Marketero · Entrevistas
How to Solve Your Digital Marketing Attribution Problem Once and for All | Perpetual Traffic EP 650
29 ideas de marketing sacadas de «How to Solve Your Digital Marketing Attribution Problem Once and for All | Perpetual Traffic EP 650», de Perpetual…
- 29 ideas despiezadas
- Fuente Perpetual Traffic
- Dura 59 min
- Publicada en 2024
Ver la entrevista entera en YouTube
De qué habla
Las 29 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 29 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 cuatro campañas para captación de leads en Google Ads
Canales de adquisiciónUna estrategia fundamental para generar leads en Google Ads se compone de cuatro campañas sincronizadas: 1) búsqueda en concordancia exacta orientada a Max Clics o CPC manual, 2) búsqueda en concordancia amplia clonada con tCPA, 3) campaña DSA (Dynamic Search Ads) con el 20% del presupuesto hacia las mismas páginas de destino, y 4) campaña de marca (concordancia exacta o de frase) asignando el 10% del presupuesto.
level yep so I would use a feeder strategy on search so use your exact match Max clicks or manual CPC then combine that with A tes or tcpa broad match using the same exact campaign just clone with that those two changes then I would use a DSA campaign at 20% of your ad spend using the dynamic ad targets that are also overlapping with the same landing pages that your broad match and your exact match search campaigns are also using make sure 10% of your money is allocated towards a I would either do a phrase match or exact match brand but be my kind of like first four things to
Escúchalo en el minuto 53:57
Canales de adquisiciónPara eventos de venta como Black Friday, conviene escalar el presupuesto publicitario al máximo en los días previos calculando el retraso de conversión (time lag de 4 a 7 días) y reducir la inversión aproximadamente a la mitad durante el día pico de la venta, ya que la mayoría de compradores ya interactuaron antes con la tienda y así se protege la rentabilidad.
actually reduce that time of sale almost like the day of really um a lot of time expensive my opinion is you start a week before your time lag and you scale up to as high as you possibly can afford until like the day before day of the Black Friday sale and then you you can if what you'll find is that the people that are purchasing are new to the site four five six seven days ago so you don't have to suspend it a bunch you can do a really good job actually reducing by about half and getting 99% of what you were going to do but you can save a lot
Escúchalo en el minuto 52:50
La lentitud de carga y la mala ubicación de etiquetas rompen la continuidad del rastreo
Medición y experimentosSi un sitio web tarda en cargar y las etiquetas de atribución se ubican al final del HTML, los cambios de página rápidos hacen que el tag no se dispare a tiempo; como consecuencia, las plataformas analíticas pierden la sesión del usuario y la reclasifican erróneamente como tráfico 'Direct' o no atribuido.
fast is your website y that's one of the website it fires the tag they change Pages it fires the tag again what people what this means is that if you're putting a tag on your site for attribution and that tag is deployed towards the bottom of the HTML that is going to take whatever seconds it's going to take for that thing to Fire and if it does not fire it's going to lose the person and but it didn't he wasn't on page two for four seconds it's ah it's Ralph's gone now we got this new person here Direct you log into ga4 and half the people are
Escúchalo en el minuto 5:31
Canales de adquisiciónLas campañas de búsqueda de marca a menudo se atribuyen ingresos que no generaron originalmente. Al analizar los datos desde el primer clic (first-click), es posible identificar si solo están canibalizando demanda existente; pausarlas puede ahorrar una fracción sustancial del presupuesto publicitario sin perjudicar las ventas globales.
why I kill brand campaigns for people where I was like hey we took this one campaign and we just wiped it out that was spending two-thirds of the cost and two-thirds of the revenue we just wiped it out saved two-thirds of our ad spend that week 10 G's in one week no 15 G's in a week didn't lose a that's because
Escúchalo en el minuto 20:13
Priorizar Google Shopping sobre Search non-brand al iniciar en ecommerce
Canales de adquisiciónAl lanzar campañas pagadas para una tienda de comercio electrónico, se debe priorizar Google Shopping (o Performance Max si no existe tráfico omnicanal previo) antes que campañas de búsqueda no de marca. Shopping muestra la imagen, el precio y el nombre del producto directamente en la página de resultados, calificando al usuario antes del clic sin coste extra, a diferencia de Search, donde se paga un CPC más alto…
e-commerce number one I usually don't I usually start just standard shopping first or if you have no Omni Channel traffic performance Max it could work great because you don't have an existing brand or existing customers or existing traffic to go after to mess everything up performance Max can work reason why I don't start with search a lot of times I won't even really run search for the first year for e-commerce are selling Crocs right I have to go and click on this ad and you pay Google 40% more for me to be able just to find out price I can find all of that out before you have to pay…
Escúchalo en el minuto 36:34
Canales de adquisiciónAl combinar campañas en Google Shopping: si se parte de Standard Shopping y se incorpora PMax, Standard Shopping debe retener el 80% del presupuesto; si se parte de PMax y se incorpora Standard Shopping, Standard Shopping debe recibir el 50%. Con tráfico alto proveniente de Meta, conviene escalar Standard Shopping más rápido que PMax.
you're already running pmax add standard shopping obviously and if you are running standard shopping add pmax now the difference is if you're running standard shopping and you're adding in pmax standard shopping should remain 80% of the ad spend on the flip side if you're already running pmax and you're running St and you want to add standard shopping standard shopping should actually be 50% of the asend the only variable that this differs on is if you have a high amount of metat trffic you're going to need to increase your standard shopping faster than your performance Max to see if you can…
Escúchalo en el minuto 41:52
Monitorear el costo por nuevo visitante para validar la adquisición real y la escalabilidad
Medición y experimentosEl costo por nuevo visitante es una métrica fundamental para evaluar si la adquisición de tráfico nuevo se mantiene dentro de márgenes financieros sostenibles y si una campaña puede escalar, evitando distorsiones cuando los visitantes recurrentes no son identificados correctamente.
off yeah now that's effec cost per new visitor is one of our 15 mpis very cost that you use in order to figure out whether or not you're getting are you getting new people to your site because that's why people hire us to get them new customers at an affordable price makes sense for their finances right it's usually how growth happens and if you know can you afford to scale that
Escúchalo en el minuto 10:21
Medición y experimentosLas plataformas publicitarias como Google o Meta deben emplearse para ejecutar campañas y alimentar estrategias de puja, pero no como la fuente de verdad para la atribución y medición, ya que tienden a atribuirse conversiones de clientes recurrentes e inflar métricas como el ROAS.
looking at the platforms for for metrics I'm looking at I'm using the marketing platforms as marketing platforms right don't use it for measurement platform say in AO sucks it does however we still do use it as I need a benchmark it I need it for I need it for my my bidding what the I found the same ones over and over again great look at me when in fact they like well why do I have the same amount of Revenue and it looks like you've got 400% Raz this week that still
Escúchalo en el minuto 26:12
Costo real por visita nueva en campañas Performance Max vs. ROAS reportado
Medición y experimentosLas campañas de Google Ads Performance Max pueden reportar métricas atractivas en la plataforma ($2 por clic y 400% de ROAS), pero al medir con precisión se detectó que solo el 5.5% del tráfico correspondía a visitas nuevas, elevando el costo real por nueva visita a $33 al concentrarse en usuarios recurrentes.
new visit percentage we know that Google ads is running performance Max which is why only 5.5% of these people are new and it costs you $33 every single time there's a new visit this is important as [ __ ] because if you look at Google ads it's like $2 clicks and these things are getting 400 razes no that's not true this is the health of your company I'm only we're only spending 30 grand out of 225,000 so about 10ish percent because we know that this is never going to scale and I would I would waste $100,000 a six month of my time because I didn't know that and I didn't know this and that is…
Escúchalo en el minuto 16:25
Alimentar las plataformas publicitarias desde un Data Warehouse central con first-party data
Medición y experimentosCentralizar los eventos de conversión como datos de primera fuente (first-party) en un data warehouse y sincronizarlos de vuelta hacia las plataformas de anuncios (Meta, Google) y el CRM incrementa la visibilidad de conversiones reales, lo que permite a los algoritmos de puja automatizada (como tCPA) optimizar de forma más rápida y precisa.
coming in because it's being caught in the data warehouse it's now first-party but when it goes into Data warehouse shoots back out of this this is the your reports this also passes it back to the ad platform so that the ad platform have my in app data go from 50 to 75 I'm 50% closer than I was to the actual truth excellent it means F faster optimizations it means that my slanted T has in tcpa goals that are using half data of targets because it is only half accurate can be a little bit more accurate could be sent to the CRM tools so the platforms not only for ads but the
Escúchalo en el minuto 22:12
Dependencia técnica de Content API para acceder a YouTube Affiliate
Canales de adquisiciónPara permitir que creadores de YouTube elijan autónomamente productos de un catálogo y cobren comisión por venta mediante el programa de afiliados de YouTube, es obligatorio sincronizar el catálogo vía Content API (app de Google y YouTube). Usar gestores de feeds externos como DataFeedWatch u Optimizely descalifica la tienda para participar en este programa.
affiliate program that is being released go and check it out YouTube affiliate means that it's gonna Google is going to open up the floodgates to anybody that has a YouTube channel to Simply go and select products from your catalog that they could autonomously put on their uh YouTube channel to go try to sell without you even knowing it and then you have to pay them your commission and you they wanted to the problem is you only get that through content API which is the Google and YouTube app instead of Shopify if you use data feed watch if use optimizely you will no longer be eligible to join…
Escúchalo en el minuto 51:13
Medición y experimentosLa decisión de compra suele tomarse antes de que el usuario busque activamente el nombre de la marca en Google (impulsada por canales de descubrimiento como Meta). Por tanto, el ROAS de campañas de búsqueda de marca suele estar sobreestimado y conviene reducir al mínimo el gasto en captura de marca para no pagar por ventas que ya estaban decididas.
what's your logic there actually no I spend the least amount of money on brand as I possibly can um brand is always the I'm gonna go buy it point where you said I'm gonna go buy it that is the decision to then Google search the brand you do not end up just randomly Google search the brand and say well [ __ ] I want to buy it now no the decision as to which what you want to buy happens before the Branded search so how much money do you have to spend in order to make sure you capture the person post decision and that's why branded Raz doesn't really mean that that campaign did anything that is…
Escúchalo en el minuto 43:12
Medición y experimentosLa decisión de compra suele tomarse antes de que el usuario busque activamente el nombre de la marca en Google (impulsada por canales de descubrimiento como Meta). Por tanto, el ROAS de campañas de búsqueda de marca suele estar sobreestimado y conviene reducir al mínimo el gasto en captura de marca para no pagar por ventas que ya estaban decididas.
what's your logic there actually no I spend the least amount of money on brand as I possibly can um brand is always the I'm gonna go buy it point where you said I'm gonna go buy it that is the decision to then Google search the brand you do not end up just randomly Google search the brand and say well [ __ ] I want to buy it now no the decision as to which what you want to buy happens before the Branded search so how much money do you have to spend in order to make sure you capture the person post decision and that's why branded Raz doesn't really mean that that campaign did anything that is…
Escúchalo en el minuto 43:12
El etiquetado en el Edge previene la pérdida de atribución frente a bloqueadores de cookies
Medición y experimentosCapturar y etiquetar el tráfico a nivel de Edge (antes de que la solicitud llegue al servidor de origen o se ejecuten los scripts en el navegador del cliente) permite registrar la procedencia del usuario antes de que intervengan bloqueadores de anuncios, herramientas de privacidad del navegador o restricciones de cookies.
break your privacy tools cookie blockers browser based blockers have no effect on failed at that point in time and that's but here's the edge so this is The Edge now what the edge does is essentially tagged you in the parking lot the of Costco says Ralph's here Ralph's back I know where Ralph came from exactly we know where Ralph is so it tags you because of all those blockers that are on the site exactly blockers on the site blockers from the client for the the client but the user yep the cookie blocker the app blocker they have all that stuff basically we gets iOS 14 exactly we get to tag…
Escúchalo en el minuto 12:06
El ajuste de adquisición exclusiva de nuevos clientes en Performance Max puede seguir captando clientes recurrentes
Canales de adquisiciónA pesar de activar la configuración de adquisición exclusiva de nuevos clientes en Google Performance Max y cargar listas de clientes, el sistema sigue pujando y atribuyendo ventas a clientes existentes debido a retrasos y limitaciones en la coincidencia previa de datos.
this is a performance Max campaign for new customer acquisition only and new October only bit for new customers it's like you got it we uploaded our customer it's still there like it's still we have $6,600 in existing customers and this only bid for new customers says great out of 518 that are known 12 25 of them are returning customers we got them back that is because it doesn't know it C match has to go and do enhanced conversions later on to find out guess what I still [ __ ] you over like
Escúchalo en el minuto 27:36
Sobreestimación del canal orgánico en atribución full-funnel frente a primer clic
Medición y experimentosEn negocios impulsados por pauta y remarketing, los modelos de atribución de embudo completo pueden sobreestimar los ingresos atribuidos al canal orgánico ($141,000 frente a solo $4,000 medidos a primer clic), al no reflejar que el contacto inicial provino de campañas pagadas.
spoil it so here's the full funnel model this is using full funnel attribution it says Hey organic had 141,000 of your Revenue out of 789,6 now this is my personal brand I have never run anything but a little bit of you know Facebook remarketing which is why this thing is 0.5% new visits man that's really accurate but the revenue says Hey 141,000 Revenue uh from organic and then we said wait a minute that can't be possible we don't have organic traffic we we're we're reselling of luggage like Samson night no one knows who travelo is that's never come from anywhere else but but the return…
Escúchalo en el minuto 19:13
Canales de adquisiciónCuando una campaña de Performance Max (feed-only) canibaliza y bloquea el tráfico de Standard Shopping, se puede reactivar la entrega configurando un tROAS muy agresivo y bajo (10% a 20%) en Standard Shopping y duplicando su presupuesto respecto al de Performance Max.
standard product ads manual CPC uh I use a low T row ad it seems to be really aggressive so if you a pmax feed only and it's doing really really well I will go down to like a 20 even a 10% T Raz inide of a standard shopping campaign and I will double the ad spend compared to the pmax usually just launches it to
Escúchalo en el minuto 54:50
Canales de adquisiciónCerca del 80% de los usuarios toman la decisión de compra antes de buscar el nombre de la marca en Google. Por ello, la inversión en anuncios de búsqueda de marca (branded search) debe calibrarse únicamente para no perder clics reales sin sobrepagar en CPC, incluso en compras de alta consideración.
happen before they Google the brand so spend as little as you can there that pay that person money typically you make your decision before you Google you do purchased not necessarily I will actually spend enough to not lose out on every click okay and that's where I say don't try not to overspend and that's people make most of the time 80% of the time ish they're making their cons their purchase decision before they Google the brand to then go and buy it or go to
Escúchalo en el minuto 45:14
Medición y experimentosIdentificar el origen exacto del tráfico y dejar de sobreinvertir en usuarios que buscan la marca orgánicamente y convertirían de todos modos permitió un ahorro de $33,541 en 14 días.
are all people that we know when you when you count the people that are coming in organically they're Googling the brand name why they didn't start there so don't overspend on those people because you will still get it anyway this being able to see this saved US $ 33,541 14 days if that's not a use case
Escúchalo en el minuto 30:20
Medición y experimentosUn ROAS elevado en Performance Max suele ser engañoso si la plataforma destina hasta el 80% del presupuesto a tráfico de marca o clientes reciclados; es fundamental segmentar y auditar la adquisición de clientes nuevos frente a los recurrentes.
how many people are newe how many of those that 900% row as on your pmax is actually new customers or just Google recycling old ones it's like how many new and returning users are there when they say don't know I say well how many new and returning customers are there course but a lot of times performance Max is 80% of your ads spend to those people yeah and that's what it is is if
Escúchalo en el minuto 33:12
Superar el Ad Rank de Performance Max mediante Smart Bidding agresivo
Canales de adquisiciónPara ganar prioridad en la subasta y mejorar el Ad Rank frente a campañas Performance Max en Google Ads, conviene configurar campañas de Smart Bidding agresivas fijando un objetivo de tROAS bajo o un tCPA alto que dé margen al algoritmo para pujar más alto.
but I would I would run this thing on a low T Raz or a High tcpa um have it be smart bidding and have it be really aggressive what that does is it basically it does the same thing as performance Max does but that campaign says I can spend twice as much and so it earns a better ad rank now if you're you that has worked in every account that
Escúchalo en el minuto 55:18
Canales de adquisiciónLas audiencias preconfiguradas por plataformas como Google durante picos estacionales (como Q4 o Black Friday) se basan en correlación y no en causalidad, asignando etiquetas de compradores estacionales a usuarios por simples visitas fortuitas a sitios web en oferta.
for audience if it's not a known audience be very weary of how Google classifies an audience right now in Q4 everybody is tagged with Black Friday SE Monday shopping and Christmas shopping even if they're like Jewish just because they went to a website had the had a sale like it doesn't mean they're actually even interested in even Christmas so just be very mindful that to be running so the audiences are just more like their correlation not causation so be very mindful the person
Escúchalo en el minuto 57:08
Canales de adquisiciónEn campañas de adquisición en Google, es preferible dejar que el algoritmo optimice con usuarios propios en vez de forzar retargeting a visitantes web provenientes de Meta; la tasa de coincidencia (match rate) de Google tag a Google tag es significativamente superior a la de pixel de Meta cruzado con Google.
to establish familiarity no I'm actually doing only Google algo because I don't want to necessarily Target the site visitors as an audience I want them to do that on their own but if I take a tag that has been filled up with a bunch of meta users I don't really want it I want the tag to come from fresh Google users and the match rate between a Google tag and Google tag is much better than a meta pixel to Google tag so if you pre-build it you're G to get a whole bunch of metap pixels
Escúchalo en el minuto 49:48
Pérdida de precisión en la atribución nativa por restricciones de privacidad
Medición y experimentosEl rastreo analítico tradicional basado únicamente en etiquetas y píxeles ejecutados en el navegador ha sufrido una fuerte degradación por bloqueadores de cookies, ad blockers y modos de consentimiento; debido a esto, la atribución nativa de plataformas como Shopify solo alcanza aproximadamente un 55% de precisión en Estados Unidos y un 45% en regiones asiáticas.
good two years ago that's been heavily deprecated in the last about 720 days um to the point where all of these interrupts privacy settings consent mode cookie blockers ad blockers have whittel down to what Shopify has even said on their own data shopify's data is 55% accurate in the United States 45% accurate in Asian territories and that's the only two two attribution model that attribution model even if you set to like first click and last 30 days that attribution model will show approximately 55% accuracy if
Escúchalo en el minuto 4:04
Canales de adquisiciónAl pausar las campañas de búsqueda de marca que consumían 36.000 $ de un total de 62.000 $ y destinar un incremento de solo 6.000 $ a escalar campañas amplias de prospección, la cuenta generó 250.000 $ en ingresos ahorrando decenas de miles en tráfico que ya era orgánico.
we said okay we can actually take all of our branded traffic that we that we spent 36,000 out of 62,000 on and I'm like let's just pause it just kill it we don't need it just save us 35 Grand there and the broad campaign here that spent you know $6,000 more which is 500% more because they started to scale it up all of a sudden just lands with a quarter million in two
Escúchalo en el minuto 29:47
Importancia de la captura de datos en el Edge para la consistencia de la atribución
Medición y experimentosEl despliegue de etiquetas de seguimiento a nivel de Edge (CDN) desde el primer contacto garantiza que los datos de atribución se propaguen correctamente en el tiempo; si el rastreo inicial falla al capturar la primera visita, los modelos posteriores quedarán corrompidos de forma permanente.
the way that the tag deploys is extremely important because it's like you have a amazing tag that's your engine of your Ferrari but the Ferrari doesn't have wheels like you can have amazing tagging capabilities but it's can't go anywhere it's not fast enough to actually deploy and so that's kind of the analogy there is the way that things are deployed Matters from that visit for the first time for the next years really if it fails at the beginning it will never propagate later correctly at the end right so the edge tag this is that goes into the data warehouse the data warehouse then feeds…
Escúchalo en el minuto 15:33
El CPC en Search suele ser un 40% más caro que en Google Shopping para ecommerce
Canales de adquisiciónEn comercio electrónico, el costo por clic (CPC) en campañas de búsqueda (Search) non-brand suele ser aproximadamente un 40% más costoso en comparación con el CPC de las campañas de Google Shopping.
great non-brand search hard it's always more expensive than e-commerce than shopping usually buy about 40% more your CPC is from search to to shopping your 40% more search CPC than shopping but click on this ad and you pay Google 40% more for me to be able just to find out everything's good so don't pay 40% more
Escúchalo en el minuto 36:50
Medición y experimentosEl aumento de bloqueadores, consentimientos y fallos de carga obliga a las plataformas (como Meta o GA4) a rellenar los datos no atribuidos con modelos de IA estimados, lo que genera reportes inexactos y decisiones de inversión de alto riesgo en estrategias omnicanal.
northbeam triple whale it's like throw on our Ai and we'll tell you what yeah we're gonna model it that's a very expensive decision and then even meta is use modeled conversions we don't really multimedia Omni Channel mix so now you got a whole bunch of unattributable Blended together and half of it's unattributed you're like well [ __ ] I don't know what to do now I have half my your site isn't fast enough but also there's pop-up lockers there's at that sit on the browser yeah consent platforms to model or the third party attribution systems to model the data to
Escúchalo en el minuto 7:20
Tiempo promedio de Largest Contentful Paint (LCP) en pequeños negocios
Medición y experimentosEl Largest Contentful Paint (LCP) promedio para sitios web de pequeños negocios oscila entre 8 y 16 segundos, comparado con los 4.1 segundos de grandes marcas como Nike.
don't quite understand is that even Nike themselves takes 4.1 seconds to load a it is that is a 4.1 seconds for largest contentful paint which means the page your average largest contentful paint for like a small business is anywhere between like eight and 16 seconds and
Escúchalo en el minuto 5:44
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.