What is the reason for the traffic loss for independent news sites? Google Search spokesperson explains

Google has caused an uproar among digital publishers with its non-transparent algorithm changes. Google has violated competition rules by shaping the 'Discover' section, which should be shaped according to user habits, through its own impositions. On Google, many news sites do not even appear as the first suggestion to the user when searched by name. Publishers have called on authorities to put an end to Google's anti-competitive practices.

Eray Özer from T24 spoke with Danny Sullivan, the international spokesperson for Google Search, about the major decline experienced by independent news sites following the algorithm change.

While traffic is of great importance for independent news sites, as the number of visitors drops, ad impression counts also hit rock bottom, and advertising revenue—which constitutes almost the only source of income for publishers—decreases to a level that makes it impossible to continue.

Here is what Danny Sullivan had to say about Google Search's algorithm problem...

I want to start here: As artificial intelligence continues to develop, do you think Google Search will soon understand us more semantically and that we will start getting more detailed results on the first page? In other words, could search engines start working like Large Language Models after a while?

We have actually been trying to go beyond matching words for years. I think we have managed quite well, too. Sometimes people search with different words than the text they are looking for. For example, if you are trying to fix something or trying to get information about an illness, you use a different term than a plumber or a doctor would use. We have been using artificial intelligence and machine learning for a long time. Probably for over a decade… There are also features like the “AI mode” that we have recently started testing. And I think this will continue.

If people expect searching to evolve in this direction, then as a search engine, we need to listen to this demand. However, there is one more very important point, and as Google, we are very, very aware of it: you can only do this if the underlying web ecosystem is truly successful, that is, if there is a wealth of knowledge. Therefore, what we are trying to do is to find a way to bring this explanatory part to life, while at the same time keeping sources or more information accessible to people and allowing them to continue from there.

We think that in this way, search will continue to grow and the web ecosystem will also continue to develop. Because if you do not cite sources and do not enable people to reach the sources, the web ecosystem will not develop either.

When you start sharing content, do you have a plan to pay royalties? For example, OpenAI made a deal with NewsCorp. Is Google considering paying royalties to its content partners at some point?

I mean, this is an administrative matter beyond my scope. It would not be right for me to talk about copyright. But I can say this: For example, having a licensed agreement with a source does not mean that source will appear at the top of search results. We have similar agreements, and the only thing that is certain is that even these agreements cannot affect the ranking. That is not how we work. So, we do not want to make agreements where people have to pay to be displayed in a better place. That is for sure.

Let's come to the famous algorithm changes… This is a topic that is also heavily debated in Turkey. We have experienced two major drops in the last year. One in August 2024 and the other at the beginning of this year… The drop here was not something that could be described with just the word “drop”; perhaps the right word would be “collapse.” Because the numbers hit zero, especially on Google News and Google Discover. What was the reason for this?

It is difficult for me to go into detail for any specific site. However, speaking more generally, I can say that we made a ranking change and determined our steps globally rather than locally. But of course, when we look back later, we may realize that what we did affected some other things in ways we did not anticipate. When this happens, we may conclude that what we did was not useful. Then we ask ourselves: Is there a way to make this better? And we try to understand how we can improve this general prescription.

Again, speaking generally, there is a tendency to improve things in general for a wide variety of queries. So, the situations you mentioned are probably due to such a reason. We are already trying to understand the problem and bring it to a better state after these changes.

Another issue is to keep in mind that there is a distinction between Google Discover and Google Search. Discover is part of Search, and sometimes a lot of traffic comes from there and then it can disappear. Then it comes back again. So, you may not have met some criteria for the content I will highlight, and you become invisible there. Then you meet these criteria on another topic and take your place there again.

For this reason, we generally tell media outlets “do not rely on Discover.” I understand that this creates a more difficult situation, especially in news publishing. Because you get significant traffic from there. But it is a much, much more uncertain area compared to the Search section.

Also, we are increasing content options for Discover. When we started, content was limited to news only. Then it expanded. And of course, there are also those who experienced the exact opposite after these updates. We need to look at that too. Although you said this situation applies to news sites in general, but…

Finally, something else strange that we could not foresee might have happened. It is possible that we will get better performance from the settings here when we go over this with the next update, in this respect.

But you understand the situation, right? This is not unique to T24. Many publishers, especially independent publishers, have been negatively affected by this. There have even been news reports about this.

I understand. But it is impossible for our goal for this update to be to block independent news sites. From where we stand, those media outlets are just websites. The system does not perceive these media outlets as “Oh, this is an independent news site. This one is not.”

The system makes a ranking by taking multiple signals into account. The platforms you mentioned may not be meeting some of these signals. Also, perhaps the system is not working efficiently enough to reward the content of these sites as much as they deserve. This could also be the case.

And then, as I said… Maybe at some point we will say, “Wow. This is a surprising situation. We did not expect this.”

So, in such major changes, we really know that some things are going better and are useful in some points. Because we test them beforehand. If it does not provide an improvement in general, we do not release that update anyway. But in such a niche area, an unexpected result can indeed emerge, and we may say, “This is not an ideal situation.” Then we try to reach a decision by reviewing this situation.

There is also a situation specific to Turkey. Especially these independent publishers do not have a financial structure like large companies. Therefore, if there is a need to do SEO here, they may not be able to afford to set up separate teams or technology teams for this. This puts them behind in the race.

We do not want media outlets to think they need to do SEO to be found. I hope these words of mine will comfort you a little bit that this is not our goal. We already have a guide on SEO. Just make sure we can access your content. This is not a very difficult thing either. If you are doing this, you do not need to do much else. Although sometimes people block us without realizing it. This is also a situation we encounter. If we can access your content easily, you do not need to do anything special for SEO.

Even if you do not like sitemaps, we do this for you. Maybe it would be useful to pay attention to a few structural things on the sites. That is all… Other than that, continuing to provide the right content for readers is enough.

Also, think of it this way: Until these recent glitches, large companies doing SEO did not have that much of an effect. So the problem is not in not doing SEO. What happened was related to us making a ranking change, and this change did not yield results for content that did not meet certain types of signals for a group.

I would like to say again: What I have described does not prevent us from pursuing the situation that occurred. Something did not comply with the change made. But this is not a situation related to doing SEO. We constantly need to make some changes to understand your content correctly. That is the issue.

I am not saying that SEO means nothing at all. Do not get me wrong. It might be useful to consider SEO, but it is not a requirement.