is smarter than you. And that is half the reason that you're never going to be nearly as successful as possible. spend is earned. If you're trying to
Un porcentaje de clics (CTR) elevado suele producirse artificialmente porque el denominador (impresiones) es muy bajo debido a un CPM alto, o porque se optimiza hacia clics baratos trayendo tráfico de baja calidad. Reducir el CPM amplía el alcance y puede generar más volumen absoluto de clics y ventas netas, aunque el porcentaje de CTR disminuya.
if for the same money I reached 2,000 people. So, instead of a $20 CPM, it's a 10. And out of that, instead of getting 10 clicks, I only got I I got 15. And out of that 15, let's say I got three sales. Now, it's my CTR looks a lot worse because it's not a,000 divided by 10. It's 2,000 divided by 15. That's That's worse, right? 10% or 1% would have been 20, right? So now our CT CTR has gone down, but we made three sales. One, because we got 50% more traffic, and two, because our ads are better. How do we know that? The CPM. The CPM is a understand um the reason your CTR is high is because the…
En campañas con optimización de presupuesto a nivel de campaña (CBO), lanzar constantemente nuevos ad sets suele resultar en que el presupuesto permanezca en los anuncios originales, ya que el algoritmo prioriza el histórico y aprendizaje de conversión que ya posee sobre las nuevas pruebas.
cheaper. Easy. I launch CBO campaign with three adset every 3 days that create new adset, but the initial adset takes 99% of spend. First off, yeah, that makes a lot of sense. Why would the new ads get any spent? The machine already knows how the other ads work. throwing a ton of spaghetti at the wall doesn't work because the machine doesn't like any of the new ideas because it already knows what you're doing someplace else.
La primera compra suele ser la menos rentable o incluso asumir un margen negativo en publicidad, pero adquirir clientes netamente nuevos construye el valor del negocio a largo plazo porque las compras recurrentes posteriores tienen un costo de adquisición mucho menor y un ticket promedio (AOV) más alto.
campaign that's been 104 is 90% net new customers. We lost $12 on the sales. If you look at the campaign right below that, it's only 60% net new customers. And we made $57 profit on every sale. We were hyper profitable because the first ad is the first sale is the least profitable. Every sale after that is worth way more money. If you notice, the AOV on the more profitable is is substantially higher. The CPA is lower. more time. When you buy a new customer, their value to the business is far greater than getting somebody who already bought to buy again. So when you combine these two…
Replicar el volumen masivo de creatividades o las tácticas de marcas multimillonarias es un error grave para empresas pequeñas o medianas, ya que las grandes corporaciones resuelven problemas de saturación de escala y dinámicas operativas internas que no aplican a negocios de menor tamaño.
But if you're copying how a nine figure business does work, understand that you're chasing the wrong thing. Don't You don't have the problems they have. Club. They are trying to solve problems 10 times bigger than your entire business. And they're trying to do it in a way that also makes sure everybody in the business keeps their job. their needs are very different than yours. Don't follow their strategies. It's a great way to destroy your business.
Métricas habituales de plataformas publicitarias como ROAS, CPC, CTR y hook rate a menudo no tienen correlación directa con el dinero neto en la cuenta bancaria; optimizar campañas persiguiendo exclusivamente estos indicadores intermedios puede destruir la rentabilidad del negocio.
to your Facebook ads. Your dashboard is crap. The metrics you've been taught to ROAZ doesn't mean anything. Facebook ads. CPC completely irrelevant. matter. CTR completely irrelevant if rate has nothing to do with bank account. ROAZ has nothing to do with bank account. CPC has nothing to do with bank account. CTR has nothing to do with bank account. In my world, if the information isn't actionable, if it doesn't help me save time or make more money, it's basically and why chasing down all of the metrics you've been told to care about destroyed their profitability.
Un costo por clic (CPC) o costo por lead (CPL) muy bajo suele reflejar usuarios de baja calidad sin intención real de compra; optimizar anuncios únicamente para abaratar el costo unitario de tráfico atrae grandes volúmenes de personas que nunca convertirán.
the cost per click has nothing to do with the level of intent or interest or decent people. It's like leads. The lower cost the lead, the shittier it is. The lower cost to click, the shittier it are good. It just means that measuring the cost per traffic or a or or a qualified user in some way is a terrible way of running a business because none of that information is actionable. Remember the actionability. If this low CPC is the success, then only run ads with a low CPC. you're just going to have a super high volume of people who will never buy and run out of money.
Al reducir la inversión en anuncios no incrementales de 13.000 $ a 7.000 $, el CPM disminuyó un 33%, el costo por compra cayó de 150 $ a 100 $ y el beneficio neto total se multiplicó de 1.300 $ a 5.400 $, aumentando la ganancia por venta de 15 $ a 75 $.
75% of the cost. So 25% cheaper on total cost per reach, 33% cheaper on CPM. What happened? Our cost per purchase went from $150 to 100. The GPT went from $15 to75. Let me explain this again. Okay. We cut the ad spend from $13,000 to 7 by getting rid of the ads that people and by pulling back the budget where it wasn't incremental. As a result, we dramatically cut the CPMs and our cost per purchase. And as a result of doing that, our profitability per sale went from $15 to 75. What does that mean? It means every sale was five times more profitable. from 1300 at nearly 13,000 in spend they…
Los anuncios nuevos suelen entregarse inicialmente en el fondo del embudo donde el ROAS parece elevado pero la frecuencia y los CPMs son altos; al intentar escalarlos se saturan rápidamente porque la audiencia disponible con esa intención es limitada.
you can't scale into an audience that is already limited by frequency and understand the interplay of frequency to be low. And as soon as you start scaling up, the frequency and CPM and skyrockets and the ad fatigues, quote unquote. And it's only because that ad was only effective at a small budget or happens because every time you launch a new ad, it launches at the bottom of the funnel because the machine wants to see how
Suprimir los periodos de prueba gratuita (free trials) en programas de formación o comunidades de pago filtra a los usuarios oportunistas que entran solo a copiar material, mejorando significativamente el nivel de compromiso y la calidad general de los miembros.
took away the 7-day free trial uh because we had a lot of people coming in basically to steal all the information and then leave. And to be fair, I don't want people that don't take it seriously. The quality of the community has grown 10x since making it not a 7-day free trial. So, all of that we we
Implementar seguimiento específico de eventos de nuevos clientes mediante herramientas CAPI (server-side) permite identificar y eliminar los anuncios que no traen compradores netos nuevos y solo consumen presupuesto en clientes existentes.
can't have new customer percentage if you don't have a new customer event. If you want a new customer event, my number best conversion side or server side conversion tool uh cappy tool on the and you can get to understanding how much did it cost to bring in a new customer. How many of my ads brought in We attached this and began to immediately just eliminate the ads that didn't do a good job at bringing in new
La verdadera diversidad creativa en publicidad no consiste en producir decenas de variaciones menores del mismo anuncio ganador, sino en construir grupos de creatividades con ángulos y conceptos completamente independientes y diferenciados entre sí.
volume, spinning up 100 ads, but that's not how you roll. Why are they wrong? And what's the real deal when it comes coming out really soon. But diversity just means don't have 12 ads, five ads, three ads. You don't need to cover that same number. You don't need to hit that concept that wins over and over again. tremendous way to have five different pools of creative content that are completely diverse and how they
El ROAS basado en atribución de último contacto asigna erróneamente todo el mérito de la venta al último anuncio interactuado. Los ingresos de hoy no son causados exclusivamente por el gasto de hoy ni el gasto de hoy impacta solo en hoy: el viaje del cliente es prolongado y se apoya en exposiciones previas, demanda acumulada y otros canales.
let's break down why return on ad spend The concept of return on ad spend is built around attribution. Basically saying whoever was the last to touch something takes credit for all The bottom line is the last touch is But the last touch isn't the reason something happened. And to prove that, if you turn all of your ads off, you're going to get sales tomorrow and tomorrow's rorowaz is essentially going because people bought on ads they saw today and yesterday. And that ad they saw today and yesterday is not the first fair, the customer journey isn't days or weeks long. It's a lifetime. The…
Lanzar creatividades de venta dura y directa incrementa los CPMs por rechazo de la audiencia y agota rápidamente el público listo para comprar, forzando a los anunciantes a una espiral insostenible de producir cada vez más anuncios de baja calidad para perseguir conversiones residuales.
absolutely. Also, your CPM will be high if people don't like your ad. If your ad is a hardcore like buy now from me because it's somebody that's really good at email or search running your ads, you're going to be just pay higher CPMs. bad over and over again. And the more ads you launch, they're bad, the faster it's going to depreciate. And the only reason your new ads work is going to launch at the bottom of the funnel. So the only way you're successful is by launching more and more shitty ads. And the only way you stay ahead of the depreciation wave is by launching more and more ads. you…
El tráfico proveniente de motores de búsqueda posee una intención implícita de compra y preparación para adquirir un producto, mientras que el tráfico de redes sociales funciona como una valla publicitaria que interrumpe la navegación pasiva con niveles de intención sustancialmente menores.
in search there is an implied intent and sentiment generally don't search for things products and then click on the ad in search because they have no interest. aware. They do that because they're ready. They're looking they have their credit card out looking to shop. Social media isn't like that. Remember that is somebody going into the store trying to find something. Social media is a billboard in the sky blocking the sunset. Those are two very different levels of intent.
Tolerar un margen bruto de beneficio negativo o nulo en el primer día en campañas de alto volumen es viable y estratégico siempre que atraigan clientes netos nuevos que alimenten el valor de vida del cliente (LTV) y la recompra futura.
is well over a 90. No other campaign in volume of sales and how many of them are net new. Yes, the GPT on this is these are overwhelmingly net new individuals, meaning this campaign is buying future cash flow that will scale the business. If you're only worried about making a profit today, what you have, you are a salesperson, not somebody who runs a business. You need to understand the profitability that you on and the acquisition of better customers that come back and buy again
Tolerar un margen bruto de beneficio negativo o nulo en el primer día en campañas de alto volumen es viable y estratégico siempre que atraigan clientes netos nuevos que alimenten el valor de vida del cliente (LTV) y la recompra futura.
is well over a 90. No other campaign in volume of sales and how many of them are net new. Yes, the GPT on this is these are overwhelmingly net new individuals, meaning this campaign is buying future cash flow that will scale the business. If you're only worried about making a profit today, what you have, you are a salesperson, not somebody who runs a business. You need to understand the profitability that you on and the acquisition of better customers that come back and buy again
Un anuncio no crea el interés profundo en un producto ni fabrica una necesidad en el usuario; su función es capturar y ganar la atención en redes sociales de personas que ya cuentan con un contexto de vida, identidad o disposición previa compatible con la oferta.
over 20 years. Like my entire life has brought me to the point where this shirt The point to this is your ad is not the reason somebody's interested in the product. Your ad is not the reason somebody discovers that there's a solution. Your ad is not the reason somebody's interested in buying. Your ad does nothing other than earn somebody's attention while they are in the process of enjoying social media. Your ad is a
Las páginas de aterrizaje excesivamente estéticas a menudo convierten peor que páginas sencillas o poco estilizadas cuya estructura se enfoca puramente en la claridad de la narrativa de ventas y la alineación con el anuncio.
they're getting more." Look, shitty pages sell. All right. Really sexy pages are usually the reason that you can't Sounds to me like you need to really ramp up the quality of your landing page and tell the story and then just get ads designed for that really great sales
Aproximadamente el 20% de los usuarios que compran tras hacer clic en un anuncio adquieren un producto diferente al que se mostraba en la creatividad. La función principal del anuncio es captar la atención hacia la marca o tienda, no cerrar únicamente la venta del SKU específico anunciado.
other products on my store, odds are at max 80% of the people that buy are buying this thing. Meaning 20% of the people that click on that ad are buying something else. You're like, why would somebody click on the ad and buy something else? My response is, have you ever gone anywhere and spent money on anything that wasn't promoted? The ad gets your attention. That's all the ad is. Okay.
El fenómeno de la fatiga del anuncio ('ad fatigue') utilizado para justificar caídas en el rendimiento de ventas es en realidad una falla de configuración y gestión estratégica del comprador de medios (operator error) y no un agotamiento inevitable de la audiencia.
metrics. Okay. Ad fatigue. Our media buyer brought it up since our sales decrease, but I call BS on ad fatigue. Yeah, ad fatigue is operator error. Um,
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.