Marketero · Entrevistas
Stop Wasting Ad Spend: This Tool Guarantees Accurate Customer Data
17 ideas de marketing sacadas de «Stop Wasting Ad Spend: This Tool Guarantees Accurate Customer Data», de Perpetual Traffic. Sobre todo de medición y…
- 17 ideas despiezadas
- Fuente Perpetual Traffic
- Dura 44 min
- Publicada en 2026
Ver la entrevista entera en YouTube
De qué habla
Las 17 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 17 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.
Evitar recortar presupuesto por incrementos a corto plazo en el NCAC si el tráfico nuevo crece
Medición y experimentosAl incrementar el presupuesto publicitario (ej. +17%), un aumento temporal en el NCAC por encima del objetivo (ej. superando los $100 tras subir 30%) no debe motivar una reducción inmediata si las visitas de usuarios nuevos crecieron proporcionalmente (+26%), ya que la conversión efectiva requiere considerar el tiempo de maduración o retraso de compra (time lag).
So this ties back into what John and I were discussing last week with time lag. And you can work with time lag when you have this because you have additional information. So there was a 17% increase in spend because we wanted to scale, right? So when we do that, the NCAC here, you can see it's come up 30% over a hundred or $104. We're above our target. Okay. So previously what people would do, they look at that and go, oh no, we're above the Let's scale back that 20% we just did on Meta. However, because I have this information, I can look at this metric here that is 99% accurate And I can…
Escúchalo en el minuto 39:00
Esperar el periodo de retraso de conversión (time lag) antes de evaluar el escalado de campañas
Medición y experimentosAl escalar presupuesto en canales de prospección (como Meta), el coste de adquisición de nuevos clientes (NCAC) suele incrementarse de forma inmediata en las métricas superficiales. Es fundamental conocer el tiempo de retraso de conversión (time lag) —desde la primera impresión hasta la compra final en otros canales como Google— y esperar a que madure esa ventana (por ejemplo, 14 días) antes de pausar o ajustar la…
I know the time lag on this business is two weeks. I just want to wait another week to see what happens once people move down the funnel and convert on Google with that extra time that they need, because Google's telling me that's in August when you scaled up a Meta campaign by about 16 or 17%. In that week, you saw NCAC starting to rise, but you knew because it takes about two weeks, but you know, in this particular case with this customer, it takes about 14 days or so for a client to, or a new customer for them from first impression or first click to purchase. So hit that as the next week…
Escúchalo en el minuto 40:23
El riesgo de tomar decisiones de apagado o escalado basándose únicamente en el reporte de Meta CAPI
Medición y experimentosMeta Conversions API (CAPI) ayuda al algoritmo a optimizar la entrega de anuncios, pero los reportes dentro de la plataforma siguen conteniendo datos modelados y estimaciones debido a bloqueadores y restricciones de iOS. Tomar decisiones humanas de pausar campañas basándose solo en los reportes nativos de Meta puede llevar a apagar anuncios altamente rentables cuyas conversiones reales no se reflejan en el…
And there's stuff you could be turning off because you don't think it's doing well, but it's just model data. It's not real numbers that are actually the performance of that campaign. But I think with Cappy, you might have a false sense of security because what you think you're seeing isn't actually what you're seeing. Even with Cappy installed, you still weren't getting accurate data because there's this large portion. The point is, is there is a certain portion that meta and Facebook are just never going to capture no matter what. And they're taking a guess at it. And that guess might be…
Escúchalo en el minuto 20:16
Riesgo de pausar campañas rentables por guiarse únicamente por el CPA nativo de Meta
Medición y experimentosEvaluar y pausar campañas basándose exclusivamente en el CPA o ROAS reportado dentro de Meta Ads puede llevar a descartar anuncios que en realidad cumplen el objetivo, especialmente cuando la plataforma reporta costos que parecen exceder el target por pequeños márgenes ($5 a $10) debido a compras que el pixel no detectó.
how I was speaking about earlier, where you're always going to see it lower. And DataSuite, because you have more data, you're getting those purchases that meta doesn't know they're actually in existence. But there could definitely be situations where, you know, you're $5 over the target, you're $10 over the target. And for your business, that is too much over the target and you have to cut something. But the reality is it's not over target if you were looking in a place that had all the data. And I think a lot of people can get held back by things like this.
Escúchalo en el minuto 32:05
Medición y experimentosLas decisiones de optimización y escalado deben tomarse con base en el costo de adquisición de clientes nuevos (NCAC) a nivel de campaña y anuncio, en lugar de métricas combinadas de CAC general que mezclan recompras y distorsionan la rentabilidad real de captación.
So to add to that, then what we are making those decisions on really is not this all CAC number because we're focused on new customer acquisition, as you and John have, you know, So I am looking at this North Star metric here of new customer cost or new customer acquisition cost of a current customer. It just doesn't exist in as manager. This tells me every single campaign, ad set and ad, what is the actual new customer acquisition cost for that asset? But you need to know accurately, what am I acquiring a new customer for? And that's the difference between DataSuite and everything else…
Escúchalo en el minuto 35:32
Captura de datos de atribución en el Edge frente a la API de conversiones estándar
Medición y experimentosCapturar el tráfico en servidores perimetrales (Edge) antes de que el navegador procese la solicitud evita que los bloqueadores de anuncios y restricciones de cookies eliminen las señales de seguimiento, permitiendo cruzar datos de primera parte (first-party data) en un data warehouse propio antes de retroalimentar a las plataformas publicitarias.
And then the second thing is it's going to send that back once it's fingerprinted, it matched those people up. It's going to send that back directly to the ad platforms and you have far richer data. So we're not losing any of the data because we're doing this before the browser ad blockers take the signals out of the data, essentially. So we're getting all the data at this step and it goes into the data warehouse, all gets matched together and then sent back to the platforms directly. And the edge server basically captures that user before they actually enter your store. It's coming to the…
Escúchalo en el minuto 25:30
Medición y experimentosTomar decisiones de inversión basándose exclusivamente en los datos nativos de la plataforma de Meta es arriesgado, ya que el modelado estadístico que estima compras y añadidos al carrito suele diferir significativamente de las ventas reales registradas en el backend del negocio.
The reality is these days, and I think other people are the same who are using Jire11 Data Suite is we don't really look too much at the in-platform numbers anymore because it is just a lot of nonsense. But model data is a big problem because if you are looking in-platform, basically what meta is doing is they're going, okay, we don't have oversight into what is actually happening on the website. So we're just going to try and estimate based on how many people are clicking away from our ad, how many people are, we're able to then match again later with our hashing from the conversions API…
Escúchalo en el minuto 18:48
Medición y experimentosEnviar eventos mediante la API de Conversiones (CAPI) en lugar de depender exclusivamente del píxel del navegador mejora la puntuación de coincidencia de eventos (EMQ) y alimenta el aprendizaje automático de Meta con datos de primera parte más fiables.
So you're using the pixel in this case, the pixel then becomes the third party in this whole thing. And the reason how, why API that, why Cappy works through the, through the API is because you are the first party today that you, you actually own it. And one of the things that we would always sort of check with the EMQ score inside meta specifically, we're talking about meta here, just in this specific example is the EMQ score was our indication that we were getting a better match rate on specific events with Cappy. Yeah, well, I mean, a better EMQ essentially for us is meaning we're getting…
Escúchalo en el minuto 15:26
Riesgo de quiebra al escalar tráfico pagado sin calcular el costo de adquisición permitido (NCAC)
Medición y experimentosIncluso empresas con facturaciones de 7 millones de dólares pueden escalar por coyuntura favorable o producto sin conocer cuánto pueden pagar por adquirir un cliente (NCAC). Iniciar o escalar campañas de tráfico pagado sin determinar previamente este umbral unitario puede llevar al negocio a la quiebra en un plazo de uno a tres meses.
It's a $7 million business. They don't know how much they can pay to acquire a customer. It's like, if we just started running traffic to this group without an NCAT call, we would have helped them go out of business faster. And one to three months, they would have been gone. And probably we might actually bankrupt them if we did what they told us to do. So like the example you're talking about, you know, $7 million in revenue. If you have a good product or you're in like a good period where the market's just hot, you can get quite big without knowing these numbers. And then they're now either…
Escúchalo en el minuto 6:52
Las tres palancas fundamentales para hacer crecer la facturación de un negocio
Estrategia y decisiones de negocioTodo objetivo macro de facturación se descompone y logra a través de tres únicas vías de crecimiento: adquirir nuevos clientes, aumentar el valor promedio de compra (AOV) e incrementar la frecuencia de compra a lo largo del tiempo (LTV).
So then from that goal, we sort of begin with the end in mind and then figure out, okay, how do we get there? Is it new customer acquisition? Is it getting customers to buy more when they buy, meaning increase their average order value? Get them to buy more often, maybe increasing LTV? Those are basically the three ways in which you can grow a business. And so the combination of those three are through the MPIs. The MPIs are just a means to measure how to get to the big goal.
Escúchalo en el minuto 3:31
Discrepancia del 20% al 25% en el CAC entre Meta Ads y la atribución first-party
Medición y experimentosEl costo de adquisición global (All CAC) reportado de forma nativa en Meta Ads puede diferir entre un 20% y un 25% (mostrando costos más altos por compra no registrada) frente a plataformas de atribución conectadas a CRM o Shopify, incluso tras esperar dos semanas para que los modelos de Meta terminen de rellenar datos.
at all CAC, which is the same metric being compared, we have $77.56. So you've gone $95.64, $77.56. The point is, is that the all CAC, the ACAC, which is the entire cost of acquiring a customer, is about 20%. There's a 20 or 25% difference between what you're looking at inside meta and what you're actually seeing inside the interface, Wicked Reports and Data Suite, which is not insignificant. And yeah, it's 20% difference. You're actually giving it a two week grace period to give meta even more of a chance to backfill. But now it's two weeks later.
Escúchalo en el minuto 33:09
Funcionamiento y limitaciones de la API de conversiones (CAPI)
Medición y experimentosLa API de conversiones (CAPI) mitiga las restricciones de rastreo y bloqueadores de navegador enviando datos de primera parte (nombres, correos hasheados) directamente desde el servidor web a la plataforma publicitaria. Sin embargo, su tasa de emparejamiento es imperfecta y pierde datos significativos cuando los usuarios emplean correos o identidades diferentes entre su cuenta de compra y su perfil en la plataforma…
So conversions API, the thing that we told everyone to install to enrich their data basically is a way of getting around app blockers and browser blockers. Things that are basically removing your data from coming back to the ad platform, specifically meta for conversions API. What was happening before when iOS 14 came into play was that click ID was being removed by the browser and meta was losing any awareness of who that person was. So what they did is they brought into play conversions API. And what that would do is you're still losing that click ID, but the server that your website is…
Escúchalo en el minuto 11:54
Optimizar y escalar campañas mediante NCAC en lugar de métricas superficiales de plataforma
Medición y experimentosAl escalar campañas de adquisición de forma agresiva, es común que las métricas superficiales de la plataforma publicitaria empeoren; por ello, la toma de decisiones debe anclarse a indicadores clave de negocio como el costo de adquisición de nuevos clientes (NCAC).
The industry has been focused on in-platform metrics for so long now and even just front-end metrics. So people are coming on to NCAC and stuff now, but it's still an education. They understand exactly, well, why are we actually making you go to an NCAC target? This is why it's good. And this is why when we scale aggressively, sometimes some of your other metrics are going to come down. But it always comes back to those MPIs.
Escúchalo en el minuto 4:12
El costo de adquisición de nuevos clientes (NCAC) requiere datos exactos fuera de la plataforma para escalar
Medición y experimentosAl planificar el crecimiento de ingresos (como duplicar la facturación de 6M a 12M), la adquisición de nuevos clientes suele ser la palanca crítica, haciendo que el NCAC (New Customer Acquisition Cost) sea la métrica clave de control; por ello, este dato no puede depender de las métricas aproximadas o modeladas de las aplicaciones publicitarias nativas, sino de una fuente de verdad precisa.
Like I said before, if you are, if your goal is to go from 6 million to 12 million in revenue, okay, it's a $6 million gap. How are you, is it all new customer acquisition? Chances are probably a new customer acquisition is a very important part of that. And that's the reason why NCAC or new customer acquisition cost is such a vital stat, such a vital goal. Like you as a media buyer, you're looking at that all the time and measuring against it, benchmarking it. But that data, that number has to be as accurate as possible. And if you're relying on in-app, it might be right, might not be right…
Escúchalo en el minuto 22:40
Medición y experimentosLos informes dentro de las plataformas publicitarias como Meta utilizan modelos y estimaciones sobre el comportamiento del usuario que a menudo difieren de las transacciones reales registradas en la tienda (como Shopify), por lo que optimizar basándose únicamente en métricas in-platform genera decisiones financieras erróneas.
There is a lot of money on the line here, and you need to be able to make accurate decisions based upon true data. Model data is a big problem because if you are looking in platform, basically what Meta is doing is we're just going to try and estimate based on how many people are clicking away from our ad, how many people are purchasing, what percentage of those people are taking what action. This looks like it is very accurate. You've had all these sales and you can look at your Shopify dashboard and you're not getting anything.
Escúchalo en el minuto 0:00
Medición y experimentosLas métricas dentro de las plataformas publicitarias y los modelos algorítmicos de atribución de terceros (como Google Analytics o CAPI estándar) están cada vez más modelados y distorsionados por restricciones de privacidad, por lo que las decisiones estratégicas de crecimiento deben basarse en datos directos y métricas de impacto puro del negocio.
And we found that, yeah, the in-app metrics used to be the real judge and jury, like the source of truth years ago. That is no longer the case because there are a lot of things that now block that from being the source of truth, which we'll get into here in just a second. So you need a data solution in order to be able to read the tea leaves to get the right metrics to move the business in the right way. Otherwise, you know, if you're just using Google Analytics or if you're just using Cappy, and we'll talk about that here today, that data might not be as pure as it possibly could be. And the…
Escúchalo en el minuto 5:16
Medición y experimentosLas restricciones de privacidad aplicadas a usuarios de dispositivos Apple provocaron caídas repentinas de entre el 60% y el 70% en el reporte de conversiones dentro de las cuentas publicitarias de Facebook.
And especially for, I remember for some of our, I wouldn't even say like higher end, but iPhone users, Apple users tend to be, you know, higher end buyers. And I remember one account in particular, we lost 60 to 70% of our conversions literally overnight.
Escúchalo en el minuto 17:33
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.