Meta Reklamları Eskiden Neydi, Şimdi Ne Oldu? (Andromeda Öncesi ve Sonrası)

Hi Ozan, how are you? Hi Doğu. I'm fine. How are you?

I'm fine too. Last week in the office you said something like this: We

're developing and implementing MetaAndromeda, but nobody else is. We need to teach what we've learned

[music]. So what happened to the method? What changed? Well, if you remember from the meeting—

or rather , we've been talking a lot about Andromeda in our

Wednesday meetings for quite some time now—

we're talking about creative marketing ideas and

how to adapt them to this update, rather than just the technical aspects of the strategies. So when you said that, it just

came up automatically. We need to explain this. It's not something that can be covered in one or two videos

. We need to go into a bit more detail. Then we said okay, we'll do it. How should

we do it? Should we explain it using screen recording?

Or should we go with a Q&A session, etc.? We decided to go with a Q&A session

. We decided to start with that. This is the first video. Let's start.

Let's see. Of course, there will be things we need to show. We'll continue to show

those things here in videos via screen recordings. So,

if there's something visual while we're talking,

we'll continue with examples to support what we're explaining. But this isn't something that can be done in just one or two videos.

As you said, we also have a responsibility to teach what we've learned. So

let's start explaining. How long will it take, how many weeks?

Let's talk a little about marketing. I mentioned in the previous video that in 2026, we'll be on YouTube, and

we'll continue teaching only the new things we've learned here. This is

one of the important topics, and from what I've seen,

it's not really understood. When you mention Andromeda, many people, many brands,

when I look at their libraries, think they need to improve their ad creatives.

Beyond that, from a technical and marketing perspective,

they haven't delved into the details, as far as I can see. So

I think we need to explain the technical aspects a bit. Yes. One video isn't enough, but

what I've seen is that people don't realize what was happening before in the meta. What

was happening before, and what did Andromeda bring? Before we answer that,

we need to talk about this. If we think about the past, we actually

need to look at the past. So, at the end of the day, Meta is a tool. It could be a TikTok ad,

it could be a Google ad. Meta

invested in its algorithm to automate what it previously did manually, simply to improve

its performance. They physically

integrated this into their servers and ad distribution processes. This essentially allows us

to do the manual work of testing creatives

based on target audiences to find the best way

to show the right ad to the right person at the right time. Now, the algorithm can do this automatically

the moment you upload the ad to your account.

Naturally, this changes the way we approach things 180 degrees. If we

continue to handle things the way we did before, we'll just burn more money.

This doesn't mean that because it's also available on Andromeda, all the old strategies are obsolete

. It's unnecessary. Why? Because when you do manual testing, you only focus

on the tests that fail – and you know that if you try 10 things in ads, 9

of them fail. There's no need for that. To use your budget more effectively

, the way you launch campaigns, the way you think about ads,

understanding your target audience, correctly addressing and highlighting personas, and

knowing their market share have become crucial. This

isn't a radical change; it's simply what should have been happening, now that technology allows it

. Think of it that way. We used to make offers based on the steps of a prediction,

we created ad strategies. So what's the difference now

? What's changed? As I said, it's

just the way we eat yogurt that has changed. Actually, the ultimate result, the ultimate goal is the same.

Here, when we launch ads in the algorithm, we manually test single creatives

and find concepts from those creatives, and

find which target audiences they perform well with. This had to be done manually

before. That manual process is gone. What does this mean? The more I

diversify my creatives, the more different personnel I can include in the broadest segment of my target audience

. In my target audience with the broadest demographic structure...

By allocating the right proportions and budgets to different segments, and

leaving it to the algorithm, I can spend my budget much more effectively and

achieve much better results and campaign returns.

That's actually the main difference compared to before. To understand this, you need to have a good grasp of marketing.

Actually, positioning strategies, which we've already discussed on YouTube,

have been a very positive update to how we do business.

Things have become a little easier for us. This is where we have strong points

. But while it was possible to differentiate ourselves technically before,

it's become a little more difficult with this update.

So, are the fun steps no longer necessary? Have the old videos become obsolete?

Well, they're no longer necessary. It says here: Your target audience equals your

new target audience. In fact, there was a sentence in one of the webinars: "New

targeting is creative." New targeting or targeting equals creative.

It means your target audience or targeting equals your creative.

This means you should explain who you want to target in your creative

. For example, in a product group X... In the previous video,

I used clothing as an example, but

you can't say to someone looking for clothes to wear out at night, "Click here to see my pants." At that moment,

they're looking for dresses, not pants. When I say clothes, understand it as dresses.

They're not looking for pants, they're looking for dresses. Therefore, you need to approach them with an ad tailored to their needs.

You need to go into more detail and

know which awareness stage each persona is at. Similarly, you need to create different messages for them and divide them according to these

funnel steps. Does that customer know you?

Have you ever interacted with them before, or do they not know you at all?

You need to think about these funnel steps and your target audience in advance in the creative process.

So, in the previous working system, was the algorithm

punishing us for its own shortcomings? Let's not say it was punishing us, but

as people managing the ads, we had to solve these issues. We had to

analyze the numbers correctly, analyze the behavior of the target audiences correctly,

and direct the next campaigns accordingly. But now,

because the advertising distribution process has changed and become much more powerful,

the algorithm can automatically decide

which ad a person will see at what time from millions of ads.

Therefore, the more variety we can increase in these auctions, the more

people and personas we can reach, and consequently,

the more we can increase our sales potential.

Think of it this way in the previous year. You created an ad. Let's use the

previous example. I always say this to compare apples to apples.

For example, your creative has the season's trendiest dresses. But

do all the people interested in women's clothing products

have the chance to be looking for dresses at that moment? No. How was it before? It tried to reach

people who were exhibiting similar behaviors to those who clicked on your dress product and interacted with your ad

. The algorithm now already knows this in advance. Therefore,

it gives you this freedom. If you know your customer well, you know that

some people are looking for dresses, but others are looking for baggy pants. Some

are looking for a certain type of jacket. Some are looking for a leather jacket, right? And

there are specific reasons why they are looking for these things. The more specific you are in your message

and the more variety you can provide in your visual pool, the better. Because yes, some people

are different in terms of marketing message, product group, or personnel, but

the types of content they are drawn to are also different. Some click more

on real-world ads, while others click more on catalog ads

. Some click more on visuals, and so on;

you need to provide that variety in that area as well. When you do this, the algorithm rewards you.

By simplifying and streamlining the campaign structure as much as possible, you reduce internal

conflicts , and because you significantly reduce the possibility of

your own brand and your own ads clashing,

you can use your budget much more effectively. Let's say your main budget is 100,000 TL, and your

daily budget for a brand is 100,000 TL. Normally, that might be 40-50,000 TL.

While Turkish Lira is being wasted, if

there isn't an opportunity in those 40-50,000 Lira ranges, or if creatives aren't producing good enough results,

the algorithm can't afford to spend money on them. Therefore, this algorithm

needs to understand that distribution processes are being done in advance in this update, and we need to focus much more on creative preparation processes

. For example, in the last two years—

I'm not counting this year much, I'll say 2023 and 2024—the focus in 2025

was mostly on testing and establishing the most accurate structure. So,

if we were working with 10 creatives in 2023-2024, we now

need to work with 100 or 500 creatives. Of course, not every brand can adapt to this. That's why

we're getting more successful results in brands that can adapt.

Because when you look at it, the shooting side can't be achieved at the same speed in every sector

, or the product stock side, or in a way, that's

actually a disadvantage. If there isn't enough demand in your product line,

if you're not getting enough sales on the algorithm and sales side, then

managing that area becomes a bit difficult. Providing more variety requires more investment

. The main solution there is to use AI tools, or AI visual and video tools, more

intensively, and that's what we're trying to do with many of our brands. Yes. So

, when we ask what Andromeda really is,

we can say it's about showing relevant ads to the target audience.

Showing relevant ads is fine, but

the listeners need to understand it very well.

Showing relevant ads at the relevant time and with the relevant product. Therefore, your creative needs to directly touch that person.

They need to say, "Yes, exactly," when they see that ad. To achieve this,

if you provide a lot of variety, you need to increase the number of those "yes" oils.

It's not like the previous scenario. In the previous scenario, we could do these things individually

by spending money. Now, that's not necessary. The previous scenario

doesn't equal being completely wrong. These two need to be clearly distinguished. It's still correct, but

it's not necessary. If you've worked correctly from the start, if you're sure about your creatives and, more importantly, your marketing

efforts—who you're showing what to, when, and how often—and

if you've produced and created these efforts accordingly, then

there's no point in burning your budget or using it less effectively.

Technology has advanced. We're still trying to do business with old technology.

It's like we're trying to insist on it. There's no need for that.

And the timing part is very interesting. What's the technical aspect of this? How does Andromeda

understand this? How does it find out when to show it?

I can say this about this: We already created a target audience beforehand, for example

. People who follow Zara or X trend,

people of this age and gender, or similar audiences, or a 1%

similar audience to a specific audience. Here, what we were trying to do was

to get the algorithm to do it, both with the optimizations we did

and the campaign structure we created. Right now,

because they've integrated AI into the ad distribution processes

at Andromeda , the moment you upload your creative, it already analyzes it and

the algorithm knows who that ad is suitable for even before you show it. By the way, these

aren't just my personal opinions. Of course, Meta

had plenty of webinars during this process. Both the information there and

the hundreds of thousands of tests we've already conducted have led us to the conclusion

that forcing the old methods won't yield any results

. You might say, "Now, marketers

really need to do marketing. They need to find those personas."

Well, that's how it's always been. We were working that way, but largely

because you inevitably see advertising accounts. Even if they're not our clients,

you come across accounts. I see accounts.

Many of them are still doing things the old way. I can say that comfortably. That's why

we decided to film this series and explain it. Here, we're spending a lot of our time on this creative

thinking process, marketing processes, understanding the target audience, understanding our product,

understanding personas according to our product, extracting the right angles, and the funnel...

We need to approach

this issue by providing that creative diversity in our steps. What did you say at the beginning of the last question? You mentioned time. That

time is already

being achieved thanks to this GPU, the Hopper GPU, Nvidia's Hopper GPU, even before it's shown.

Yes. It's clear why the price of graphics cards has increased. [laughs]

You know about Nvidia's growth in the last 3 years with artificial intelligence

. You mentioned the meta seminars earlier.

There, they say creative equals new target audience.

Yes. So what exactly is that?

What exactly is it? As I said earlier, your creative

should clearly explain which persona, which problem, which person it addresses.

Therefore, while diversifying this creative pool,

instead of focusing on the general public, we should focus on very niche target audiences in our creatives and

our messages, and we should work on our creatives accordingly

. In my opinion, our direct targeting

shouldn't be based on this or that. Instead of targeting

a very broad audience, like the trendiest summer dresses of 2026,

we can focus on segments within the audience that seeks trendy dresses, while providing this variety. For example,

the trendiest ruffled dresses of 2025, patterned dresses, or

the trendiest satin dresses of 2026 – each

appeals to a different audience, right? So, perhaps we should target

more niche audiences in our creatives, even including our product purchasing and production processes

. What did we do before?

We tried to find this in the target audience section. There was the question, "Who would buy this product?"

We created a target audience based on that question. For example, we'd say, "

Show these creatives to 1% of the audience that bought a ruffled dress from our site in the last 180 days

." Now, without needing that, we can think of it as a scenario where all the

dress creatives are together,

offering each one individually and leaving the targeting part

to the algorithm, using a broad advantage targeting approach. I don't know

if I've explained it clearly enough. So, when we're multiplying creative content,

we're focusing on quality rather than quantity. Both are important;

quantity is crucial. Extremely important. Because think about it, as we said,

it's not enough to just say "trendy dresses." It's not just these dresses, or those dresses

. This time, there's another audience focusing on fabric type, another

message. Each of these messages has different funnel steps,

air stages.

There's an audience completely unaware of what you're saying ,

people who are directly experiencing the problem in their own world. There are those who are aware of the problem, or

another audience that is aware of both the problem and the solution. Therefore,

it's necessary to develop separate narratives for each. This doesn't mean, of course,

that you'll work so well in a best practice, perfect scenario, that

you'll create so many creative variations that it's the most beautiful, most perfect scenario. As

I mentioned in the meeting last week,

in many of the big brands, the number of active, opened, and closed ads in their ad libraries has increased absurdly in the last 5-6 months

. It has increased absurdly, and

we have a similar rate in our brands as well. This actually stems from this: How much

creative power do you have to appeal to a large audience? Of course,

when you look at Nike, when you look at Adidas, they have the power to produce those creatives.

They can make that investment, they can provide that diversity, or they already

have many in their archives. Therefore, if you proceed knowing what you are doing,

you will get the most effective result in the current algorithm. If your knowledge level is not yet sufficient in these areas

, or if you don't yet know how to create creatives based on personas, A-stages, marketing messages

, funnel steps,

I think it is still more

logical to continue with the old method. You need to know, but

you should always know business, but you haven't adapted to this situation. What I see is that maybe in

a shorter time, maybe a year, let's say 6 months, we need to learn these things effectively

. We will explain each of these things.

We'll explain the topics one by one. You can just say, "

Hey guys, maybe instead of going straight to Q&A, let's

explain topic X. Maybe that would be more sensible.

Let's continue with the videos in a conversational style, but if something more intense

needs to be shown, let's show it on screen." It can be complex too.

We thought about it a bit more, so let's start like this. Because it's easier to consume.

You can even consume this video on your way to work. Because we're explaining something with voice

. The other one will be more like an educational video, so it

's harder to consume. It requires a different focus, etc. Please ask your questions in the comments below the video

. Speaking of difficulty of consumption, you're talking about it in detail, but

will campaign designs also become as difficult as you say?

On the contrary, campaign designs need to be simplified. Because, as I said,

we don't even need to compete with ourselves anymore while doing these tests. We create a

creative, work on it, and it doesn't spend any money. Just delete it,

it doesn't matter. Even if it did spend money, it wouldn't bring anything significant anyway.

We've tested this dozens of times. 100% yes,

it's not that a creative who doesn't spend any money doesn't generate any sales, but I'd say that's the case 95-96% of the time.

Therefore, at this point, the time you spend on advertising

needs to decrease significantly. The time

you dedicate to marketing efforts, targeting your audience, creating creative content, or

introducing new products

needs to increase considerably. Actually, it was like this before too. There's not much difference.

Yes. Great news for those who like simplicity. So,

can you briefly explain the new approach to advertising in a single sentence?

Just one sentence? Just

one. Then I can say this: We are now in an era where brands that best understand their customers,

can generate emotional needs, and can build a community

will have more successful marketing efforts

. That's a long sentence, but it's just one sentence. So, in short,

for the first video, we'll outline the main points, covering the before

and after, etc., and

we'll need to talk a bit about marketing, which is something we love doing. As you know, in our team meetings,

we've been focusing more on testing and understanding advertising techniques for years

. Because

technology changes. It's always been like that. I even made a

video a few years ago about how meta-ads, especially

with the arrival of iOS 14, would go to a completely different place in a few years.

The acceleration of artificial intelligence has further accelerated these processes. That

's good news for us. We just

need to interpret it correctly and play by the rules. It shouldn't be like the old days where a creative works, and then you just use

the same creative for product B, product C, product D, and all the math

is based on that. On the contrary, each

creative needs to be diversified. Diversity is the most important thing here, the most

important word. Knowing diversity and personas well.

Knowing the brand's personas or the product's personas well and

distributing them correctly across the funnel steps. Reals videos, visuals, catalog

ads, influencer campaigns, UGC videos, product

test videos, etc. By providing plenty of this diversity, you have the chance to improve your brand's advertising results much more successfully

. This

[music] was never possible before. Thank you.

They call it marketing, but they're just marketing. [laughs] Thank you.

See you. Next [music] videos. See you.