Part 1 Video start 0:13 - Google Search Dominates Digital Advertising
Part 2 Video start 8:30 - Google’s (AI based?) Self-Updating Maps reflects failure to engage SMBs
Part 3 Video start 14:47 - The Importance of images in the local search results for rank & conversions
- Google Ads Domination
- How AI and imagery build a self-updating map
- The Visual Future of Local Search - Part 1
- The Visual Future of Local Search - Part 2
Transcript Ep 60:
Greg: Welcome back to Episode 60 of the Near Memo with David, Mike and Greg. And we're talking as always about search social and commerce with the local flavor, local inflection. I was out next week and Joy Hawkins said, and for me, that last week, And I'm sorry, Freudian slip. I was in Joshua tree with a friend.
It was beautiful down there. And in California, for those that don't know where that is. And we're back this week full strength. We'll be talking about Google, Google, and Google, I think, across the board.
But I want to acknowledge that tonight is the first night…. so it's Friday, April 15th when we're recording this. So tonight is the first night of Passover. I believe Ramadan also has begun and Easter of course is on a Sunday. So it's a very unusual convergence of these three holidays. So happy holidays to everybody out there, just wanted to acknowledge that and
Let's then jump right in after that long preamble to this week's stories. And I wanted to talk about the IAB report, the 2021 full year advertising report, which I thought was very interesting. It was really kind of huge year for digital advertising: 35% growth, the highest growth since 2006. The total number was something like 190, very close to 190 billion.
And, you know, making digital advertising the largest medium by far, really now in the United States in terms of advertising revenue. And what was one of the things that was quite striking about it was the strength of search, right? I mean, there's been this sort of perception that over time search would not maintain its momentum and would not be able to maintain its growth.
And it's 41% of digital advertising revenue across desktop and mobile. The next highest category was display at 30%. The highest growth categories were digital audio and video on a percentage basis. But those are much smaller numbers, but it's really remarkable how search has continued to maintain its dominance of digital advertising and, really in some sense, increased it last year.
And that's almost entirely Google revenue, right? I mean, the actual numbers, the actual search revenue number was $78.3 billion and display was $56.7 billion. So really $22 billion more. In real dollar terms, the increase of over 2020 was $50 billion.
So there was $50 billion in increased digital ad revenue, which is a huge number, almost a shocking number over 2020. And another, another thing that's worth calling out here is that about 80% of that revenue is concentrated in 10 companies. That's sort of been true for years, but it really just it's kind of a remarkable thing because in real dollar terms, You know, that's about the $50 billion increase about 40 billion of that was pocketed by those 10 companies.
So massive, massive concentration of revenue. And in particular, in the search category, there's no. Digital advertising category. That is, that is so dominated by a single company as search. And I'm also worth mentioning about 71, 70 2% of revenue was mobile and 28% or 29% was desktop. So there's. You know, I haven't looked at these numbers in a couple of years, but it's, that's another really striking number from my point of view.
I have other things to say, but Mike, it sounds like you want to make it
Mike: Well, a couple of things. One is, is this a function of, I mean, is this a zero sum game? I mean, does this come from someplace else or as you saw in the census reports today or yesterday, you know, we've seen huge jumps in retail and restaurant employment across the United States.
We've seen huge jumps in new business starts for lsat year, whatever. So is this a subtraction from other advertising medium or is it new or both?
Greg: Well, both I think is the answer. Some of it is net new. as you, as you apply with your comments, and if you look at the, the, the, the IB report is, is basically compiled by Price Waterhouse, and they, at the end of the report, they have a comparative chart, reflecting different media channels, internet, TV, magazines, newspaper, and so on.
And interestingly, many of the traditional media channels actually grew last year, newspapers were up you know, three and a half percent. Radio was up almost 13%, which is kind of amazing, magazines were off and a couple other channels were often and out of home was also up. So there's some, some net new dollars in there, but also other traditional media have been bleeding for a long time and that's, and the internet has siphoned some of that away.
So I would say it's a mix of things going on. I spoke at an investor conference this week for Evercore and we were talking, I was on a panel with a couple of other people in the industry, and we were talking about the sustainability of Google's position in search and its revenue growth.
And, you know one of the things that really plays into Google's hands is all the privacy stuff that's going on because search is really, uniquely positioned versus display advertising and other types of advertising to, to benefit from, from tighter privacy rules because it's really, they're almost no privacy implications to straight up search advertising. So, I mean, it's amazing that Google is such a huge company, $2 trillion market cap now you know, and is really essentially a monopoly and paid search and is generating by itself almost 40% of digital ad revenue.
David: Yeah, I was going to say that the growth doesn't surprise me necessarily. I developed a graphic probably five years ago at this point where my hypothesis was that even if search didn't grow as a thing that Google would just monetize more and more and more of each SERP. And I think we've seen that play out with all of Google, especially the Google shopping advances in the COVID era.
I see more and more shopping carousels. We've talked about the inventory situation before. So I think that's also a contributing factor is just, there are more ads than there used to be. And the ads that are there are probably getting more expensive, although I'm not involved in enough ad words, or I guess just ads accounts to, to know if that's the case,
Greg: but that's the anecdotal evidence for sure that certain things are much more competitive in pay per click.
I mean, we could do that analysis. Earnings, but that's my. Um, but you know what the, the, the, the increased monetization of the serve, the new ad units, you know, the pages that have nothing but ads above the fold on mobile devices. You know, this is a source of enormous criticism now from people in the industry about Google, you know, that they're just putting too much, too many ads, too much ad clutter in the, in the results.
And there's a perception that search results are getting worse, even as the technology and the backend becomes more sophisticated. It's kind of crazy that the market is so demanding that Google has to continue to squeeze more and more revenue out of the syrup when it is so dominant, you know otherwise it just seems, it seems really unfortunate because they're just making money, hand over fist and yet investors want more and more and more growth.
And some of that is propelling, Google to do all this monetization.
Okay. All right. Let's yeah, let's, let's, let's go to to you, David, you also Google announced that they were going to be using various AI tools and methodologies to automatically update local business information on their own. And you know, that got a mixed reaction on Twitter. When I saw that kind of news coming out, what, what is your name?
so tell us what that means.
David: The, okay. So the announcement was not interesting at all to me in and of itself because I think it's, at worst, it's a continuation of kind of what Google has been doing and at best it's a moderate enhancement to what they've been doing. So it didn't strike me as anything revolutionary, but essentially Google is programmatically updating the business hours of businesses that haven't touched their Google business profiles in awhile.
And in some cases, even businesses that have touched their Google business profiles based on data that they're collecting from Android users about being in a particular venue. They also mentioned that they're using Street View technology and reading operating hours signs, which strikes me as pretty ludicrous given the infrequency that the street view cars are driving around. And they also mentioned that they're using Local Guides to, you know that they're soliciting our updates from, from Local Guides. I don't, other than the street view thing, which may be new. I didn't see anything in here that they haven't been doing.
Greg: well over the last several, several years, the Duplex stuff, calling businesses, the Duplex stuff
David: So, so maybe they're also using Google Duplex to, you know, programmatically call business owners and ask, ask them for hours. Um, again, That's something that's been around for the last three or four years. So I didn't think necessarily that it was that this particular announcement was newsworthy, but it is a really important public statement that sort of reinforces the.
A couple of things that Mike has been talking about for years. Number one, that your Google business profile is not your Google business profile. It's Google's and they're going to do whatever the hell they want to do with it. Even if that's updating your hours to something that isn't true
Number two, it speaks to their emphasis on crowdsourcing with the Local Guides program, which has long been a thorn in the sides of Marketers.
And number three, it speaks to their notion, their sort of company ethos of updating things at scale. I wouldn't call this artificial intelligence necessarily. I would call it just the vast amounts of data that they have access to.
And so I think that all three of these things are important trends that are relevant for Google business profiles, not just when it comes to hours, but I feel like there are things… I'm seeing more and more programmatically written descriptions of businesses, whether that's, I shouldn't say programmatic, Google written descriptions of businesses, whether that's coming from AI, whether that's coming from Local Guides, whether that's coming from an outsourced human team,
…but we're starting to see Google, I think ramping up the degree to which it is writing Google business profiles because they've had such a hard time engaging business owners to do this stuff themselves. I think that's the other side of the story is that they've had such poor, they've had such poor engagement with Google what is now the Google business profile manager or whatever it's called that they've had such poor engagement with its predecessors that they're sort of forced into this situation of if they want to present accurate information, they have to find out, to do it through other means.
Mike: So there was a couple other things that I thought struck me about the article. One, was the title self-generating maps or something along those lines, which I thought was very upity
David: a self updating maps, AI rebuild a self updating mess.
Mike: So that was one, two. They said they were going to update over the next six months, 20 million businesses internationally.
This is roughly one fifth or one sixth of the total, we don't know exactly, but in that ballpark, but I thought also it reflected their need to position themselves as an AI first company. Right. I thought from a marketing point of view, it's attempting to say, oh, we're, we're at the leading edge of, of AI even if it's stuff we've been doing for 10 years and we're repackaging it and a little bit new and shiny or boom. We're at the leading edge of AI.
Greg: Well, they've been, they've been saying a version of that now for several years. So we're an AI first company. They went, they went from a mobile first company to an AI first company.
Mike: Correct. And so this is reinforcing that as well. So that's the other big trend, is that desire to be perceived that way.
And, and it's largely how they, I think they do use this a lot, but
David: on the listings that I manage, you know, I've seen this basically on a monthly basis for the last several years that Google is updating hours on probably 15% of the listings that I look at in a given month. So it just strikes me as, Mike is a great point, it's a packaging exercise. There's, there's not much actually new behind the scenes that's happening here
Mike: And yet they don't provide tools in your debt, in your situation, which is a very large number of locations to easily reject those choices or to choose them. Right. So it's like, it also reflects this long term hubris about the value of this data or the unwillingness to deal with the 10% that's wrong. And is it an inevitable outcome of Machine-learning
Greg: okay. So there's more to say there, but let's move on to our third item today. And you might you've written a, a three part article on the ways in which image recognition image understanding is impacting Google's search for.
And the larger role that they're playing in, in sort of local search. And so why don't you talk about that? We've got two of the articles up. The third one is coming either later today or Monday.
Mike: So this took me five months to write. Just I've been thinking about it for a long time, just about the long trends involved in it, as well as the amount of technology that's involved in it.
And this is one area where Google's AI is actually a very innovative and leading technology, but to a large extent, particularly on mobile, primarily on mobile, the SERPs, the search results have been increasingly visual. When I did an actual analysis of the number of pixels allocated in a search result in 2016 to today, it went from 2% where images and visual stuff to almost 40% are images and visual today.
So a dramatic increase in the amount of imagery. Graphics on the search results. Page one assumes that because Google has testing it. And that's what people really want. Also a vast increase in the number of images and space in the local pack in organic, but also in new universal format where they're bringing in panels of images from, and I've written about this house where from their image collection, from.
Sweet from a local shopping, they're bringing in panels of images and providing a more Pinterest like interface to access those. So one big trend is Google just jamming the Serbs full of images. And the other that we're seeing, it's just, they're increasing ability to understand incredible nuance about what's in an image.
The content of the image, the entities in the image, the colors of the image, the style of the clothing people are wearing in the image they claim now to be able to understand and there, and in the article, Either later today, or next week I will be detailing how this is impacting rank and conversions and what a business can do about it, but it is a huge trend in local and all of mobile search results that I think opens up all new areas of technical SEO.
Greg: Well, it definitely makes sense for mobile search results. I mean, even on a, I don't know what the size of the normal iPhone is today, you know, six, six inches verdict. Or diagonally. Um, it's still very difficult to read a lot of texts, right? So a text-based kind of surf in a mobile on a mobile device is less user-friendly than one that's broken up by by images.
And so I think that makes sense from a usability standard. Um, th th I don't know if you mentioned, you may have mentioned a product inventory images, you wrote a piece about, about that. That's also kind of two, two trends coming together, right? Google's emphasis on local product inventory and product images in, in the serpent, in the local pack as well.
So that's another, another thing that's really interesting. Um, and you've got, you've got a number of pieces. I don't know what the total number is. It's like five or six that, that touch on images in our analysis section on. which I would encourage people to check out, but this is a major trend, as you say, and it's something that people should not be ignoring
Mike: and their AI is incredibly sophisticated.
Although at the same time, it's like all AI has a certain blind spot. You can see that in. The second article where I compare two pictures of a dentist, one Google clearly identifies as a dentist with her and the other, it thinks it's about medical gloves or something. It's just fascinating, the small differences in image that so dramatically changed.
Google's understanding of what's in an image and it really, it, it will be interesting to see how that, that, that. The reality of SEO and photography work to sort of change. What, in images going forward, will images become less emotive or less substantial or less creative? Because they're being asked to conform with Google's understanding of what an image.
Greg: Well, I mean, it almost certainly once those once people sort of reverse engineer what's what works about an image and what boosts rank, you're going to see a lot of people conform to that aesthetic or that standard, whatever that is. I mean, I'm sure. Can you give, can you give people a little bit of a sense of, of some of your top level recommendations that's in the third.
It's yet to publish, but sure.
Mike: I mean, firstly, that every image that you're using should be run through their image, AI API, either at scale or dropped in on the website to see how Google understands it, to make sure that that's the case in terms of local, you want to be sure that you have images that reflect the categories of products that you're selling and not just one, but you need multiple images in a category because then Google will.
Well then segment your local profile by that, by that and your
Greg: example, it was jewelry, right? So different categories of jewelry, right?
Mike: Rings and earrings and necklaces and engagement rings, et cetera. Also, you need to, every image needs to be a good one because the query is going to dictate what image shows.
So you never know which image you're going to show you a very little control over that. Also, you need to be sure that the profile image is a profile image that I didn't talk about this in the article, but it's just a case similar to what Google is showing for that category. For example, if they're primarily showing.
Front frontage footage of your business. And you want a great picture of that. If you try to use a logo for that, Google may then take user generated content, which is much lower quality. And then finally you need to be watching. If you're in a business that's generating user generated content, then you need to be careful.
Um, and there's some interesting stuff that I didn't talk about my article, and I know Joy's going to talk, be talking about it the next, hopefully you, but also how images. Influencing are now being included with reviews and are influencing the persistence and visibility of reviews
Greg: and updates from users is
Mike: one of the places there's.
One is updates and users, and the other is in reviews, both places now except images. And it seems to influence their visibility. So images are going to play a big role or are.
Greg: One thing that struck me, I think about the second, the second piece you wrote, and this is kind of consistent with what David was talking about with Google.
doing this sort of a involuntary data replacement is Google is going to determine, I mean, if you've got, if you, you, you talk about the user generated images and the challenge for business owners, where people are uploading photographs. What speak to that for a second. Like, so
Mike: Google at every opportunity is.
Local Guides to upload images. If you go, if you're in a physical area and you walk by business, you may get a ping from Google, obviously. And other businesses. It's common that people upload images, I've uploaded images just to get points. As I pointed out in another article in their local guide program, and those images were not.
Not the ones I am necessarily most proud of. And they may be seen three or four or 5 million times by Google scouting, which even if it's off by an order of magnitude, it's still a lot. These images get seen it seeing a lot. Now it means that you need to control your images by keeping them from.
Continuing to upload new images so that new image, there's a strong predilection in Google's image, ranking algorithm for freshness. So if you're in a business where there's a lot of user generated images, either you have to work with users that there help them get better images, or you need to upload your images more frequently to keep the.
Down in the vet view because they may or may not reflect your brand image and then also report them if they're terrible, your dog.
Greg: Yes, she can. She, she, she, she agrees with the strategy of, of, of periodically updating images. But I mean, that's a big, that's a big problem. If you're a business owner and you spend a bunch of money creating some, some nice images from, with a professional photographer and you upload them and they get overwritten or superseded, not overwritten, but superseded by a bunch of user-generated mediocre images, that's gonna be.
Mike: And we're seeing some indication. I have some early research from patient pop that indicates it dripping images in provides more uplift to visibility than just loading while up and forgetting about them. So it appears to be that Google is at some level rewarding, persistent image uploading, which hasn't been.
It hasn't been tested that broadly in the past, although it's consistent with other signals they're sending. They did also, as we noted, I think last week update their how to rank in Google by adding photos and product inventories are both in there now. So there they're clearly sending signals that they want images.
They're cooling, showing images a lot more. And I think it, in some ways it's sort of, it's the subtle signal that people convert I think have great images. I'm trying to find the research on this, but I am convinced that having a great image. That that communicates to people is really one of the best ways to get conversions.
So I'm bullish that.
David: Yeah. I remember a being survey from, gosh, probably 10 years ago now, maybe even longer saying that at least in the restaurant vertical, it was like the number two thing that people wanted from a restaurant profile. Hours of operation. Number two is images
Mike: and Google research, which wasn't published.
It's a similar research that was done in 2014, but it was redone in 2017 and they gave me access to the data. Um, and it was a big number. I don't know, 90% of people trusted businesses, mold. With images and 74% used it as a primary buying into consideration. Yeah. And
Greg: Yelp and Yelp has research that validates that two profiles with more images and better images do better than profiles with fewer images or less, you know, a lower quality, lower quality images.
I mean, it makes, it makes intuitive sense because what we're trying to do as consumers, when we go, when we search online is we're trying to get as much of a sense of the business, you know? Pieces of information, reviews, descriptions, and images as we can. And so more images give you a sense of what the reality of the offline experiences.
I mean, how many times have you looked at the the, the vibe area in, in, you know, on the Google profile to see what the interior of a restaurant is like? Right. I mean, the food is great, but is that a place that you want to hang out or is it like a, you know, grab and go or something? So it just, it just makes complete sense.
I mean, in this case, Google is really, you know, generally speaking, Google tries to do what the users, what they think users want and what their research reflects users want. And, and sometimes that works for the business and a lot of the time it doesn't so. All right. Well, we've come to the end of another interesting discussion, hopefully hopefully interesting for you, certainly interesting for us.
And we'll be back next week with episode 61. Um, we also want to say that we are working on. A monetization plan and we will be communicating about that pretty soon. We've got some exciting new things coming in the beginning of may and we'll be letting you know what those are and your opportunity to pay for them because we have to support this somehow.
Um, and so we're, we're, we're starting to starting to do that. So hopefully you'll be into. Um, any last statements from you guys, I am,
Mike: would put out a request. Anybody listening that if you know, a photographer who is very thoughtful about commercial photography, who has explored the issues involved with presenting images at very low pixel resolutions, which is what local is.
150 by 200 pixels. I would love to chat with them. So send them my way and Blumenthal and your media data seal. I'd love to chat with them. So I'm just looking to increase my understanding of commercial of the constraints around photography in a commercial environment for digital.
Greg: Okay. All right. Great.
So thanks very much for listening as always have a great holiday, whatever a tradition you celebrate or have fun if you don't celebrate any of them and we'll be back.