Marketero · Entrevistas
Enemies No More: Using Google and Meta Together for 10x Returns Part 2 | Perpetual Traffic EP 564
26 ideas de marketing sacadas de «Enemies No More: Using Google and Meta Together for 10x Returns Part 2 | Perpetual Traffic EP 564», de Perpetual…
- 26 ideas despiezadas
- Fuente Perpetual Traffic
- Dura 57 min
- Publicada en 2024
Ver la entrevista entera en YouTube
De qué habla
Las 26 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 26 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.
La casilla de 'solo clientes nuevos' en Google Performance Max puede incluir hasta un 80% de clientes existentes
Medición y experimentosDebido a que las tasas de coincidencia de datos (match rates) de listas de clientes en Google rondan por debajo del 20% (12-17%), la función de Performance Max configurada para dirigirse exclusivamente a nuevos clientes termina impactando a un 80% de clientes recurrentes o existentes que la plataforma no logra identificar.
interesting so performance Max has a new customer only check box if you check that box on average pmax takes 80% of your old customers and new customers now continuity numbers pmax takes 80% old customers as new customers okay look at the last two examples I showed you 12% match 17% match so 80% new those are the Google I only want new customers 80% of the people are your actually older existing customers that's because you're getting somewhere sub 20% match it's
Escúchalo en el minuto 31:25
Combinación publicitaria de alto rendimiento: Google Standard Shopping con Meta Advantage+
Canales de adquisiciónEmparejar campañas de Google Standard Shopping con campañas Advantage+ en Meta genera una sinergia probada de captación y conversión para la adquisición de clientes.
to want this these are campaign pairings that we know work really well Google standard shopping and met meta Advantage Plus this is a wrer downer people phenomenal pairing yeah this is the hack
Escúchalo en el minuto 39:40
Estrategia 'Full-Funnel' tridimensional: construir demanda en lugar de solo capturarla
Canales de adquisiciónLimitar el presupuesto al tráfico con intención inmediata de compra agota rápidamente las audiencias. Un crecimiento sostenible exige operar en tres ejes simultáneos: mensaje relevante en todo el embudo (eje Y), distribución multicanal (eje X) y ejecución continuada a largo plazo (eje Z) para crear y nutrir demanda fría.
that's your y- axis then you have to run across all applicable channels that's x-axis then you have to run for an extended period of time that's your z-axis and to be frank that's the answer you take a very relevant message put it in front of a very relevant audience across the entire funnel across all applicable channels across a protracted see but if you only spend money on that traffic can be likened to corn and traffic that buys is popcorn and all of us every business in every industry has decided to just start paying for instant pot popcorn because that's the easiest and the fastest but…
Escúchalo en el minuto 35:06
Medición y experimentosPara tomar decisiones fiables sin depender de las métricas infladas de las plataformas, se debe contrastar la información usando al menos tres fuentes: datos in-app (Google y Meta Ads), analítica de tráfico independiente (como GA4) y datos de origen 'first-party' (Shopify o CRM), apoyándose opcionalmente en herramientas de atribución externa.
have to establish at least three measurement sources preferably more so that would be inapp take what Google and Facebook are telling you know that it's a lie but you still need to look at it G4 is surprisingly adapt at broader yeah and then whatever is first party for you so Shopify your CRM those are your three sources uh extra credit if you're using attribution attribution software Wicked reports or North beam those are the two that we
Escúchalo en el minuto 36:30
El peligro de apagar campañas de prospección basadas en métricas directas de conversión
Canales de adquisiciónDesactivar campañas de Meta Ads que aparentan ser perdedoras según la atribución directa (en este caso, 6.000 $ gastados para solo 0,9 transacciones) puede colapsar el rendimiento de Google Ads si esas campañas son responsables de alimentar la parte alta del embudo (generaban el 76% del tráfico nuevo).
interesting this is a tier 11 meta campaign that shows a losing Facebook campaign so we're dealing with one of your clients I'm the Google agency you're spending six grand you're you're you've got 0.9 transactions this isn't transactions you spent $6,000 and you got not even one full trans action so your client turned this campaign off oh this campaign this losing campaign when they turned it off it tanked everything all Google campaigns instantly just plummet and what we go to find out is the losing Facebook campaign was responsible for 76% of all new traffic wow so this responsible this one…
Escúchalo en el minuto 8:29
Medición y experimentosLo que las plataformas y analistas llaman atribución de 'primer toque' (first touch) es en realidad el primer contacto que la tecnología puede rastrear. En la práctica, suele ocurrir tras cientos de impactos previos no medibles y situarse muy cerca del final del ciclo de compra, por lo que categorizar campañas como pura adquisición basándose en esta métrica induce a error.
and last touch is such a misnomer because Google can't see the first touch this should say first touch that we can track and last touch that we can track touch it's the first touch we can track profound first touch we can track which by the way is probably one of the very last touches in their sales cycle prob the first touch we can track is probably their 498th touch it's the second to attribution but that's the thing is all these data driven marketers are like oh first touch attribution and this is clearly acquisition campaign no it isn't
Escúchalo en el minuto 12:53
El reconocimiento de marca previo impulsa la captación en servicios de necesidad urgente
Marca y posicionamientoInvertir en campañas de alcance y notoriedad de marca en la parte alta del embudo mejora la eficacia de las campañas de último clic, incluso en sectores de respuesta inmediata o emergencias (como fontaneros o electricistas), ya que los consumidores recurren por búsqueda directa a marcas que reconocen y en las que confían cuando surge el problema.
his brand awareness campaigns improve the efficacy of his last click emergency plumbing so I don't care what industry you're in I don't care how last click you think your [ __ ] is you need to run top of funnel you need to run full funnel you need to run multi Chan so like you think of those guys and you go do a brand search spend so much on all this brand stuff because they are exactly like what you're talking about like when you need them you need them in a pinch you need them now and they're a name that you recognize and they're a name you trust cuz name you trust consumers aren't that the…
Escúchalo en el minuto 50:10
Baja tasa de coincidencia en audiencias de Google a partir de tráfico y compradores reales
Medición y experimentosGoogle Ads puede llegar a registrar y emparejar únicamente entre el 12,5% y el 17% del volumen real de usuarios en su creador de audiencias (tanto para visitantes web generales como para compradores con first-party data), lo que reduce drásticamente la cobertura esperada en campañas de retargeting o listas de exclusión.
the show two million viewers in 60 days we said Hey Google create an audience for us of of all those users Google matches 321,000 out of 2 million for those of you that are bad at math that's 17% I know because I used a calculator Google is able to match 177% now you deeper uh 654 th000 website purchases in a one-year period 654 th000 this is now this is last click converted users with first-party data that landed on my thank you page 654 th000 customers paid me money gave me names addresses logged in blah 654 th000 in a 365 day period zoom out to 540 days with Google and say create me an…
Escúchalo en el minuto 18:21
El sesgo de la atribución por clic en plataformas de consumo visual rápido como TikTok
Medición y experimentosEn plataformas de vídeo corto como TikTok los usuarios consumen a gran velocidad y casi no hacen clics, por lo que los modelos de atribución basados únicamente en clics muestran un rendimiento falso (por ejemplo, un MER de 0,65 sobre 60.000 $ de gasto). Al apagar el canal guiándose por esa métrica deficiente, la adquisición total del negocio se desploma al ser el motor real de la demanda.
clicks on Tik Tok yeah well if you're looking at Tik Tok from an attribution and the attribution model of course is Click only because there's no way to never be able to attribute Impressions there's no such thing as a click stopping is the equivalent of a click in Tik Tok it's like oh I stopped and I saw information right so here's the media efficiency ratio of a Tik Tok ad being attributed at 0 65 $60,000 spend with the myrr of 65 and yet when you turn off Tik Tok the whole thing goes to hell in a handbasket that client's entire acquisition was Tik Tok here's another
Escúchalo en el minuto 10:23
Invertir en la parte alta del embudo reduce el costo por adquisición combinado (Blended CPA)
Canales de adquisiciónEl tráfico en la parte superior del embudo (Top of Funnel) no solo es más económico de adquirir, sino que incrementa la eficiencia global del marketing, reduciendo el CPA combinado (Blended CPA) y abaratando las conversiones en todas las etapas inferiores del embudo.
about this for years is the cheapest traffic dude well not just that not only is it the cheapest but it makes everything else throughout your funnel cheaper work yes it lowers your overall Blended we did this branding we 10x our brand spend on on Facebook and we lowered the CPA like a 100 like 50% and then another 50% the more we did at top of funnel so it makes everything cheaper not just the traffic that you're buying at the top which is non-conversion traffic so um
Escúchalo en el minuto 53:07
Impacto del reconocimiento de marca previo en servicios de compra impulsada por emergencias
Marca y posicionamientoIncluso en mercados de respuesta urgente y ciclo de compra inmediato (como servicios de fontanería de emergencia), la exposición previa a campañas de notoriedad de marca y elementos visuales incrementa la probabilidad de que el cliente elija hacer clic en tu negocio en el momento crítico.
that's true generally but I'm in emergency plumbing and in emergency plumbing if somebody goes I need a plumber right now today I know for a fact that's last click and it's short-term and it's not a bu long buying model here's what's crazy about it I know somebody who runs an agency that only Services Home Service mostly Plumbing when he runs brand awareness campaigns and lets and puts people in front of here's the logo here's the brand here's the URL here's the phone number here are the colors and then when they do have you know hey there's feces spewing out of my toilet even in that urgent…
Escúchalo en el minuto 49:33
Ejecutar campañas sin objetivo de conversión directo para adquirir tráfico a menor coste
Canales de adquisiciónLanzar campañas en Meta o Google optimizadas para interacción o reproducciones de video sin rastreo de conversión activo evita que los algoritmos encarezcan y 'comoditicen' el tráfico al detectar compradores; el propio contenido del anuncio se encarga de cualificar a la audiencia manteniendo los costes publicitarios bajos.
Strokes are run non-conversion campaigns and here I am now telling people run campaigns without conversion tracking but here's why you don't want meta and Google to know when somebody converts because then they start to optimize for it and when they start to optimize for it they start to commoditize your traffic and you pay more so if you're running a relevant message to a relevant audience but they can't see who's converting you'll notice that your costs actually stay static you've got copy or ads that intrinsically qualify an audience then run no audience targeting and just let meta and…
Escúchalo en el minuto 41:50
Sinergia cruzada entre anuncios de Meta y conversiones por visualización en YouTube
Medición y experimentosPausar las campañas en Meta puede reducir a la mitad las conversiones por visualización interesada (engaged views) en YouTube Ads, evidenciando que los canales publicitarios operan de forma interconectada y que la atribución aislada o de último clic distorsiona el impacto real entre plataformas.
the week that it was disapproved to the two months all of October of November and engage view conversions make up twice the total conversions so that have 360 conversions 140 or clicks 220 or engaged views and what we realized was meta's ads were cooperating and cooperating with YouTube and dude that are meta the minute meta we were getting double the conversions from click to engage view engage view is doubling the conversions that clicks is getting meta gets shut off and then engage views worth
Escúchalo en el minuto 15:25
Medición y experimentosUna campaña publicitaria en plataformas como Meta puede no mostrar conversiones directas dentro de la plataforma y aun así ser de alto valor si aporta la gran mayoría del tráfico nuevo que alimenta el resto de las campañas del embudo.
if you're trying to figure out which campaigns we feeding which campaigns thoughtful person goes okay I'm not seeing the conversions from meta but what am I seeing 76% of my new traffic is coming from this one meta campaign so even though meta is not showing the conversion that to me is a high value
Escúchalo en el minuto 0:00
Evaluación del rendimiento mediante modelos de caja agregada (MER/MMM) a largo plazo
Medición y experimentosEl éxito de marketing debe evaluarse mediante modelos globales de dinero invertido versus dinero ingresado (MER, MMM o ROI global) observados en ventanas temporales amplias, evitando juzgar la rentabilidad únicamente en ciclos artificiales de 30 días.
last point about what to do is measure success using cash in cash out models over time this is Media efficiency ratio media mix modeling or or return on investment myrr mmm or Roi here's what's consideration Opex like shut up it's how much I put money in the magic box and then money came out of the magic box and you should be tracking and you should be tracking over time this is the maybe the again marketers have the bad habit of looking at things in 30-day cycles of course there's no business in the world that I can think of that operates in clean 30-day Cycles we have to look over time and…
Escúchalo en el minuto 38:37
Uso de métricas tempranas de interacción para evaluar campañas de adquisición no atribuibles
Medición y experimentosAunque métricas como reproducciones, clics o CTR no generan ingresos directamente, sirven para diagnosticar la salud de campañas iniciales de adquisición cuyo impacto comercial real no queda registrado de forma inmediata en las plataformas publicitarias.
that I shared in the presentation 76% of my new traffic is coming from this one meta campaign so even though meta is not showing the conversion that to me is a high value campaign right am I seeing views am I seeing engagement and dude what's my CTR am I seeing clicks so don't use vanity metrics you can't eat vanity metrics but you can use vanity metrics to give you a senses to the health of a campaign and as long as we assumption that this campaign is participating in the success of the overarching marketing mix earlier than I'm able to track and
Escúchalo en el minuto 45:13
La ilusión de las exclusiones y segmentaciones precisas en Google Ads
Canales de adquisiciónLas campañas que pretenden ser de prospección pura o retargeting estricto no logran aislar audiencias de forma limpia: debido a las bajas tasas de emparejamiento de datos (con márgenes de error estimados de más del 80%), las plataformas publicitarias terminan operando con públicos amplios y opacos.
such thing because Google can't even match its own damn data here's what this means there's no audience exclusions you can't exclude website visitors which means you can't do real prospecting campaigns you can't exclude converters which means you can't do remarketing to everything's Broad everything's opaque you can't exclude existing customers which means you can't do acquisition think is oh I I'm getting it 80% right with 20% margin for error no you're getting it 12.5% right per the date I just gave you or 17% right with 87% margin of error or 83% margin of error gray most of what you get…
Escúchalo en el minuto 20:07
Canales de adquisiciónCuando la mayoría de los anunciantes concentran sus presupuestos exclusivamente en Google o Meta, los canales tradicionales (correo directo, vallas publicitarias, radio, llamadas en frío) y las redes secundarias se convierten en océanos azules con menor saturación y costos competitivos.
first thing that I'll say is if you're only running singular channels and I don't just mean just running meta and Google like you should be running every applicable channel to your audience period full stop you should be running back to traditional advertising like uh acquisition tool it was direct mail and it's because everybody used to use Direct Mail when they all transitioned to Google they took all their money out of direct mail and they put it in Google but they did it all at once so then Direct Mail became a blue ocean I think the whole world needs to start looking at those blue oceans…
Escúchalo en el minuto 32:53
Canales de adquisiciónUna estrategia de tráfico efectiva requiere activar campañas en todas las etapas del embudo (prospección, conciencia del problema, conciencia de la solución y reconocimiento de marca en el top of funnel) en lugar de depender exclusivamente de campañas de conversión directa en el fondo del embudo.
champion this narrative you have to run full funnel traffic so remember we have our XR Y and a z-axis the y axis is the funnel and people who aren't running top of funnel campaigns brand awareness campaigns prospecting campaigns um you know make people problem aware make people solution aware you have to run full funnel traffic that's number one
Escúchalo en el minuto 34:45
Desfase temporal en el impacto de ajustar presupuestos publicitarios
Medición y experimentosModificar el presupuesto publicitario al alza o a la baja rara vez muestra su efecto completo en ventanas de 30 días; el impacto en el marketing mix general suele requerir semanas o meses para manifestarse.
in in a more sophisticated fashion what if I ratchet spend up a little bit what happens what if I ratchet spend down a little bit what happens and most importantly when does it happen because it's not going to be in 30 days maybe you ratchet spend up and then the impact you see is seven months later true this interesting or S weeks later this is the farming corn analogy cuz you know we're no we're not driving a race car we're sailing a boat right it's like I'm going to turn a little bit I'm going to tweak a little bit and we're going to see how planting the corn in the spring and I'm…
Escúchalo en el minuto 46:08
Apagar campañas de prospección para verificar su impacto en cascada en la búsqueda de intención
Medición y experimentosPara comprobar de forma empírica si las campañas de prospección en redes sociales (como Facebook o YouTube) están alimentando las búsquedas de marca o de intención en Google, pausarlas temporalmente permite observar el impacto en cascada en las conversiones finales.
belong to he's like I just shut off all those campaigns to see where the holes were do you recommend people doing that to say oh Facebook is actually you know feeding my bottom of funnel you know scientific method I think that's probably the only way to know definitively where the joints are so if campaigns we feeding which campaigns by shutting one off you can see where the cascading impacts are but if you have a
Escúchalo en el minuto 43:51
El modelo del cubo de atribución publicitaria
Medición y experimentosEl tráfico publicitario funciona como un cubo tridimensional compuesto por la etapa del embudo (eje Y), la interacción del canal desde la impresión al clic (eje X) y el tiempo (eje Z); los sistemas de analítica estándar solo registran la esquina inferior frontal derecha correspondiente al último clic inmediato.
but just as a refresher all traffic is uh I'm is represented in the analogy that I'm offering as a cube and the cube is made up of uh a 10x10 X and Y AIS the Y AIS is the funnel top to bottom the x-axis is every potential ad channel uh left to right and starting on the left is impression and moving to the right is click and then the z-axis is time so that becomes 10 x 10 by 10 so you have a th cubes and roughly speaking you can see one you can see the extremee bottom extreme right extreme front right you can see the bottom of the funnel the the last click the most recent that's all you can…
Escúchalo en el minuto 6:19
Caída drástica de la tasa de coincidencia en audiencias de Google Ads
Medición y experimentosLa tasa de coincidencia (match rate) de listas de usuarios en Google Ads cayó de rangos históricos del 80%-90% a niveles de entre el 12% y el 15% debido a la transición hacia políticas y tecnologías de privacidad.
they had I mean dude our Google match rate in some instances historically was like 80 90% yeah so how can they drop from 80 to 90% to 12 to 15% when they
Escúchalo en el minuto 27:05
Transición del rastreo publicitario 1 a 1 hacia modelos agregados por cohortes
Medición y experimentosFrente a las presiones regulatorias y los cambios del ecosistema tecnológico, el seguimiento publicitario abandona el rastreo individualizado (1 a 1) para adoptar atribución y segmentación agregada en cohortes o grupos psicográficos/demográficos (como FLoC).
gdpr California legislation ADA ecosystem we cannot rely on the way that do tracking modeling attribution Etc and something called flock which is Federated learning of cohorts and it was they did instead of tracking users on a one to one basis they put users into clouds that were indicative of who that user is demographically and psychographically and then they provided you data on the cloud instead of on the right yeah without actually tracking
Escúchalo en el minuto 25:24
Medición y experimentosEl modelo tradicional de atribución publicitaria está obsoleto debido a la opacidad del ecosistema digital. Es prácticamente imposible rastrear cada visualización, impresión y punto de contacto a través de todos los canales en herramientas de terceros durante toda la línea de tiempo de conversión de un usuario.
attribution as we know is certainly as extinct as we knew it 10 years ago is digital marketing and 2024 and it's not going to change anytime soon it's not going to get less opaque so you might as well just just unless somebody can figure out how to track a view uh and an impression in some third party tracking software which is not happen in every single Channel across every single channel the entirety of a timeline that it takes somebody to convert I I'm just Leap Forward but like you're never going point is is like this is the reality of what we live in right now so make sure
Escúchalo en el minuto 55:33
Diseño tecnológico según el espíritu de la regulación futura
Estrategia y decisiones de negocioEn sectores regulados o con presión sobre privacidad, las plataformas tecnológicas reducen funciones (como el rastreo de datos) adaptándose al 'espíritu' de las normativas venideras y no solo a la ley vigente, ya que la legislación avanza más lento pero se actualiza con medidas restrictivas y de barrido amplio.
Google has pulled back on what it allows us to see from a targeting perspective made a strategic decision to change its Paradigm its approach in in order to assumes will be the regulatory environment moving forward and there's he go he said we can't build to be compliant with the regulation we have to build to be compliant with the spirit of the regulation because technology moves so much faster than regulation can keep up with you as the technologist have to Intuit it what is it that the regulator actually wants and means because when they go to update this they're going to have to Broad…
Escúchalo en el minuto 23:37
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.