← Behind the Numbers: an EMARKETER Podcast
Behind the Numbers: an EMARKETER Podcast
Digital Lessons 2025 — The Triopoly’s Power, the Sleeping YouTube Giant, AI Search Behavior, and More | Behind the Numbers
00:25:17
Loading summary
Transcript44 lines
- [00:00]
A
Consumers skip ads. Those sneaky individuals. But they don't skip rewards. Thank goodness. That'll be madness. Fetch drives performance. With over 12 and a half million monthly active users and 11 and a half million receipts scanned daily, capturing 88% of household spend, your brand becomes the reward. That's nice. Earning real engagement, verified purchases and loyalty. Fetch, it's America's rewards app where brands are the center of joy. Hey, gang. It's Friday, October 17th. Eleni, Yuri and listeners, welcome to behind the Numbers new market video podcast made possible by Fetch Rewards. I'm Marcus and join me for today's conversation. We have VP of Global Content Operations living over in the uk It's Eleni De Galaki.
- [00:55]
B
Hey. Hi, Marcus. Thanks for having me. Hi, everyone.
- [00:57]
A
Of course. Welcome to the show. We're also joined by a principal analyst who heads up our media, tech and ad teams based in New Jersey, it's Yuri Wormser.
- [01:04]
C
Hey, Marcus. Glad to be here.
- [01:05]
A
Hey, fella. Happy to have you. Today's fact. Elevators, or lifts are the single most used form of transport transportation on earth. Now, I'm playing fast and loose with the word transportation, okay, but bear with me. Every day, people take roughly 18 billion 1, 8 billion elevator trips in a single day. That's more than the world's car, plane and train journeys combined. That's too much elevator. Like. Like all transport, there is a lot of waiting around. The average office worker, you'll be furious to learn, spends more than a year of their life just waiting for elevators to come. Destination dispatch systems, like the ones we have in our New York office, works to group folks together who are going to the same place to the same floor to help save time and energy, I guess. Altogether, elevators move the equivalence of the Earth's population, 8 billion people every three days.
- [02:22]
C
That is amazing. Those are amazing numbers. I never would have guessed if you had asked me what was the most used form of transportation. Elevators would have come in about on the third day, toward the end of a wait for an elevator up to our office.
- [02:37]
A
In fact, it doesn't seem like it's transportation's generous.
- [02:43]
B
Yeah, it's generous. But those numbers did make me feel good that I'm working from home three times a week, probably missing more than like two, three months a year on this transportation.
- [02:59]
A
You're welcome. That's. Yeah, the good parts of working from home, save that time. Anyway, today's real topic, what we've learned about digital in 2025. So, Eleni, you and Vlad Hanslick, who runs content for us, just published a New piece of research looking at the must watch trends digital leaders should be paying attention to this year and beyond. I like this piece of research because it shows kind of the main ways that digital has changed this year and then obviously you're looking to the future, but I thought it's just kind of nice to recap a lot of times towards the end of the year we look forward and we're just looking at 20, 26 trends and it's kind of nice just to say, well, what's changed this year? We rarely spend a lot of time looking at from January to now what's different and acknowledging those changes. So I asked Eleni and Yuri to tell me which charts or pieces of data best define how digital has changed the most this year for them, either from this research or from some other place. Eleni, what's one of yours?
- [04:06]
B
Well, the way you framed it makes it kind of inaccurate because my first one looks at how it's changed over the last five years. Marcus? Five years.
- [04:18]
A
Okay. Gone five. Great. Okay. But from five years to today.
- [04:23]
B
Okay, yeah, okay, yeah. And this research is more of a compilation of some of our favorite charts from this year that showcase exactly how digital is shifting. So they all are housed in different reports and are part of other analysts research as well. But I think when brought together they also tell a very interesting story. And so the first one that I have is one that looks at the triopoly share of total ad spend and also digital ad spend. And triopoly is definitely one of our words. I don't know how much we've popularized it, but that's Google, Meta and Amazon. And so what we're seeing here, which is quite interesting, is that their share of total ad spend is rising. It's risen roughly 12 percentage points over the last five years, but it's only increasing because digital is increasing as a whole. So they're riding that digital wave, but they're not actually gaining ground within digital. So in fact their grip on digital is slipping. It's felt it has declined 2 percentage points over that same period, which is sounds modest, but I think it's quite meaningful in that it indicates a more, a less concentrated market, a more competitive market and also I think it's even more stark if we take Amazon out of the picture. Amazon has been riding the retail media wave over that period and their share has actually increased. So if we take that out, I think Google and Meta have seen their share drop by roughly 6 percentage points over the last five years. So I think that's quite an interesting number.
- [06:03]
A
Yeah, a lot happening over that time. Just to key in on what's happened from 2024 to 2025 with regards to those, those players, Google has lost over a percent, 1.2% share. Meta is up a tiny bit, but pretty much flat. And Amazon has gained as well. To what Eleni was saying, Amazon is the outlier there. That's kind of keeping the triopolis share up. You mentioned the words competition, Eleni. Where is this competition coming from the most?
- [06:38]
B
Yeah, so if you see in the, in the graph, there's this other digital that's been growing its share over the same period and it's sort of becoming like a fourth pillar of digital advertising and there are a lot of fast growing channels there and platforms across connected TV, retail media and social. So the likes of Walmart, TikTok, Reddit, Peacock I think are some of those platforms that have been growing the fastest over that period.
- [07:08]
C
That said, I mean we shouldn't overstate it as well. I think it's a totally valid point. But there's still around 60%. I think that's right. There's still around 60% share holding somewhat steady or dipping a little bit. But, but the dips are pretty small and Meta, Amazon and Google are all pretty well positioned to, to grow in the AI era. Google is probably the most threatened just with search, but it is also has some of the best AI technology out there, so it might actually ride the wave and, and do pretty well. Completely interesting and valid that it's, it's, the numbers are going down a bit but at the same time just want to point out that we don't want to overstate it as well.
- [07:53]
B
Yeah, I agree. It's a good point, Yuri.
- [07:56]
C
Yeah.
- [07:57]
A
And the total dollars as well are still extremely healthy. So even though the shares are going down, we have looking at Google here for a second, they're making, in the US we expect them to make 4 billion more dollars this year than they did last year. We expect Meta to make nearly 7, $8 billion more this year than they did last year. Amazon obviously going up as well. So the dollars are looking very good. It's just the share because the pie is growing and they're not growing as fast as the pie. So you mentioned Google here. Yuri, you're one of your major digital trends is to do with YouTube owned by Google. What, what do you have for us?
- [08:39]
C
One of the charts in this report that Eleni and Vlad put together show just the, the, the huge number of minutes that YouTube captures in the US it's the leading media in the US in terms of time spent and, and if you look at, on a global scale, more people in the world view YouTube than any other media, two and a half billion people. So it, it's a huge media in terms of attention. And yet in the U.S. youTube only nets about $10 billion in ad revenues. And if you look at the gross figures, so the numbers that are actually earned by YouTube, some of which shared with creators, that that's only 20 billion. When you look at Meta, it's 73 billion. Google is a little bit more than that. Google as a whole is more than that. So it's, it's a very under monetized media property, very profitable. But I think that differential between time spent and attention in general and ad revenue is going to be something that will probably shrink in the years going forward. It's a significantly under monetized property.
- [09:59]
A
So what's going on there? Because you're spot on. I was looking at this as well. Despite having 20 million more US users than Facebook, YouTube has four times fewer ad dollars. How does YouTube Yuri close this gap and can they, is that something that they're able to do?
- [10:18]
C
I think they can close the gap. A part of it is just through AI targeting. I mean it's always been relatively lightly monetized compared to its attention. I think Google has shifted its attention to YouTube. Some of it is also it has a subscription YouTube TV which you know, has longer programs with less advertising. So both, you know, I think as YouTube figures out how to monetize kind of the longer form TV as well as targeting better with new ad formats and new shoppable ad formats within YouTube over, you know, in all of its content. I think that that should drive up the ad revenues. I'm not sure if it'll ever get to meta's level, but it, but it should shrink the gap.
- [11:13]
B
Yeah, I think there's definitely an opportunity gap there. We have a report, the YouTube opportunity that talks about the sort of, some of the key barriers that are maintaining that gap and it's, I think part of it is how fragmented the platform is. There are many different formats and many different sort of contexts for content, creative content and there's not a lot of premium content. So brand safety is sort of an issue and then measurement is quite difficult and attribution when you have this sort of also fragmentation and complexity of the platform. I do think it's a big gap though and the, there's a real opportunity there and I do expect it to shrink over time, like you already said. And I think maybe with AI adjusting creative to, to fit different types of formats and context might will also become easier, which will be to the benefit of advertising advertisers looking to take more advantage of YouTube scale and sort of time spent and attention.
- [12:21]
A
Yeah, I mean YouTube scale is still growing in the U.S. they're going to add a another 5 million viewers this year. And each of the next four years I think they're adding about 4 or 5 million up until at least 2029 when our forecast goes out to and they're adding a further 60 million 6,0 million viewers worldwide this year alone. So still somehow swallowing up people on the planet and adding them to their user base. Eleni, what do you have for us next? What else do you think people should be paying attention to in digital?
- [12:53]
B
All right, so next I'm moving into E commerce to mix it up a little bit. And there's an interest infographic in our report that shows that the categories with high E commerce penetration will be the ones to increase their lead over time. So looking now from today up to 2029 essentially, and you have some categories with very high E commerce penetration like books and toys and computers and all those categories will increase by at least 6 percentage or up to 10 percentage points over that period in terms of penetration. And then you have other categories that are lower in penetration and will sort of stay low. So the increase will range at 1%, 2%, which is much lower. So that spread will widen. I think it's an interesting story. It might be somewhat intuitive in that it sort of signals that the categories that are largely offline are offline for for a reason. There's some structural barriers that keep them there and that's why they won't shift online as meaningfully as some other categories. And similarly, some other categories like toys or books, like the tactile experience doesn't matter. It's quite easy to ship them. You understand how they work. You don't need to see them to buy them. And so again, it's sort of intuitive as to why they'll be growing their lead. But I think what's kind of most interesting there is that some other categories in the sort of high penetration like apparel, which will grow like 6 percentage points over that period, or even within household, which is one of the low categories, but one that will gain Quite a bit, 4 percentage points, even more if we look at beauty as a subcategory there, they have some of the same barriers that food might have in terms that I feel like the tactile experience is important. Trying clothes on and sort of seeing if they fit or not some of the return logistics. But they've made meaningful strides I think in those areas and they'll continue making potentially AI will also help with that some of the 3D try ons and so on. And that's why they're moving more meaningfully than auto say or food. So it's interesting to see those categories that have sort of broken through those barriers and actually a parallel will be I think E commerce first subcategory by the end of that period like it's going to be 51% online. So it's quite a big milestone.
- [15:32]
A
Yeah, yeah, this one is really interesting to see that because you're saying like the very high, highly penetrated categories, they said books, toys, computers, they are like over 50% kind of some, some of them well over closest, like 60, 70% and they have the biggest growth, they're seeing the biggest growth from now to 2029. And then some of the other ones on their end, food, beverage, they're very low penetrated. They're like probably around 10% or lower in some instances. Auto and parts lower. It's quite interesting, I mean Yuri, what do you make of this? Because on the one hand I understand the food and beverage piece like people like going to the store and a lot of it's produce, it's not all produce. A lot of it is just like it's paper goods, it's things in boxes, it's stuff which you buy all the time. You know it's bottled water is the things that you could just replenish. You don't need to go and see them. Same with auto to a certain extent. You know brake pads are brake pads and I don't think like taking them, picking them off a shelf and looking at them. I don't, you know there's some of these categories, I think they should be more highly penetrated. What, what's your take on this chart?
- [16:36]
C
I mean I think that there are different reasons for auto and for, for food. So for food you're right. A lot of these parish, these packaged goods could easily shift into, into E commerce. The thing is a lot of people will still go in for perishable goods, frozen foods, produce and while they're there they'll pick up the packaged goods. So I think that's probably holding back some of the packaged goods. The E commerce share for packaged goods with car there are all types of barriers. You have dealer barriers. AI could meaningfully change the Experience of buying a car in the sense that you could really do a lot of the comparison shopping. You could almost, if you, depending on how AI develops, you could almost take a AR and augmented reality test, drive up to different cars and so forth. But I think one of the things that's holding it back is people love shopping for a car or oftentimes part of the experience is trying the different cars.
- [17:38]
A
Yeah. Elena, you were saying before we hit record, you were talking about how AI shopping assistance might change some of this category growth. Are there any here that you could see it really affecting?
- [17:50]
B
Yeah, good question. I want to plug Karina's recent report on AI shopping assistance because I fully buy its arguments and its main take, which is that in the near future it won't meaningfully change the penetration because they do offer greater convenience in many ways. They collapse the funnel. We might see an increase in sort of average order values, but they don't really right now address the bigger barriers that have to do with what's keeping in store strong. And again, those are the things like the tactile experiences or the immediacy that I want something now and it's going to be quicker versus waiting on a long delivery window. And also they don't. So some of the barriers, but also some of the things that consumers want. So consumers apparently, like I say, apparently because I don't relate, I do everything online, but the, the interact, the social interaction that comes with shopping as well as not just like that tactile experience and trying something on, but just seeing people talking to people. There's an appeal to that. Not for you. I love talking to you, Marcus.
- [19:09]
A
Okay, thanks. Yeah, this doesn't count your insights.
- [19:14]
B
True, true, you caught me. But yeah, so I think it won't be in English shift online, any of the categories in the near future.
- [19:26]
A
Yeah.
- [19:28]
B
Yuri, would you agree with that take?
- [19:30]
C
Yeah, I do agree. I mean the one area where I think that might get kind of an early impact from these chatbots is going to be an apparel with better try on features. I think as the AI gets better and you can actually try on these outfits to see how they're, how they, how they look without actually going into the store and trying them on. I think that might boost apparel online purchases. But I otherwise, I basically agree that it's, you know, part of, part of what shopping is, is entertainment. And I think that won't meaningfully be changed by chatbots at least in the near term.
- [20:10]
A
Speaking of, of chatbots, speaking of AI, Yuri, your last one is, is in that Space it is.
- [20:18]
C
So I'm, I'm going to, you know, in this chart you can see that OpenAI or ChatGPT has gone from 200 million weekly users in February 2024 to 800 million in October. So right now about 800 million users. So just a huge growth in. And ChatGPT has really entered the realm of a mega app. It hasn't hit that billion user threshold that is often used for that, but it's heading there and it'll probably reach that pretty soon. At the same time, it's still very early days in terms of its impact on advertising. On E commerce, they've introduced a few e commerce tools. Very, a very small amount of referral traffic is actually coming from these chatbots. Now part of that has to do with zero click. You know, you're getting your answer onto chatbot. But part of it is people aren't really using it too much for commerce type of inquiries yet. My colleague Nate Elliott just produced a report, just published a report that showed only about 3% of discovery time, which means kind of search, search and inquiry time is happening on chatbots as opposed to traditional search. That Data is from comScore, so just a really small amount of discovery time. And in terms of Referral, it's about 1% according to SEO Company or GEO Company, generative engine optimization company BrightEdge. So its impact is still small, but the size of the audience indicates that big change is coming.
- [22:07]
A
Why is it still so small?
- [22:11]
C
Part of it is that people are still going to Google for commerce information. They have a lot better and richer data when it comes to what's available in the nearest store, cost comparisons. People are often going to ChatGPT for kind of deeper research. A lot of it, a lot of the time spent on ChatGPT doesn't have to do with trying to compare, you know, ask a question about a product that has to do with themselves. You know, what's, you know, what type of thing is, you know, you know, what is this a symptom of, you know, a health thing. But even more than that, they're using it as a tool to write things, to create things, and even to use it as a therapist, even though that's probably a fairly small portion of it. But there are a lot of use cases for it that aren't commerce related.
- [23:12]
A
Eleni, what are your thoughts on this one?
- [23:14]
B
I think it's also about that it's early consumer behavior where we haven't yet learned to ask AI to search for things that we want to buy that require clicking through. I'm thinking back to when it first launched and it didn't even have like, you know, up to date information. And it's sort of like the way we learn to use it is just for more knowledge based questions. I think still the way that we use it is to inquire and to ask questions is more that's like the primary use case. But it's just as Jori said, it's not about it's like information seeking rather than something that has to do with commerce.
- [23:54]
A
Well, we'll see if that changes because we just did an episode on Wednesday of this week where I was guest hosting the retail show and then we Talk about how ChatGPT's instant checkout changes things. And that was with Karina and Zach. So that's brand new. Maybe that will move the needle somewhat. What Eleni and Vlad's full report is called 10 Charts that Define Digital Markets in 2025 and Beyond. Pro plus subscribers can find it at eMarketer.com link courses in the show notes or all those other reports we mentioned from Karina Perkins on AI shopping assistance, I think we mentioned and then Nate Elliott who wrote about AI. We've got the YouTube advertising opportunity report as well. Tons and tons of stuff there. So head there if you are a subscriber or if you're not and become one. That's what we have time for. Today's episode. Thank you so much to my guests. Thank you. First to Eleni.
- [24:49]
B
Thank you Marcus. Thanks for having me.
- [24:50]
A
Yes indeed. And thank you to Yuri.
- [24:52]
C
Great being here as always. Thanks Marcus.
- [24:54]
A
And thanks of course to the whole editing crew. And thanks to everyone for listening in to behind the Numbers and the Market video podcast made possible by Fetch Rewards. Make sure you subscribe and follow and leave a rating and review if you can. We'll be back on Monday. Happiest of weekends.