Validar anuncios con pocas conversiones o micropresupuestos no predice su rendimiento a escala

So the one campaign structure says Facebook is a machine learning algorithm. What we are trying to do is present a control and then use the scientific method to see if we can change the outcome. So we have a winners ad set of proven ads for whatever offer it is. And then the purpose of other two tests is simple. When I launch this test, does it earn spend? And if so, what happens to the bottom line of business? So this is simply scientific method applied to a data machine. So let's just use the machine in way it was designed to be used.

What I Learned from One of Meta’s Top 100 Global Advertisers - Charley Tichenor IV - E65 [English] · 2025 · comprobar

Whatever wins there, I'm then going to put in my scaling campaign. Here's the problem. What works at $20 is not in any way predictive of what works at 300 or a thousand because the load You were good at making me one. When I asked you to do exponentially more work, my expectation of you producing the same level of work, well, is setting you up for failure. is to see what gets lucky early on a small budget. Cause also when you launch new ads, they're generally launched at the bottom of the funnel because the machine is trying to understand with ads. That's why all of your ads look great for…

Meta advertisers, we've got a problem... · 2026 · comprobar

it looks like to move from ad to ad to ad. When you have 25 to 30 ads, there is no way for you to look more money to exclude a bunch of good people just so that you can have dumber machines with 25 to 30 ads. That's nearly a hundred ads running. Do you have the budget to have any statistical significance across 75 to 90 ads? No, you don't. I don't care what your budget is. That number is astronomical. The amount of variations that you would need to get through just 50. Let me give you a concept of what that looks like. Let's say you did one variation a second. You would need 78 billion years…

Meta Ads expert reacts to $1M/month Strategy · 2026 · comprobar

with it and I would say the minimums really you're looking at doing is if you're running um a campaign you want to be at least getting every creative to have three to five times your target CPA and spend per week so if you have a a certain budget for the campaign as long as every creative in that budget could get fre to five times a CPA they might

Why Your Marketing Tracking Sucks (Even If You Think It Doesn’t) and How to Fix It | PT EP 674 · 2025 · comprobar

problem is you weren't able to say this is my problem. Let me test to solve it. You just I have a new idea. Does it make things better? Those are two very different things. One is a pointed understanding of a hypothesis, a testing framework, and then proving out whether you're right or wrong, and the measuring the impact. is I have a new idea. Let me throw it into the randomizer and see if the output is good. That's just not as valuable and actionable a way of

Andromeda just changed Meta Ads forever · 2025 · comprobar

mistaken opinion that your testing starts your testing campaign but it doesn't your testing campaign is seeing what lottery ticket do you want to actually install to your scaling campaign and your test doesn't actually start until it hits that scaling campaign because it doesn't actually have to spend any meaningful money until it hits there so you're pulling Lottery tickets to seeing what you think your best chance is to actually start your creative test okay great your winner had what three sales five 10 how many sales when were day on for me that's a very low Integrity data set built on an…

Dynamic Creative & Machine Learning | Adspend | Charley Tichenor · 2023 · comprobar

inside of Facebook split testing doesn't make any sense for several reasons one like you're doing it one of three ways that all objectively hold no weight as soon as you apply like logical thinking to it one of them would be I'm gonna launch 10 new ads a week that's gonna have a budget of let's say but I'm sending a thousand a day 100 bucks is gonna go here and so you're splitting that up so that if it was going completely even split each Advocate like 10 bucks a day do you know the impact that Those ads have on your Omni Channel marketing mix no there's no way in that environment you could…

Cracking the Code: Effective Strategies for Facebook Advertising · 2023 · comprobar

take your um take uh you know know the learning phase conversions how however however many learn conversions you need to exit learning phase set a budget of at least 21 conversions per day and then tangible results back and run tests for interrupting it's 20 a day we you you need to get out of learning phase with with 50 conversions 50 within a within a within a seven day period seven period however the first

How to Use TikTok Ads The Right Way to Grow Your Business Part 2 | Perpetual Traffic EP 629 · 2024 · comprobar

uh not really a niche it comes down to uh spend amount for us you know it's really spending at least ten thousand dollars a month in advertising if you're below it it's not really the best system because it is it is a larger system to implement into your ad account and if you don't have the budget for something like that you really do need to focus on efficiency because you need to drive those signals um from a media buying perspective

Ashton Shanks of HemonX & Charley T - Founder of Disrupter School Interview · 2022 · comprobar

first off part of the issue is scale like if you're spending if you're investing 10 bucks across 20 ads or whatever like you're getting the actual results you're seeing are so small whether it's like one purchase or two purchases and then you're basically saying all right now we're gonna take all this budget and applied on that the results will be different in an environment where they're artificially small numbers versus you're saying now we're going to put a hundred thousand dollars behind that is that more or less and if that ad crushes it 20 bucks a day

Cracking the Code: Effective Strategies for Facebook Advertising · 2023 · comprobar

uh not really Niche it comes down to uh spend amount for us you know it's really spending at least ten thousand dollars a month in advertising if you're below it it's not really the best system because it is it is a larger system to implement into your ad account and if you don't have the budget for something like that you really do need to focus on efficiency because you need to drive those signals um from a media buying perspective

Ashton Shanks of HemonX & Charley T · 2022 · comprobar

first off part of the issue is scale like if you're spending if you're 10 bucks across 20 ads or whatever like you're getting the actual results you're seeing are so small whether it's like one purchase or two purchases and then you're basically saying all right now we're gonna take all this budget and apply it on that the results will be different in an environment where they're artificially small numbers versus you're saying now we're going to put a hundred thousand dollars behind that is that more or less

Cracking the Code: Effective Strategies for Facebook Advertising · 2023 · comprobar

Entonces, la estructura de una sola campaña dice que Facebook es un algoritmo de aprendizaje automático. Lo que intentamos hacer es presentar un control y luego usar el método científico para ver si podemos cambiar el resultado. Así que tenemos un conjunto de anuncios ganadores, anuncios probados para cualquier oferta que sea. Y luego el propósito de las otras dos pruebas es simple. Así que esto es simplemente el método científico aplicado a una máquina de datos que, en mi opinión, pasó de montar caballos a llegar la Luna en 100 años gracias al Y bueno, sabemos que Facebook es un algoritmo de…

Los Secretos del Padre del Método 3:2:2 en Meta Ads - Charley Tichenor IV - E65 [Español] · 2025 · comprobar

uh not really Niche it comes down to uh spend amount for us you know it's really spending at least ten thousand dollars a month in advertising if you're below it it's not really the best system because it is it is a larger system to implement into your ad account and if you don't have the budget for something like that you really do need to focus on efficiency because you need to drive those signals um from a media buying perspective

Ashton Shanks of HemonX & Charley T · 2022 · comprobar

inside of Facebook split testing doesn't make any sense one of them would be I'm gonna launch 10 new ads a week that's gonna have a budget of let's say 10 of your budget your overall net but each Advocate like 10 bucks a day seventy dollars at the end of it do you know the impact that Those ads have on your Omni Channel marketing mix no there's no way in that environment we also have to say let's say two of those ads are winners so eighty percent of your money went to creating harm for your business

Cracking the Code: Effective Strategies for Facebook Advertising · 2023 · comprobar

like 20 ads I see so many people like well I'm building 10 different versions of an ad or 20 different versions of an ad I'm testing a bunch bunch of different audiences they're spending maybe 10 bucks a day or 50 bucks a day or $100 a day for a week or two and they come out of it with something that did I get lucky and then if they did they come up with a winner and then they're like okay I found one thing that works my question to them is is that going to be so successful that you can pay off all the losses like if you're playing if you're if you're thinking like all right all I have to do…

Facebook Ads Cheat Codes with Kelvin Parker · 2024 · comprobar

campaign but it doesn't your testing campaign is seeing what lottery ticket do you want to actually install to your scaling campaign and your test doesn't actually start until it hits that scaling campaign because it doesn't actually have to spend any meaningful money until it hits there sh50 ads okay great your winner had what three sales five 10 how many sales when got lucky the most is that anything at 50 bucks a day 500 a day that you want to spend 5,000 or 50,000 or 100,000 a day on for me that's a very low Integrity data set built on a non-actionable set of data that I'm then risking…

Dynamic Creative & Machine Learning: Adspend with Charley Tichenor · 2024 · comprobar

so going into the testing framework and how this then translates into the ad account, this is what the testing cycle looks like, which is that you build a hypothesis, which is you define what you're actually testing. So, when we're introducing new creative, generally the test is the concept, and particularly one of those four variables that we're changing in the creative. Then we want to build a bunch of ads under that concept. One big mistake that a lot of people make is that they come up with a concept, they build it all out, it looks amazing, and then they make one ad, or they make two…

Meta Ads Creative Strategy in 2026: The Full System · 2026 · comprobar

have enough conversion you don't uh maybe you have yeah CPA but you have two conversions not out of the learning phase in your test campaign so it's something that is completely unstable Facebook says that there's no guarantee that another campaign start and will give you the same result with the same now stupid and criminal thinking that uh you maybe if your ad takes two or four conversions Okay purchase conversion or lead conversion with good CPA is foolish my yeah I tested already okay the test is over is completed I can scale it right yeah the idea of a testing campaign and a scaling…

Crushing Facebook Ads: Luca Amici's Success Story · 2024 · comprobar

pool quality data for example you don't have enough conversion you don't uh have two conversions not out of the learning phase in your test campaign so it's something that is completely unstable Facebook says that uh there's no guarantee that another campaign start that uh you maybe if your ad takes two or four conversions Okay purchase conversion or lead conversion with good CPA is foolish thinking that if I put okay the test is over is is completed I can scale it right yeah the idea of a testing campaign and a scaling campaign and and well it got three sales really just doesn't make any…

I Spent 30 Days Mastering Facebook Ads and CRUSHED It! · 2025 · comprobar

understand from a very core element that Facebook is a there's no different than anything else it's a scientific method platform so you have to understand what's my hypothesis what's my expected outcome on people don't necessarily have the budget or the or the the awareness to see tests all the way through if you don't have the budget to understand that test to a completion of statistical significance it's going to be really difficult for you to be able to take one lesson and apply it to someplace else

Facebook Ads Secrets with Charles Tichenor · 2022 · comprobar

Ahora hay que tener cuidado porque el riesgo es que mucha gente se pone a testear por testear, no saber qué es lo va a funcionar. En realidad la diferencia más que testear iba con una hipótesis de algo que tenía sentido. Entonces Creo que el error cometen muchos es que no saben ya sé qué está funcionando, entonces voy a probar muchas cosas distintas. Voy a probar muchas distintas a ver qué funciona. Y para que eso tenga relevancia estadística tenéis que gastar mucha plata, es muy caro. O si de repente capaz que vaya a tener uno, oh, esto funcionó, pero podría ser un Podría ser que funcionó…

¿Por Qué los Anuncios Feos Funcionan Mejor en 2025? - Tomás Girardi y Vale Ahumada - E37 · 2025 · comprobar

has a budget because the way that you teach the way that Facebook actually works is you don't really just put a $5 doll budget down right right yeah you need you need and not to blow completely things but like we'll say on a healthy basis you are above the you're you're spending some money you're spending some cash exactly like we don't need to give away your business but like you're investing some dollars on a daily basis for sure so um if you're that kind of person that's investing some dollars and um you're ready to learn about how Facebook actually

I Spent 30 Days Mastering Facebook Ads and Got Astonishing Results · 2024 · comprobar