Growing Ecommerce – The Retail Growth Podcast

AI's Reality Check: Financial Bubble or Foundation for the Future?

• Smarter Ecommerce • Season 4 • Episode 9

Share episode

Welcome to Episode 9 of Growing Ecommerce! In a world buzzing with AI hype, we're cutting through the noise to ask the big question: Is this the next dot-com bubble, or a long-term, foundational shift for the future of business?

Join hosts Mike and Chris as they debate the financial health of the tech giants driving the AI revolution and challenge the notion that AI has delivered on its promised productivity gains. They'll also reveal some mind-blowing data on Google Shopping's mysterious lack of market shock when Amazon's ads disappeared, and expose the "hellscape" of Google Merchant Center account suspensions.

Tune in to get the full picture behind the headlines and learn what’s truly happening in the world of ecommerce.

About smec (Smarter Ecommerce):

At smec - Smarter Ecommerce, we specialize in transforming business goals into optimized ad campaigns. With over 16 years of experience in Google & Microsoft Ads, our intelligent software and expert services help retailers achieve superior results.

We're committed to giving you the tools and insights needed to stay ahead in the ever-evolving world of digital advertising.

Make sure to follow smec - Smarter Ecommerce for more performance marketing insights:

Speaker 1:

Welcome to Growing E-Commerce. We are your hosts. I'm Mike Ryan, my name is Chris. We're really glad to have you back for another episode. Hello sir, hey Chris.

Speaker 2:

Oh, what a pleasure.

Speaker 1:

Chris. I'm going to address the elephant in the room. Okay, I noticed something. Anyone who listened to the last couple episodes Chris was in the middle of a stressful move, yes, and his closet wasn't set up.

Speaker 2:

No, not at all, but it looks like you're doing, I'm getting there.

Speaker 1:

You're getting there, I'm getting there.

Speaker 2:

It's not there where it should be, but I'm on my way. All right, I feel way more confident today. All right, good, good, I should feel confident. Right, a big debate about to happen.

Speaker 1:

Yeah, let's see we're going to continue a conversation, the last time about the AI bubble or lack thereof. Yeah, what's?

Speaker 2:

your take on it For the ones who didn't listen the last time, because we kind of did an intro to make a statement where we are at.

Speaker 1:

Yeah statement where we are at yeah, I mean you're right.

Speaker 2:

Did you change?

Speaker 1:

your mind, or are you still? Well, there are so many dimensions to this and no one has a crystal ball. And I'm always mr nuance, so you know let's be nice.

Speaker 1:

But, like chris, this is the stuff, this is I'm. I think a lot of ai attention has been directed at like work, work for placement, topics like this, productivity gains, all kinds of stuff. Recently, salesforce said they let go of 4,000 customer service roles, but let's see where that goes, because who was it? Was it Klarna? Who did the exact same thing and then they had to backtrack it massively. And so I'm always wondering about this and in general, I think a problem that we face wow, this is we're zooming out here. Hold on, let me try to keep this brief. But you know, productivity has been flat for the past couple of decades, despite the boom in, or maybe because of the boom in, software and like all kinds of productivity software, there's no shortage, and yet productivity is kind of flat. Kinds of productivity software there's no shortage, and yet productivity is kind of flat.

Speaker 2:

And meanwhile we have like this aging population and we need it populate productivity to keep growth.

Speaker 1:

It would actually be. People worry about job replacement. But actually, ai, I would love to see it boosting some productivity. But this is one of the biggest reasons why I'm a bit bearish is we haven't seen this and it's always like, well, the next six months or next year or next year, and I've heard that for a couple of years now, or a few cycles, and maybe it's really true, we don't know. But I'm just kind of that's. What I'm missing is the bigger impact here, chris.

Speaker 2:

A strong statement which I don't really want to fight, because what we see at a very small scale and I think you can agree on it in or with our company is we're roughly 100 people. We want to be an AI-first company, which I think we truly are, but we see it on a daily basis that these low-hanging fruits we thought are like everywhere with AI they're not that low-hanging. I give you that. I think AI is not that easy to be implemented to really have a massive ROI impact or productivity impact right away. I give you that. I think it's harder than expected. Some big companies have to crawl back to their old ways of work because they were too bullish. I give you all of that when I'm coming from and that's why you're absolutely right in your statement that there are so many nuances to it.

Speaker 2:

What is a bubble? How do we define a bubble? I'm rather coming from the financial perspective. Are we in the midst of a financial bubble? Is the I dream way overinvested and is it about to burst? And I think we are not even close to that and I have some facts with me which I think we can discuss from a financial perspective and we can then jump back to the operational layer of this whole AI question.

Speaker 2:

But from a financial perspective, for instance, the last bubble we really had was, I think, 2000,. The dot-com bubble. The Standard Poor's was trading at the price-earning ratio of 25, 32, sorry, standard Poor's is trading today at 25. So just from that very meta level, 30,000 feet view, I think that there's not that much overinvestment going on. And I mentioned it the last time. This whole AI game is now driven by fundamentally sound companies. Right, we have Microsoft, we have Apple, we have Google, we have Tesla to a certain degree. These are companies which are fundamentally sound. They are not overvalued. We can discuss about Tesla, probably. But if you look at in media Microsoft, these are fundamentally financially sound companies which have a massive capex going into ai. Yes, they do and they have.

Speaker 2:

But the companies which are driving this ai game are way healthier than, for instance, the last big bubble we had in the early 2000. This is one thing why I, just from a financial perspective, I think it's not the prototypical bubble to look at and I think we are just at the beginning, because the return on investment gains. They will come. That's what I really, really believe in, one thing which I give you also from this financial perspective, where I think even the big companies have to be a bit careful.

Speaker 2:

I had a look at it's quite interesting at the CapEx to EBIT ratio for the biggest companies of the standard pool and that rose from 0.5 to 1.3. So the companies are investing way more into something where we not fully know how this will play out, which is the AI vision, and I think that there might be some risks that some companies are slightly over-invested and the return on investment gains are not coming immediately. But again, the companies which are driving the AI game are so fundamentally sound that I just don't see a bubble from a financial perspective. What's your take on that?

Speaker 1:

I just feel like, at this point, this is a bigger part of the economy than all of consumer spending is, and this is insane to me, and it feels like we're moving from conversations like we need to keep consumer spending rising to conversations like we need to keep compute rising, and that feels odd to me, yes, particularly where we just don't quite know where this road is headed. Yep, and what happens if we reach a point where you know if there's an AI, winter or more like… by the way this will happen.

Speaker 2:

By the way, I think that the path will be a tough one and, for instance, to jump in on that, Mike, the market evaluation of this AI market industry is right now at a whooping $371 billion, which is rather big, but the projections by the end of 2022 will be $2.4 trillion, the 2032 will be 2.4 trillion and I think the whole CapEx, the whole investment, the bullish investment which is happening now is, of course, based on something which is not materializing now. I fully give you that. Will this return mass gains come? Will this productivity gains really happen? So I even think that there might be some corrections.

Speaker 2:

I give you all of that, but the bubble, as far as I understand it, that you know like the industry, will implode. The companies which are tied to AI will go bankrupt. I just don't see it happening, just from that perspective, that the companies are just too fundamentally sound and they know what they are doing. Compared to this small, fractured, structured dot-com bubble where, I don't know, thousands of companies were jumping in on, I think it's now way more consolidated. There are a lot of startups don't get me wrong which are benefiting from AI, but the major companies driving this are super sound and I believe that they know what they are doing.

Speaker 1:

Yeah, I mean, there are a lot of startups out there and we'll see what happens with them.

Speaker 2:

A lot of them won't exist in five years anymore.

Speaker 1:

There's like, if you, if you've ever seen this, like classic MarTech landscape chart where it shows every year it's like bigger and bigger how many, and that just like nearly doubled but it absolutely took off. There's so many startups out there, but, yeah, I mean, I think that in a worst case scenario, these companies can withstand this, and that's a little different from you know. I don't think we're going to need like a big tech bailout.

Speaker 2:

Like we had a bank bailout in the housing crisis and not a tarp.

Speaker 1:

But the market is at all-time highs, it feels overall frothy and it feels like a huge amount of that is based on this giant speculative bet where we just don't know exactly what's going to happen, and I think you know that could be just damaging. I'm talking about the whole market If something goes wrong here. I don't know, I follow a lot of financial meme accounts and stuff and you always see these memes about NVIDIA being the last thing that's holding up the whole global economy or something. But there's truth to it.

Speaker 2:

There's absolutely truth to it. I mean that's what I give you. I mean there is a certain concentration risk because the whole market bullishness which is still more or less carrying the markets.

Speaker 2:

risk because the whole market bullishness which is still more or less carrying the markets. It's max seven and almost all of them are somehow tied to this AI vision. I give you that If we're about to witness a bubble or bursting the bubble of AI, then we have probably way bigger issues than the early 2000s. That's what I give you. I just don't see the bubble, but there is a concentration risk and all of the big companies are betting on this AI vision. I give you that One thing where I'm which, basically, is something which maybe contacts a bit your statement in the beginning that you just don't see this operational, low-hanging fruit related gains.

Speaker 2:

There is not this immediate return on investment gains. Productivity gains, yes. Another thing which might be a positive here is that I think a lot of positivity, or productivity gains or return on investment gains, are about to happen or weren't even planned to be happening. Why? Because a lot of foundational investments have been made. Right, it's not about having this one killer app or this one killer AI approach. A lot of foundational investments are happening right now which even put us in a position to realize certain return investments.

Speaker 2:

I think there's something which we must not forget A lot of stuff which has been happening the last couple of months, maybe years, was foundational investment, building the infrastructure. That's something which also gives me some confidence that these huge institutional investors the Mac7, they are really embedding AI in their infrastructure right and it's just not that easy to kill that, to cut that away. It's embedded in everything we do. That's another reason why, or argument why, I think it's not a bubble, because it's just way too embedded, embedded in everything we do. That's another reason why, or argument why, I think it's not a bubble, because it's just way too embedded already in everything we do and everything we're building.

Speaker 1:

Yeah, I mean, I just I definitely see that they've done that work of embedding and they've made that investment. But I'm also like it's left and right inside of Google Workspace that we use and some of the features like median transcriptions Awesome, that's great. Some of them are very much in search of a use case and I just, I don't know, I have this feeling that we're going to go through the classic hype cycle where we come down to this. Maybe and maybe I'm already in there, maybe I'm excessively pessimistic I think it's going to actually take time for the use cases to really be discovered it's going to take time for this to really trickle through markets, Like there are still companies operating with fax machines today there are some.

Speaker 2:

Yes, there are some. I know one, for instance yeah, a beloved client of ours. No name dropping, no name dropping, no name dropping.

Speaker 1:

Do we fax invoices every month? No, no, no.

Speaker 2:

It's happening in a digital way, anyway, yeah, on that, by the way, mike. Just one final.

Speaker 1:

Yeah, how do we wrap this up, Chris?

Speaker 2:

What I would like to wrap up with is what we are talking about here now about the AI vision it. What we are talking about here now about the the I, the ai vision. It's very much, very much tied to to the digital world. One thing which is, I think, hardly priced in is what is going to happen with the eye in the in the physical world, in the real world, talking about robotexes, talking about robotics. I don't see that priced in in any way, shape or form, and this is by far also to the big companies which are driving this. It's probably the even way bigger market than the software or digital AI game. It's the physical AI game.

Speaker 2:

Right, I'm not envisioning an AI robot scenario. It might happen, but I think we are about to really have a massive disruption in what AI can take over in our real, physical world, and that game is not priced in at all. So I think there's way more leeway. You couldn't convince me, mike, honestly, that we're in a bubble. What I give you, however, is that these immediate return on invest cases which I think everyone envisioned. I think it's not that easy. I think way much work had to be put in, or have to be put in to make AI really a valuable asset, but I believe this is the next big thing for mankind and I think also from a financial perspective, we are just at the beginning.

Speaker 1:

Well, let's leave it on that note. We'll leave it on that, I'm going to give you the final word on that.

Speaker 2:

We'll talk about that again, maybe in one year. We know more about that. Yeah, after the biggest financial shock in human history. Yeah, but I won't show up then.

Speaker 1:

Exactly. Either in a year from now, there will have been the biggest financial shock in history, or I'll be saying, Chris, it was another bleeping year went by and more promises deferred. We'll find out. Yeah Well, let's move on. A quick follow-up from a topic we discussed in a past episode. We're talking about how Amazon left the Google shopping auctions.

Speaker 2:

You have some insights here. Yeah, more insights. Exactly, you have some insights here.

Speaker 1:

Yeah, More insights. Exactly so they returned everywhere except the US and as of time of recording, that remains true. But what I want to talk about, because the big question is kind of so what? What's the impact?

Speaker 2:

There had to be an impact, Mike, please. There had to be an impact, right.

Speaker 1:

Yeah, it had to. What if I told you Come on, there had to be an impact, right, yeah, it had to. What if I told you Come on, there wasn't Come on. I mean, in a way, no impact, basically nothing. I mean how?

Speaker 2:

Tell me how.

Speaker 1:

But there wasn't really impact when they left and then, and so, and maybe in some strange way, it makes sense that there wasn't such an impact when they came back. That's a consistent statement, yeah, but I just Still surprising. I just felt. Like you know, google Shopping's largest advertiser turns off all their ads overnight, there should be a huge shock to CPCs, 1,000%. There wasn't. But then you think, okay, well, maybe when they return like you know, teemu's not going to necessarily yield the ground and like then it must apply price pressure on this side.

Speaker 1:

So what if there was no?

Speaker 1:

dip when they left, but now there's a step up. But I don't know, it could be a flaw in my methodology here, but I have the feeling that this will bear out in conversations, in real world conversations, because that's what happened last time. Last time, I looked at the seven days after the change, after they turned off their ass, and I took you know, you know, so that you know Monday, tuesday, wednesday, however, and I compared them to the median of that weekday year to date to see how this varies or doesn't vary from a typical Wednesday. And there was no variance. And now I applied the same methodology.

Speaker 1:

I compared the seven days after to their year to date medians for that day week and it's the same methodology.

Speaker 1:

I compared the seven days after to their year-to-date medians for that day a week and it's the same thing. It's crazy, and last time that proved to be a pretty okay methodology, because what we heard time and again from advertisers is that nothing changed, and we heard it from other sources as well. It didn't seem like there's a huge impact or as much as expected, so I don't know how to account for it just to paraphrase what you just said, because I think it's worth A.

Speaker 2:

We will monitor this data anyway ongoingly. But just to paraphrase what you said, the probably biggest e-commerce player I mean we don't have stats, but I think Amazon is probably the biggest Google ad spanner there is- yeah, probably yeah, I think so.

Speaker 2:

They abruptly leave the auction and they come back basically in all of the markets within a day. Yeah, and there's no shock to the system. So all crucial KPIs, impressions, especially CPCs, there's nothing crazy going on that you might have a story to tell that, yes, this is the day where Microsoft shut it down. This is the day where Microsoft shut it down again.

Speaker 1:

Yeah. You don't see anything there Correct and I mean this is shocking to me, it's weird. I mean it could like maybe it just speaks to the amount of control that Google has over this fully automated system. Yeah, we should talk about that, that they're able to somehow absorb this or modulate it, or what should have happened.

Speaker 2:

Let's see Amazon exits the auction. What should have happened, According to?

Speaker 1:

your there, I would expect a system shock. You know, basically, there's all of a sudden there's a huge amount of ad inventory available and you know the question is, how does that basically that demand? How does that match up with the advertiser supply levels, budget, unspent budget that's sitting there ready to spend, and and these are limited, really, like they could, they could flow in, flow into and compensate they could flow in and compensate for that but not a one-to-one ratio, because there was a reason why this demand was latent because they weren't willing to bid as high as maybe amazon.

Speaker 2:

Exactly it was suppressed somehow. So the bottom line is we would have expected some dip in the CPC. Well there's. I don't want to put you on the spot here, but I think this is the interesting question, right.

Speaker 1:

No, no, we would have expected that. And I mean you mentioned that they couldn't have held that spot because they weren't bidding high enough or whatever. It can also be because Google looks at multiple factors, not just the bid, it's not just purely an auction.

Speaker 1:

They're also looking at the predicted click-through rate and stuff like that For the website and stuff like that. And Amazon is very strong on pricing, on delivery. They've got their reviews in order, all their review data, and these are all things that, in addition, could support their dominance in addition to that. And so it could also be that the bids were high enough but amazon was winning purely on a quality.

Speaker 2:

Yeah, that's one hypothesis, but they had to be massive, latent, uh, massive, massive, and and one thing, which this is the probably the the weirdest thing to me because I can't drop any names, but we are working with a very big book retailer and a couple of years ago there was also a shock to that very specific industry when Amazon they were basically shutting down their ads in one specific country, their shopping ads back in the days, and they basically activated it again, basically activated it again. The shock to just the auction environment in this industry of books, mate, you can't imagine. You can't imagine. It fucked literally everything up. Every crucial KPI of this very big book retailer changed immediately when Amazon entered the auction again. So this is. It feels very weird to me. We talked about also that this, this impression pike again. So amazon activated all their campaigns and like they were never out. I have no, no, no conspiracy theories going on here, but no, just from an observing perspective. It's weird, yeah, I mean it's weird, yeah I.

Speaker 1:

I don't know what more to say about it.

Speaker 2:

We can leave it like that.

Speaker 1:

Yeah, listen, I could be wrong, my methodology could be wrong. Others can look at this and see what they can find, and also, you know, just on a, basically people advertisers in their accounts.

Speaker 2:

See what I love to hear what people find by the way, shout out to the listeners, guys.

Speaker 1:

If you have everything or anything which maybe supports or contradicts what what we just discussed, we're super happy to to talk with you about that just fire, fire your your findings? Definitely, I love to be contradicted. I love to contradict myself. Love debates, yeah. Yeah. So contradict me, make me smarter, but for right now, that's what we're seeing.

Speaker 2:

It's a bit weird up to the next one. On to the next one. Next topic, actually our last topic today. We're seeing it's a bit weird. Up to the next one.

Speaker 1:

On to the next one Next topic, actually our last topic for today we're going to talk about Well, chris, do you hear about merchant center account suspensions often? Yes, yes, here and there. How pleasant is it for advertisers when their merchant center account is suspended?

Speaker 2:

I would say there's not a lot of things which are more critical to online retailers than to hear that even whispering sound of a potential merchant center shutdown.

Speaker 1:

Why are you asking? I think it's a rather bad thing, but it's super easy to fix the problem right.

Speaker 2:

It's super easy Just one phone call and everything is up and running again. Yeah, exactly no, just kidding. It's the rather bad thing, but it's super easy to fix the problem right. It's super easy, right. Just one phone call and everything is up and running again.

Speaker 1:

Yeah, exactly no, just kidding, it's the worst case, yeah, and we're dealing with a combination of automation and bureaucracy that creates a hellscape for advertisers.

Speaker 1:

I can remember a large US sports brand that we worked with and actually their main channel is Amazon. They spend a ton of money on Google Shopping but they spend even more on Amazon and actually Google should be really working to make an advertiser like this happy and bring some of that investment over to Google Shopping. And they had a merchant account. They had a suspension, an account suspension that it just took ages and ages, and ages to fix.

Speaker 2:

You know the reasons by the way In that case, yeah.

Speaker 2:

No but the reasons why it just takes so long for Google, because I was in the midst of some merchant center shutdowns for sometimes the right reasons, sometimes the wrong reasons. But as far as I know and this is a structural thing with Google, as far as I know, the team which decides, or basically takes the decision, if a merchant center is put down yes or no, or can be reactivated in these cases, is completely detached from basically the teams who are working with the clients or partners like us. And this is the crazy thing, because the teams which are supporting the retailers, of course they do everything they can, but it's structurally completely detached and, you know, maybe there's a good reason for that, because you know this case I'm just talking about.

Speaker 1:

actually, that's a conflict of interest. Like Google, it would be in their financial interest to look the other way, or is this true? Like in that particular case, I think the suspension was erroneous and that's what I mentioned. But I agree, they're isolated. They're structurally isolated, but I also think there's a tremendous backlog and we're about to talk about some numbers here and also I think that then there's a huge amount of individual variance that comes in.

Speaker 1:

There's a lot of judgment involved and yeah, but I want to give you some numbers. Drop some numbers on you because Google, like all large platforms, large online platforms in Europe, they're required by the DSA to share moderation data with Europe and Europeans daily and we talked about some of that data in an early episode.

Speaker 1:

we were looking at content moderation on meta. So this is that same database. What I noticed back then, and I've been meaning to circle back to it, is that Google Shopping is the most heavily moderated platform in Europe and I think we can probably say in the world I think, or I don't know, maybe it's different. I won't make any jokes about China right now, but in the Western world, probably the most heavily moderated platform that there is and, just to put that in context, the second most heavily moderated platform. Any guesses?

Speaker 2:

Has to be Amazon.

Speaker 1:

It's Amazon that's right, and also a vast quantity of products and sellers.

Speaker 2:

And yeah, of course. So it's Amazon, yeah.

Speaker 1:

So it's very fast and you can understand how this loans itself to having a lot of moderation. But Google Shopping is about seven times more moderated 7X 7X Compared to Amazon, compared to Amazon. So I, holy shit. Moderated 7x 7x compared to amazon. Compared to amazon. So I and I, holy shit. The question is, we'd have to look at, like you know, their volume of merchants to, are they don't have seven times more merchants, or?

Speaker 2:

that's what I wanted to ask. So the 7x sounds sounds, yeah, honestly crazy to me anyway. But which indications could we look at that the 7X might be?

Speaker 1:

okay.

Speaker 2:

What is it?

Speaker 1:

Number of products Number of products on the platform number of merchants I would say so yeah, we can look this up maybe if this makes sense, if there's any data available, but 7X to Amazon, I mean when we consider the amount of unmoderated bootleg products and knockoffs and all kinds of stuff that's on Amazon. That, I think, is.

Speaker 2:

Yeah, it doesn't sound okay to me.

Speaker 1:

I actually think that Google is running a tighter ship than Amazon.

Speaker 2:

I think this plays certainly into that. Do you have some numbers?

Speaker 1:

What this means in practical terms. In Europe alone, there are between 30 million and 70 million products every single day that receive limited visibility due to policy violations, real or perceived, and this results in let's get back to the lead here about your account being suspended. Well, there are 14,000, 13.8 on average, 14,000 account suspensions every single day in Europe, every single day on Google Shopping, yeah, and so that's why it might take you a while to hear back. Wow.

Speaker 2:

So maybe it's not about the structural near separation of the teams. There are too many, there's just there are too many.

Speaker 1:

There are just too many to handle. Yeah, I mean they also have stats about the level of automation both in detection and decision-making here and like 99% of detection is automated, more than 99%. Fair enough, which makes sense.

Speaker 2:

But to reactivate it. You basically have to go through the team and if you have 14,000 a day, can you imagine it's insane. By the way, what would be very interesting from a shareholder perspective? I don't have any Google shares, but how much ad spend is basically? What's the missed opportunity here?

Speaker 1:

Yeah, there's for sure. There's an opportunity cost to that for sure.

Speaker 2:

And the question then again is, of course, how many of these 14,000 are suspended for the right reasons and which are the wrong. I for sure all the account merchant center suspensions we faced with our clients, most of them were for the wrong reasons, honestly speaking. So that's a stupidly high number.

Speaker 1:

Yeah.

Speaker 2:

Didn't expect that.

Speaker 1:

It's a lot. And, by the way, about the decision making 62% of decisions are fully automated. Fully automated, but that leaves and there's pros and cons to this, I mean it leads a lot of decisions to be made by humans.

Speaker 2:

I hope it's solved by AI in the near future.

Speaker 1:

Yes, To make better decisions.

Speaker 2:

But let's see, I don't want to commit to that.

Speaker 1:

You got me, chris, you got me. I think that's going to be our final word. Final word Christian Scharmweiler got me.

Speaker 2:

No, no, I didn't get you. I enjoyed the podcast, however, me too, it was a pleasure again, mate.

Speaker 1:

Thanks for chatting with me. Chris, all right, that's another episode of Growing E-Commerce. This podcast is brought to you by Smarter E-Commerce, also known as SMEC. You can learn more at smarter-ecommercecom. As always, please leave a review, leave a rating. Recommend us to friends helps a lot. Thanks for listening and we'll see you next time.