Part 1 Video start 0:13 - Ratings vs Review Text, Can G Ad automation replace agencies?, Will Musk torpedo Twitter ads?
Part 2 Video start 10:53 - Performance Max - Is Google’s newly updated automated ad product good enough to replace agencies?
Part 3 Video start 17:28 - What is the future of advertising on Twitter under a Musk regime?
- Swayed by the Reviews: Disentangling the Effects of Average Ratings and Individual Reviews in Online Word-of-Mouth
- Google pushing goal-based Performance Max campaigns
- Twitter: The 'Cesspool' Cometh?
Transcript Ep 62:
Greg: Here we are again for Episode 62 with Greg, Mike and David on the Near Memo, where we talk about the intersection of social, commerce, search and other things, the list grows longer every day.
David: But yeah, so you guys still won't let me get in the NBA playoff
Greg: All right. Um, you know, once again, there's kind of too much news to cover, but we pick three items and, or three and a half items as the case may be and kind of unpack those for you in about 20 or 25 minutes. And today we're going to lead off with Mike. Who's got who's dug into some interesting consumer review research. And it's implications for marketers.
Mike: So I found this through Jason Brown, who tweeted a summary article from some science magazine, but it's research from the university of Massachusetts Isenberg school of management by Zen Fe Lei. It's titled Swayed by reviews, disentangling, the effects of average ratings and individual reviews.
With on-line word of mouth, historically research has leaned on ratings as the primary determinant of driving purchase intent. Although recent research seems to have indicated that somewhere in the order of 55% of all consumers actually read reviews or the summaries of reviews, and it's usually in more important categories.
Uh, and when there's high intent and when there's bigger, risks are less known about the brand and he was, he explored with his re co-researchers, whether the top three reviews. Had as much or more influenced than rating. So we had two hypotheses and he developed a three experiment design to separate out this.
And they did it with what's called a trade-off design, where they would interview you present this information. They had manipulated the content of their views, not the length and the rating. So in one situation, they'd have. 4.2 ratings and positive to positive reviews and one negative. And the other they'd have higher ratings, but with negative content and they determined that the, the, the three, most, the three visible reviews like Google shows are more, once a threshold has been passed of whatever the rating is that people need to around 4.2, once that threshold has been fast, then the content of these reviews.
Becomes the primary conversion factor on people that actually look at them and the way Google presents them, it's pretty easy to look at them. Uh, so I think it may have more influence than, than not. So it's just,
David: you're referring to Google's presentation and reviews in the business profile panel,
Mike: Yes. Correct. Well, right. That's right. So they do show, just review ratings upfront, and then you have to go in to see the presentation. So it was interesting to me just because there's so much focus on ratings and much less focused on content. Uh, also because, you know, obviously there's some questions in product review site and on product pages.
There's some question about what's honest and dishonest. And if you show an accurate rating, let's say it's a lower rating, but you highlight relevant reviews that tend to be positive. And then give people click to see the rest of the reviews. Is that. How does that work in context of the FTC? And I would contend that's probably okay.
As long as it's easy to get to all the reviews, easy to read him, as long as you're transparent about the score. Uh, if the reviews you highlight are relevant and they point out that it's equal to a 2% price reduction in terms of the increased. Conversions you get from, from positioning them that way.
Greg: So, Mike, just to clarify one thing, cause I didn't read the study the top three reviews, whether positive or negative we're, we're going to equally influence people. So in other words, if, if they see positive reviews, they're more likely to buy. If they say negative reviews, they're less likely to buy it.
David: Wasn't a star rating. Right.
Greg: So and then how was a negative review concocted? I mean, you can have. A lot of reviews are mixed, right? Yeah. They took
Mike: basically the same language and they just had a negative to the exact same reviews. Just converted the language.
Greg: They supplemented a neutral description with a, with positive and the positive and negative,
Mike: negative, negative, and tested it.
Greg: Well, I mean, just anecdotally, that makes a lot of sense because I mean, people, you know, we've got like this migration tool. For plus review, you know, anybody who's kind of going to be in the top ranking group is going to have in most cases better than a four star rating. And so you'd have to look at the reviews which Google doesn't make an entirely easy to do.
They, they show you the review distribution,
Mike: um, and they make it easy, relatively easy to see the three. But if you want to see them all multiple clicks.
David: It depends on the surf. Depends on the surface, Greg. I think that the maps app, for example, it's a lot easier than the desktop business panel to click around and see more reviews.
Greg: So, which is, which is amazing since I never used maps for reviews at this point. So, I mean, I kind of, I kind of do I suppose, but not really for
David: you. Don't use maps as a search. Uh, discovery discovery, search interface.
Greg: Not really. I mean, I use it in limited contexts or I use it for navigation, but, um I'm using apple maps more these days, but that's a different conversation.
Um, so what, so what's the practical takeaway Mike
Mike: for? And the practical takeaway is that once you've hit whatever the threshold is 4.2, then I think the work becomes making sure you've. Enough content in your reviews, better, more informative content in your reviews and more positive content into reviews.
So that one at Google, those are more likely to show. And in your own site, I think the practical implication is if you're going to show a summary, you can design it to highlight the more positive aspects. As appropriate while still showing your actual review rating. So it would, it would mean there's no need to hide your extra review rating the way fashion Nova did that.
Um, as well. So I think it depends on the context, but I think it means businesses particularly with more expensive products and. Purchases need to focus as much on content as they do on readings
Greg: or maybe more. Now it it's, it's, it's interesting not to sort of open the door to another discussion, but it's relevant in the sense that, um the role of negative reviews in establishing the credibility of positive reviews is something that there was a sponsored piece on search engine journal.
I think that was one of the review monitor, you know, reputation companies and they
Mike: power somebody.
Greg: No, it wasn't power views of somebody else, but, but basically the, the larger point was that you, you shouldn't be afraid of negative reviews. They help establish the credibility of positive reviews, et cetera, et cetera.
And I think there's truth to that. So I don't know. How that kind of element plays into this or not, you know, if you're sort of shaping your most prominent reviews to look positive and you've got a great star rating, you know I mean, is there a, is there a risk there that is going to see that as fake or, or manipulated in some way?
I don't know. I think as
Mike: long as you make it relatively easy to access all the reviews and order them in a way that the consumer wants, I think that's really what the obligation
Greg: in. Yeah, sorry, many, many e-commerce sites will give you the ability to sort and filter reviews by most recent, highest, lowest, and so on.
Mike: And I think behavior at that point, once we start to reading reviews probably differs a great deal. I think there's a lot of unique individual individualization that goes on in terms of how you relate to the content. Their theory though, was in the top three that the language is simplified. Context of only three views was easier for people to remember.
And thus they latched on to that rather than the distinction between a four, three or 4, 4, 4, 5, they felt it was easier for people to remember the positives from the simple view of only three reviews.
Greg: There are a lot of fascinating.
David: I was going to say it would be interesting. Um, not, it's not really a follow-up, it's more of a different survey, but you know, we, I think it's pretty widely accepted at this point that the threshold is somewhere in the 4.2 range for, you know, I won't consider a business you have, or a product that has under that rating.
Right. Locals found it, these guys found it. Um, but I was wondering
Mike: research that said four seven was the upper threshold where trust started declining because they thought they were likely to be.
David: What is the optimum number, you know, and I bet you that that varies based on the industry and the, the product or service under consideration.
Like what actually is the optimum number where. Maybe people don't need to actually read the reviews as much because that, that sort of lodges as like, okay, this is a good, you know, this is, this is a good enough heuristic for this being able legitimate product that has a couple of problems, but it's overall worth buying.
Greg: That's a piece of research for us to do. I think it's four, four to 46 or 47. Uh,
David: so this is why I think it's interesting because for me, it's probably four eight just personally,
Greg: that's the high. No, that's the
David: sweet spot. That's where I'm like, yup. This seems like the right thing to
Mike: Although if it's something that you're going to depend on.
You're going to read the reviews and if it's something that's a throwaway item, it's like, it doesn't matter. I'm a BA right. You're not, you may not. I think it depends on how you relate to reviews, how much money you're spending, how critical the purchases. Well, if there's a threshold beyond which you're going to start reading reviews, I think
David: that's right.
So I think that virtual for me is probably, it sounds like it's higher than Greg is. I would say four at 44. I'm going to actually want to know like, okay, what are the downsides to this.
Greg: Well, this is, this is, this is a to be continued conversation. Cause there are a lot of interesting aspects here and certainly reviews will come up again and again and again for us.
And maybe we should consider doing some, some research on this, but I want to move on since we've, um talked at length about that too. Google's announcement about performance max, which is it's sort of automated ad program replacing shopping campaigns and local campaigns. And the significance of that for small businesses agencies and enterprise brands.
So David wanted to speak to that. Sure.
David: So, yes, I, you, I believe linked to this Greg and one of the sort of like, well, there's still an opportunity for that coming up. But so, so I discovered it through our own newsletter. I hadn't seen it on Twitter or anything like that. Uh, the, it seems to me to be more of a rebranding of smart campaigns than a brand new product there are certainly, you know, more bells and whistles, but I took a pretty deep look at smart campaigns.
Uh, probably just about a year ago. And actually kind of went through the process as if I were a local business. I think a local sort of like it company you know, trying to get more leads and that sort of thing. Um, and this, the screenshots from the performance max. Announcement that Google came out with this week look very, very similar to that same smart campaigns interface.
So I do think it's sort of an iteration of smart campaigns rather than a brand new brand new product. Um, the, the couple of big takeaways for me from, from this announcement there's clearly a real emphasis on Google shopping as. As a sort of channel that this thing is intended to help. Um, they highlight in the documentation around performance max.
They specifically highlight, Hey, if you're on one of these e-commerce CMS systems you know, there's this automated integration. Um, and so I think it's it's, to me it's most interesting from that standpoint, I think that the, the. Shopping campaigns are sort of the e-commerce equivalent of LSAs in a traditional local search where LSA, you don't really worry about anything, right?
You just say, this is, this is my business. This is the type of business I am. These are the services I offer. There's no bid management. There's no ad copy. There's no creative, et cetera, et cetera, et cetera. These products sort of smart campaign or performance max campaigns now for, for products are largely similar.
Right? You've got the image of the product, the description of the product. The price and then Google just sort of runs with it. And so it strikes me that this is their sort of continued investment in this kind of thing really puts at risk. Some, some of the strategic value, if not all of the strategic value that paid agencies for e-commerce companies can offer.
Um, I think it's a great thing for small companies. Wouldn't be large enough to afford a decent agency to help them with paid advertising. But I bet you, you're going to, we're going to see even larger e-commerce retailers who are currently using an agency to help them with. Google ad spend the smart ones.
Anyway, I think we'll probably AB test these campaigns versus the ones the agencies are working on. And while I would expect that the agency wants still perform a little bit better. By the time you add in the cost of that, the maintenance cost, the management fee that the agency is charging. I'm going to guess that these smart campaigns or the performance Mexican pains are going to perform just about as well, if not better.
Um, and so I think it's it, it does reflect a. Continuing threat that I think that the automation of Google ads pose when you have a, a product or a service that can be sort of described as a commodity in the way that LSA is doing local and the way that the shopping ads do and in e-commerce search. So,
Mike: and increasingly shopping ads doing local.
David: That's right. That's right. Uh, we're now seeing many more shopping ads and local. And of course there's the integration of the Google business profiles in the Google merchant center now. So this could sort of bleed into a local agency, local oriented agencies that work with retailers that sell products.
Well, the, the
Greg: pushback to that, to that point, though, I would say is the, the, the customer service dimension that agencies provide to, to both. Clients and to local businesses because Google, isn't going to give you any human interaction. They're just, it's just going to be a dashboard. And so there's a, there's an interpretive layer and a support layer that agencies can provide that I think has value.
And also if you have an agency that's running multiple multi-channel campaigns, right. They're managing multiple social search SEO and so on. And so. Then it's less likely that the 50 enterprise is going to cut, cut them off. Now maybe they would reduce the scope of the engagement or something like that.
But I don't know how many standalone search paid search agencies exist these days.
David: That would, oh, I don't think that they're standalone paid search as agencies, but I bet you, they have clients for whom paid search is the lion's share of the retainer.
Greg: That that may be true. That may be true. I, you know, having been in a.
Large-ish company recently. Um, I can tell you that people don't have time. A lot of people don't have time to do that kind of due diligence. You know, it takes a lot of discipline, I think, to do what you're suggesting. I think it's smart, but I think there are people there's inertia, you know, just around writing the check or whatever.
So I think that w well, some people will do that. A lot of people won't because it just takes time and energy and people are. Tapped out in terms of their capacities.
David: Yep. And I know there's a lot of skepticism and I think justifiably, so that among agencies, that smart campaigns don't perform or performance max campaigns, don't actually perform at the max relative to what an agency can provide.
Um, but again, I think that I, I think that's much more true when you're talking about traditional words, you know, with headlines, ad copy, landing pages, that sort of thing. Um, I think that in shopping and LSAs, it's a totally different ball game.
Greg: Yeah, I, I we were talking before the call started about where this began and it really, this sort of automation push really began with universal app campaigns in about 2015.
And that sort of marks the beginning of the move toward greater and greater automation on Google's part for what's that for whatever that's worth. So that's, you know, 8, 7, 7 years ago. So. Um, okay. So for our final item today, we're going to talk about ads, but in a different context, or we're going to talk about marketing in a different context.
Talk about Twitter. Clearly the headlines, the technology news headlines have been dominated for the last couple of weeks by Twitter. First speculation over whether muskets would take over the company. And now speculation about how Twitter will change in the wake of. Uh, acquisition, assuming it's still goes forward because there's some people that have some cast, some doubt on that.
And you know, rather than get into the politics or whether China's gonna hold its relationship over him, or whether it Musk is a reasonable guy or budding fascist or whatever you want to say. Um I think, I think it's interesting for us to think about what is the future for marketers and advertising on Twitter under You know scenario and my own assessment of that would be that it's, even though Twitter just showed an increase in users and it improved results this week in their 200 earnings.
I, I think that there will be negative consequences or there will be a negative impact on advertising on Twitter. If Musk goes forward with many of the things that he's talked about, reducing staff Uh, you know, opening up the content to less moderation and so on and so forth in particular, they're sort of, if it becomes a cesspool, as I called it in the, in Monday's newsletter, maybe that was Wednesday.
Um, I think it was Wednesday actually. Uh, then brands will be scared away. And so I think that, I think there that. Operates. It's a little bit of a disincentive to completely throw content moderation away. But I think also people would just get scared about operating an environment where there's more vitriol.
What do you
Mike: think the impact of opening up the API APIs to alternative uses displays presentations, but including advertising in that, how do you think that.
Greg: Well, I think that's an interesting, I think that's an interesting scenario. That's what existed in the beginning. Of course. And then Twitter shut all that stuff down really.
Maybe before its ads products were you know,
Greg: operational. I don't remember the timeline. I think that's really interesting. Some of the techno technological stuff or the, the, the product changes that Musk has talked about might be very good for Twitter. Um, I don't know.
David: Yeah, I was going to say that I think that the topic you just brought up, Mike, in terms of the, sort of the open API, more open API and We eat.
We even saw the Ben Thompson Stratec repost this week about potentially splitting Twitter into essentially two different companies. Um, and I think that it's, to me, it's an open question as to whether Musk is even going to care about advertising revenue from Twitter, because there could be so much more on the data side of things, which certainly seems to me to have been.
You know, more, more in his wheelhouse than advertising historically,
Greg: so I don't need a licensing get a licensing. Exactly.
David: Yeah. So I would say that, that if I ha if I had to place a bet on where Twitter is going to, where, where Musk in particular is going to make the revenue, he needs to pay off the interest on the loan that he took, that he's taking out to buy the company.
Um, I would probably place it more on the data licensing side than the advertising side. Um, I don't know that there's going to be. I'm not sure I can, unless he has a incredible idea, which he very well may for how to radically alter the advertising piece of Twitter. Like I just, it doesn't seem like a platform that is really very well set up to maximize ads in the same way that Facebook and Instagram are.
You're just not like the, the, the main lining of information. I think that's been Thompson's term doesn't lend itself to. Ads, nearly as well as this more passive, like, oh, what are my friends doing? What are, what are these celebrities up to in the sort of Facebook, Instagram, even Tik TOK
Greg: style things.
So, well, I mean, you know, the interesting, one of the interesting points that was made, I think in an article in the information was that Twitter has, has neglected to really build good performance ad tools. Uh, they just, they have just done a poor job of execution with performance average. So, you know, you may be right.
That must doesn't care about ads, but that scenario, it seems of opportunity. I looked up how much Twitter made in ad revenue in 2021. And it's only four and a half billion as nothing, which is nothing. So you could well be right. That there's a bigger opportunity, at least at current levels with data licensing.
But, I mean, I don't know how much growth opportunity there is over there
David: laughing, but I mean, what, what's Google,
Mike: it's just so amazing to me that the three of us could say, oh, for affiliate, it's nothing. Well, it's not nothing, you know, and it's nothing compared to his net worth and to Google's grow sales.
I get all that, but it's still gotta be so.
Greg: Yeah, so, so Mike, it's obviously not nothing in the abstract, but it's, it's, it's nothing relatively speaking, which is another conversation about the, the, just the concentration of wealth in a very small number of companies. But just to put that in context, Google just announced in Q1 this week or whenever it was my sense of time is totally distorted.
Um, 50. Bill in ad revenue. That's not their total revenues.
David: Well, isn't that just search though? I thought I thought there was another
Greg: layer. Gotcha. Well, search is different. I mean, searches, searches a component of that, but that includes, I think YouTube and in M and a and display, I'm just looking at a time.
Are you can't make me look up all these numbers
David: right now. So nothing, Mike, it was also nothing.
Greg: Well, snap snap is that
Mike: was kind of on the, on the rise sort of on the QT it's got
Greg: it's snap has snap has kind of overtaken Twitter, you know, snap and Twitter were sort of peers and snap has, has has improved and is growing much more than.
But, you know, I think one of the reasons why Twitter is such a heated topic, I mean, this is something I was thinking about this morning, you know, why are people so worked up over Twitter? You know, it's just one of, many of these companies out there is because it's kind of the last place. Uh, there's a kind of common information source, even though that's not really true, you're not reading the same newspaper.
You're, you're seeing a very dynamic product, but it's the it's, it's the one place now where everybody sort of congregates and sees information that used to be the case with the evening news in decades past, and even certain newspapers. And now the media landscape is so fragmented and people are so polarized and there's such a echo chamber you know, kind of.
Activity happening, but there's, there's kind of nothing left like it. And so I think Twitter does represent that at least symbolically to people. And so they don't want it to be become, you know, skew one way or another entirely for what it's worth. But I, I, I mean, I think whether it's whether Musk abandons, advertising and goes to data licensing, or whether Twitter becomes a sort of a toxic place for brands.
It's going to be a place where the future for ads is less bright than what once might have been
David: on that, on that note, on the toxicity. So I wish I could remember who tweeted this, but essentially they were saying that, oh, guess what? There's a, there's a social network that already tried the Musk sort of unfettered, uncensored information model.
It's four Chan. How well did that turn out for them? So.
Greg: Yeah. I mean, that's it, it's, it's another conversation about why things go in that direction, you know? Um, but, but at least, at least one reason may be. Uh, a minority, a relative minority twenty-five percent of Twitter users that are responsible for almost a hundred percent of the content.
So you've got a lot of people consuming. This is like reviews, Mike, you've got a lot of people consuming, but many, many fewer people actually generating the content in this case. And
Mike: actually that's, I have a survey that says just that information, although it's interesting how the non-producers have changed over time, but.
Greg: Okay. And with that, we are at a time and we'll see you next week for episode 63, as always. Thanks for listening and subscribe to near media. Uh, and we'll see you see you in the future.