Part 2 Video start 8:50 - Google’s recent Android Privacy Sandbox announcement: Something real or is it just more Google PR?
- Google can now summarize your writing using AI.
- How to Beat Google Title ReWrites
- Introducing the Privacy Sandbox on Android
- Ron Amadeo: ‘Android’s Toothless “Privacy Sandbox” Fails to Answer iOS Tracking Limits’
- Wix Earning Report 2021
Transcript Ep 52:
Greg: Welcome to the Near Memo, a weekly conversation about search, social and commerce; what happened, why it matters and the implications for local. Hello, again, we're back for Episode 52 of the Near Memo. It's David, Mike, and me, Greg. And as always, we're talking about the week's developments in search, social and commerce.
We are often challenged to figure out what we're going to focus on because there's a lot of interesting things going on, but we've decided here and David is going to start us off with some interesting things that Google is doing with its AI, Google docs, summaries, and the implications for search with that.
David: Yeah. I, on that note of too many things to talk about, I'm going to cop out and combine two of them. And then I found that Marie’s saying this it's
Greg: It's not a cop-out, that's not a cop-out as long as they're related.
David: there, uh, in my mind, uh, integrally related to one another. So the first was a link that I first saw, um, from Marie Haynes who noticed a product release in the, I think it was on the Google cloud blog, which I guess is the umbrella organization for Google drive and Google docs.
And so that the new feature that just came out is an automated summary that will now show up, uh, sort of at the upper left-hand side of a, of a Google doc. We'll attempt to condense a multi-page doc into, you know, a couple of sentences and it's, it is editable. So if you're using this as a productivity tool, you know, you can craft your own summary that you want people to read as an abstract or what have you.
But, uh, Marie, I think correctly pointed out just how, um, I think how, how far along Google is with its AI in terms of understanding and as a long piece of content as a human being. I understand it. And I think that the, the depth of their, um, the depth of their ability to do that has probably been underestimated by many of us, myself included.
Um, now I should say I have, I tried to demo this feature in, in a handful of very long documents that I have, uh, in Google drive this week. I, it has, apparently hasn't been enabled for my account yet, or, or whatever, but, so I can't speak to how good the summaries are, but I. Google is probably not going to release something.
Um, the substantial, um, even in beta without a significant degree of confidence. So I think that that, that was an interesting product release, uh, from Google. And then sort of combining that with a really insightful post from Cyrus Shepard, uh, formerly of Moz now of Zyppy, Z Y P P Y. Um, and Cyrus did a very detailed analysis of which title tags, uh, Google was rewriting.
Uh, which this has been a thing now in the search results for, I don't know what nine months or a year SEOs have lost their, you know what that, oh, we put all this really hard work into, you know, targeting keywords on a certain page and crafting this really beautiful title. And then when the page actually ranks, you know, something like 60% of the time Google is, is rewriting the title.
Um, and so this has been a. You know, sort of a bone that SEOs have been picking with Google, you know, ever since this, this came out. And so Cyrus wrote this post about sort of how to combat their, um, their title, rewriting a couple of high-level takeaways, keep your titles short, make sure that they match the H one use dashes instead of pipes, which is, uh, I think at least in my old, old guy, SEO mode pipes have been sort of the preferred thing for a lot of.
As CEO's who have been in the game for awhile and, and Cyrus found actually dashes are less likely to get rewritten than pipes. Um, he also found writing things as a more, more or less as a sentence, as opposed to chunking things that, you know, Google can easily crop or, or rewrite. Um, anyway, it's a terrific posts.
And the takeaway for me, of both of these things combined is just how good Google is getting. Again, understanding content, digesting content, and then as it relates to search presenting what they think users would want to see, um, about a given piece of content that's extracted, uh, and abstracted on the surface itself.
So I would, I am really, really eager to try this feature in Google docs. I think it could actually potentially be a useful SEO tool over time. If you want to see how Google is understanding the content of a given web. Um, Marie hypothesized that the Marie Haines hypothesized that probably the same AI that's driving, these summaries is being used in search.
And so, um, I'm really interested to try this new feature out. And if those of you, if there are listeners or viewers out there who have had this feature enabled in Google drive Google docs, I'd love to start a conversation on Twitter about how it's going and what you think of it. So that's, that's my item for the.
Mike: I had two, two thoughts. One was Darren Shaw's response to, Cyrus's article was he liked, and this is something that I've done it a local for many years, which has make very long title tags and let Google pick and choose based on the intent of the searcher, like title tags that reflect broadly both the intent of the site and the H one and sub heads of the site. And that has been a tactic in local and Joel who has tested this at scale, Joel Headley from Patient Pop determined that he got 15% more, uh, visibility from that. So there's two ways to approach this run is to sort of worry as SEOs do about this specific 64 characters that they want to show the other is be a little more expressive and let Google do its thing.
The second thing, and this is something I've been looking at for the last year, year and a half, two years and in great detail is Google's understanding of images. It rapidly going up concurrently with their understanding of text and they are extracting a great deal of information from images sometimes in a sophisticated way, sometimes like a third grader, but they're doing it and using it and it's going to impact results.
And I think that. The idea that images as a SEO technical SEO tactic is on there, not just on the horizon, it's with us. Now. It needs to be looked at in the same context as you're looking at content summaries.
Greg: Well, I mean, I, I I'm led to think about kind of AI as a content generation tool, which is happening, you know, increasingly and how, how far down that path are we going to go?
I mean, people have talked about it for new. And for web copy and for ad copy. And it seems like that is an inevitable by-product of this kind of advance and we'll see more and more and more non-human generated content and copy across the web, which may ultimately not be a big deal. Um, you know, for certain kinds of mundane uses, but, uh, could be good.
Just add more crap and spam. Depending on how it plays out. That's why
David: it's so important, Craig,
Greg: and yes. Right.
David: All of these, uh, Burts and all these other acronyms that Google is coming out with. Yeah,
Mike: And from a social point of view, it makes more sense to me to use this as to summarize really human content as David described, or to understand the content of images I was talking about, then it does to generate more bullshit.
We need more bullshit. Like we need a hole in that, but I
Greg: mean, this goes to a larger point about search in general, right? I mean, it's, it's like there's less and less trustworthy, really valuable content, even though there's more content overall, there's less that you want to pay attention to it. You want to just find the, really the, the most.
Thoughtful and useful bits and, and the rest is just kind of a version of spam. And, and, you know, I mean, I think this is the challenge that Google faces is really. Um, and this has been a challenge for many years is how do they get to the stuff that is truly useful and, and not replicate the, the spam? I mean, I was, um, Uh, there was an article this morning that I was trying to, uh, that I, I just copied the title.
It was in an amp page. And so I copied the title and it was trying to get to the original article and there must have been, you know, 10 links that had, had, had the same title and had some excerpts from the article. And that page has just completely worthless. It's like, I don't need to see anything other than the original source, unless it's a gated piece of content, which case I want to see some, some alternative version of it, but there's just so much of that that seems to show up in, in the Serb, you know?
So we'll see. So, um, so Mike you're, um, you're, you're on deck now here or your up. So,
Mike: yes, I use our Fridays to help me learn more. So your guys are gonna have to give me a little bit of schooling about Google's privacy sandbox on Android. I read this whole announcement and it was an announcement about an announcement, about a plan to maybe create a plan, to have some privacy on Android in the future.
I read it and I did not see anything real in it. And so I was just curious. And this opinion am I with, this is something group Gruber during fireball said, basically it's a plan to create a plan to hear sir. Now. So there was that, that is this just performance theater or is there something real? Here was my first question, is this, I mean, I went and looked and Google has with the exact title.
Privacy sandbox on Android has 25,000 articles in search already in 5,000 highly trusted news sources writing about that. So they got great coverage. Of something that didn't appear to be that appears to me to be vaporware. And then, so there's two questions here. One, what is, what was their intent in this?
And two, is there more than vaporware here? And three will Google ever really. Respect our privacy. And I turn to you guys and your dog to see if there's any answers.
David: So it struck me Mike, as a running the exact same playbook that they've run with these, uh, you know, farcical summaries of how well they're combating spam and in Google maps, right?
It is, they put out a, uh, a shiny object that the tech press covered. Without any critical thought whatsoever. And all of a sudden they're planting the seed in consumer's minds, everywhere that oh, Google actually does care about privacy. Right. And, and even if there isn't anything really behind it. So I, I truly think this is a, uh, this is a core corporate PR strategy at the highest levels of Google that this is how, this is how we get our message out.
Okay. That? No, there's just not going to be enough critical analysis by any major news outlet. They're all they're going to sort of spread, help spread Google's propaganda, um, without, without any, without ever really understanding if there's anything behind it. So,
Greg: well, let, let me, let me say something slightly less cynical, the dog, the dog agrees with you.
Um, but, but I'll say I'll take a different position. Which is, you know, the way that I see this. So the coverage was largely about apple, ironically, oh, Google's going to do this thing. That's like what apple is on. And look at Facebook, which lost $10 billion or said it was going to lose 10 billion, $10 billion as a result and look at their market cap.
So, so Google is, is, you know, that that's sort of how the coverage went. Uh, I think that, you know, Google has the privacy sandbox initiative for the desktop and that there's some real stuff there, topics, you know, coming out of there. Uh, targeting they've changed the topics, which is, uh, which we talked about previously, which is a version of contextual targeting with a behavioral targeting element to it.
I think that's what we're going to see for Android. I think it's going to be largely a mirror of that. So I do think that there is something here that they're, they're pouring over the desktop approach into, into Android. And they're going to get rid of the Android ID in the same way that apple has essentially deprecated idea.
So I think there's, there's some, there's some substance here. I think the timing is, is the, is theater, right? I think, you know, they, they, they have not implemented anything from what I can tell. And so I think their, their announcement is really designed to garner the maximum positive PR coverage that they can.
Uh, and, and, you know, now, uh, Is is also playing, not just to the public, but to the regulators. I think that's, that's a lot of it. Like, Hey, we really do care about these consumer privacy issues because they're getting dinged along with Facebook, all over the world for violating this or that sort of privacy rule and privacy regulation.
Um, so I'm not quite as cynical that it's saying there's nothing at all here, but I do think it's kind of, uh, uh, uh, you know, it's a PR move on their part to say two years from now. We will have some tangible. Thing to, to, to, to implement that in two
David: years when it's, when it's clear that that plan is going to be unworkable, they will come up with another plan that will take two years to develop and then implement.
I mean, we've already seen this, we've seen the same story before with the death of the
Greg: cookie. Right, right. So that, so, well, I think that that reflects the very real problem of what do you do? What do you actually do? That's going to turn. And, and, uh, represent a compromise. We compromise
David: as in compromising people's privacy is fundamental to a huge amount of Google's revenue in a way that it's just not for apple.
Greg: So, right. So, so that's the interesting thing to me is comparing the two hypothetical's Google's hypothetical solution with Appleseed. So I, I, I don't believe that Google can do anything which represents an opt in. I don't think that they will do anything that is an opt in. It will be some version of opt-out, which will be confusing.
And most consumers won't pay attention because all of those opt-out solutions are too burdensome for the consumer. What makes the apple solution so effective is that it's so simple and it's an opt-in. Do you want them to track, you know, no. And Google knows that if they allow an opt in that people will say no to that.
And, um, as a consequence, we're not going to see an opt in from Google, which, which I think renders this, the privacy side of it probably ineffective or, or, and
Mike: oblique criticism of apple in their piece was interesting too, because apple gave Facebook all people a fair bit of time to come up with alternate.
And attribution models. And you can see what Facebook did. All they do is complain about it. They didn't come up with, in it. Apple gave them almost two years from announcement to actual implementation. So a similar timeframe, uh, with betas along the way. And Facebook didn't do much of anything
Greg: about it. So they said now they said recently that they're getting better at attribution.
I think they've cut. Cut down on the, the under counting of Iowa. And installs or whatever the, you know, whatever the conversion versions is. So they're getting, they're getting somewhat better. I mean, I think this, this remains, this is sort of the Internet's major problem is how do you respond to the increasing demand for privacy after all the abuses and still give marketers targeting and attribution tools that are not going to be, uh, An exploitation of, of user data or deceptive in some way.
I think it's very challenging. And I mean, my personal preference is make everything opt in, but when you do that, then you get, you know, 14% or whatever the number is of people opting in. So any more on that exciting topic. All right. So then, then the, the last item of, uh, today is, um, you know, there are many things that we talked about, uh, discussing.
I mean, there are some interesting things that happened with next door this week and voice search and other. But, um, earlier this week, a couple of days ago, I think Wix reported it's fourth quarter and full year results. Uh, and their, their 2021 revenue was something like one point just under 1.3 billion, 1.27 billion, which was up about 30%.
Company's not profitable. Um, they reported, um, they reported 220 million accounts globally. Uh, with 56% of those, I believe in north America, which is mostly the us and they had 6 million premium accounts, which interestingly was less than the verbal number I got from Wix some months ago. But I think this is, this is the accurate number that may have been sort of a mistake.
So 6 million premium accounts, 200 million total accounts, most of which are free. Only 2.7% are paid account. Um, Wix is the largest of all the website builders out there. And, um, there are, who comes out at about $200 per year. And it's a, it's a really interesting story. Once again, we've talked about this before and written about it several times because by my calculations, they're about 123 million, uh, Wix accounts paid and, and free in north America.
But. 31 point something million small businesses and surveys report that between 30 and 50% of small businesses don't have websites. So this, you know, these are recent surveys saying, you know, some substantial percentage of small businesses don't have websites. So this is a really interesting like discrepancy between the small businesses that have websites, all the wakes accounts and who else is out there creating these accounts, obviously.
David: I mean, do you think there could be like, you know, pur essentially people using Wix as a personal page for like a wedding planning or whatever, you know, those, those kinds of things. I mean, you have to at a hundred million, there have to be what we would consider consumers as part of.
Greg: Well, but I mean that, I mean, why are these consumers using these websites?
Because they're either trying to market themselves in some fashion or they have side businesses or, you know, whatever they're testing the waters for something. So it's, it's, it's, it's very interesting because the w what Wix projected and it's it's, uh, it's it's report was that they were going to earn $16 billion over the next decade from their existing free accounts, basically.
So, you know, they have this enormous built-in audience and just simply by bringing some of those people over, they can generate a huge amount of revenue. And in my mind at once again, raises this question of what is the addressable audience for these software vendors, SAS companies. And it's going to depend on what sector you're in of course, and what you're serving.
But, you know, I used to have these discussions in the old days with different people about what is the addressable market for small businesses. Certainly it's not the 30 million. It's gotta be some subset, 10 billion, 15 million, you know, those numbers seem reasonable, but Wix, as data suggests, we're dealing with a very different, you know, it's a very different scenario here in the wake of COVID and, and e-commerce, and.
Everybody moving online and individual creators. It's, we're, we're in a really different sort of interesting new environment. I think
David: I would agree with that. I mean, I think, yeah, the, the influencer market, which didn't exist, you know, five years ago, seven years ago, whatever, um, that, you know, that those theoretically are perfect Wix customers as well.
Um, they just need some, they need a domain name that, uh, embed of their Instagram feed. Right. And a contact button where people can sign up to pay them to promote a product or whatever. So I think that the, the definition of a business is probably has just changed radically since the Kelsey days of the late aughts or whatever.
When so many of these discussions were, were happening, not that they've ever stopped, but,
Greg: well, I D I mean, interestingly, um, I had a conversation with Perry Evans yesterday and, um, you know, he's had ever current. And we were talking about the challenges of the small business market and in many ways, as radically as things have changed with the emphasis on e-commerce and mobile devices and all these different things that have happened over the ensuing period, um, the, the, the, the, the challenge of getting small businesses to pay attention and pay for digital marketing is still pretty difficult.
You know? I mean, you, you you've just been in a, in a version. The world, David and it's it's, it's, it's it's in some ways, nothing has changed since the year 2000. Uh, even though, even though certain other things are radically different. So it's, it's really just a very complicated, interesting market. I just want to throw in.
And then my. You can feel free to say whatever you'd like. Um, we w we're not, you know, Wix is not the only website host, right? There's Squarespace and WordPress and GoDaddy, and, um,
Mike: you know, Google my business websites and being just announced a website,
Greg: bright, smart there. Right. And Google. Right? So there's, there's millions and millions of more websites out there.
And most people probably don't have three websites on three different platforms. So it's a, it's, there's a lot of non duplication out there, and it's a much, much, much bigger world of websites than what the us government says in terms of small businesses. And, you know, especially when you consider that, uh, so many of them don't have websites.
Mike: It would be interesting to look at random at a hundred thousand sites and see what they really are. Are they wedding sites like David suggested,
David: I'm just trying to try to put some kind of hypothesis out there, what they could be.
Mike: So, but that's, that's the question? What are they? And it'd be interesting to look at a sample to determine how many are actual sort of side hustles or actually business focused versus how many are just truly personal endeavors, maybe a CV or some other use that goes, that isn't really specifically about making money more about promoting, you know, or personal capital or whatever your human personal brand, your personal brand and equally offensive term thing.
But, um, yes. I just curious, it'd be interesting to look at a sample and see.
Greg: Well, we can, we can try and get that information from them directly. We know a lot of people who are at work, so I don't know how much they would divulge that, but it's, it's just a fascinating, it's a fascinating thing because it just really reframes the whole discussion about who is.
Your target audience,
Mike: but it goes to what we were saying before. What is a local business? Right? I mean, there's the info USA definition, which is similar to Google's definition, which is a location or business with a phone number that's outperforming selling goods or services in the local market. That's one and there's rough peers to be somewhere in the order of 17, nine of those there's the census bureau definition, which is 39, but includes a lot of.
Corporations or, um, like personal corporations to manage income as part of other projects. Right. I do three jobs and I'm sure David does several. Right. And so how many businesses does that involve? Is it more than one or one? So it's a very loose and amorphous. Answer to, and a difficult question to get good answers to without really clear definitions of what you're talking about.
This is your next job, Greg, you have to do it for near media. I'm looking forward to the
Greg: yeah, well, I will totally single-handedly segment the small business market or the addressable market for all these SAS S and B SAS companies. And, um, and we will sell that for. $100 million. Um, yeah, it's just, it's just very interesting because it explodes these sort of categories and ideas that, that have, have been pretty, uh, pretty much gospel and unquestioned in the, in the world of, uh, the world of small businesses.
So, all right, we are out of time. Uh, it's gone quickly as it always does. Any last thoughts or words before we adjourn for this week? Just
Mike: stay safe out there.
Greg: Alright. And with that, have a good weekend week, evening, morning. Thanks for joining David, Mike and Greg to stay on top of the latest developments in local.
Subscribe to our firstname.lastname@example.org. Let's see you next week.