READY TO BECOME A MEMBER?
THANK YOU!
Transcript
Our transcripts are generated by AI. Please excuse any typos and if you have any specific questions please email info@digitalshelfinstitute.org.
Lauren Livak Gilbert (00:00):
Welcome to Unpacking the Digital Shelf, where industry leaders share insights, strategies, and stories to help brands win in the ever-changing world of commerce.
Peter Crosby (00:22):
Hey everyone, Peter Crosby here from the Digital Shelf Institute. The number one objective for shopping agents is getting each shopper the best product for them every time using as few tokens as possible. Product content is a vital part of fueling that efficient agentic shopping journey. Spencer Millerberg, founding partner of DetailPage, joins the podcast to highlight the most recent news and the foundational strategies for winning the recommendation at machine speed. Spencer, welcome back to the podcast. We are grateful as always to have you and your brain on, so thanks for being here.
Spencer Millerberg (00:59):
It's always a pleasure to be with you, both, uh, Peter and Lauren.
Peter Crosby (01:03):
So you live very deeply in the SEO and now AO world. Uh, you're... I mean, it's stunning to me how you and your organization are able to keep up on top of the constant changes that are happening. And I know a lot of data backs that, that knowledge up. But what a wild time to be in all of this right now. Consumers are optimizing for a, a new audience agents, and we wanna break down what actually matters on the PDP for Amazon, ChatGPT, Walmart, and more. We have a lot to cram into the next 35 minutes. But to start, let's chat about some of the latest news. ChatGPT announcing their, uh, new businesses capability. Gartner's take was that, uh, CMOs must add managed presence to their answer engine optimization strategy. To protect brand integrity, CMOs must govern the content feeding these agents. Okay, so do you agree, and is this in fact a new task for CMOs?
Spencer Millerberg (01:59):
Yeah, I mean, absolutely. Uh, let's just put it simply, what they're talking about here is relevancy, right? Uh, it, it used to be that you could win based upon a bid rate if you're a CMO. You just bid higher and higher and higher and bigger budgets would win. But then bid rates started being impacted by search ring and now you don't even have 32 slots that you have anymore. You have one to three spots that you have within this place. Relevance is the whole game because the agent needs to know who you are, what Lauren wants, what Peter wants, exactly what it is. And so not only are the recommendations that they're going to give you personalized and specific to who you are, but the advertising slots now are going to be really, really specific in this. So ChatGPT was the topic of that Forrester paper, but the rules end up becoming the same no matter, uh, if you're talking about them or if you're talking about Amazon or Walmart, uh, that you have here, uh, because it's matching the query of what the customer's asking for, matching the persona, and that can be who it is, when it is, where they're gonna use it, why they use that product.
(03:06):
And that's what makes it to where you can win those backups that you have there.
Peter Crosby (03:10):
Spencer, I also noticed that Gartner said that OpenAI's business hub further commoditizes AEV tools with native metrics and business agents. Can, can you give me your response to that? Like what does that mean to you and, and do you agree with it?
Spencer Millerberg (03:26):
Yeah. What they're really saying here is that they're giving you the information that it takes on where you're winning. The part that's, uh, the part that I don't agree with is they're not telling you what you're not winning. Uh, the, you know, there's the old classic you, there's things that we know we know and there's things we know we don't know and there's things we know we don't, we don't know we don't know.
Peter Crosby (03:46):
Oh, Donald Rumsfeld, that was a callback.
Spencer Millerberg (03:49):
Yeah. <laugh> It was a bit of a callback, right?
Peter Crosby (03:52):
Yes.
Spencer Millerberg (03:52):
Uh, but like this is, um, I think what they're doing is they're telling you the things that we, we know we know, but what we don't know is the things that we don't know, right? And I see it again and again and again for, uh, businesses. They'll go out and, uh, they, they have this inside speak and they oftentimes talk about things that, uh, they think their customers care about, but it's not necessarily stuff that customers are querying. So you can be winning and ChatGPT's, uh, metrics will show you, "Hey, you're winning on this," but it may not be something that the customer actually cares about.
Peter Crosby (04:28):
So it's really the, the need here is understanding the constantly shifting opportunity for improvement in what you've got, and those opportunities need to be actively uncovered - Yep. Which is not something that currently businesses is doing. It's just kind of telling you how you're currently performing. Is that -
Spencer Millerberg (04:44):
Yeah, e- exactly, because, um, you know, like this is, uh, when I worked at Amazon, people would come into me and be like, "Hey, I grew 78% last year. Is that good or is that bad?" And, uh, that's exactly what's happening here is this, uh, this dashboard's gonna say, "Hey, you grew the X percent, but it's on stuff that you have that you know. What we really care about is share," right? Back then, at Amazon, they cared about share, because if they grew 78% but their competitor grew 180%, they were falling behind. Same thing happens here is, uh, you can have these queries and you may be winning on queries, but you need to have the share of the entire voice that you're talking about here, not
Lauren Livak Gilbert (05:25):
Just
Spencer Millerberg (05:25):
Your specific share.
Lauren Livak Gilbert (05:27):
And do you see... I just, uh, have an additional question on this, Spencer. Do you see this being, getting expanded from a ChatGPT perspective that they're gonna offer more like that? 'Cause I'm, I'm curious, like, from your space-wise, like, this is a focus area for you. Yeah. Is that something that you're thinking about or...
Spencer Millerberg (05:44):
Yeah, we actually just launched a, uh, share of AI voice measurement tool that has ChatGPT. Gemini and Alexa for shopping are the three biggest ones, because that's, uh, representing approximately 90% of all customer queries from a shopping perspective, uh, that you have. And so you need to know, um, that share of voice, how you're cons - how you are across that market, uh, that you have there, uh, not just on one. And so are they going to expand? I mean, ChatGPT obviously is not going to expand beyond ChatGPT because that's the knowledge base that they have. Uh, but from, uh, CMOs, you know, this becoming commoditized, yes, you have to have this. This is one of your core metrics. This is the metric that tells you, are you winning in this new age of shopping?
Lauren Livak Gilbert (06:28):
Makes a ton of sense. And, you know, we've already talked about a lot of acronyms on this podcast. Uh, the new one was AEV for me, answer engine visibility. I feel like everybody's using acronyms now, but let's go with the acronyms of SEO versus AEO and how they work together because SEO is not dead. It is supporting AEO. So can you talk a bit about those kind of two things working together?
Spencer Millerberg (06:52):
Yeah. Um, I would say not only is SEO not dead, it's actually the majority of what you have to do, because SEO is about what the product is, um, right? GEO is about that personalization, the who, when, where, why. Uh, Amazon talks about it in Cosmo search. They have a great, uh, white paper if you need some good sleeping material. Um, and when I talked to Amazon though, I, I'm, uh, was meeting with one of the guys at, uh, a Harvard presentation I did a couple weeks ago. He actually said it's the most important part of what their personalization engine is. In fact, they're not only, uh, they're working to expand that, uh, to where it's personalized even more so that even when, uh, you, you search and who within the household searches, you would get different results based upon who searches, right? So if I search for headphones, I'm gonna get something about office and running.
(07:42):
Uh, my son's gonna get, uh, information about our headphones about gaming and school, right? Different things for even within the household who searches based upon the device they're searching on. And now, AEO is that last one there. That's all about access. That's the linkage. Does it actually index? Does it think that those things work together? Um, so I can give you an example. Sometimes, uh, it's easier with an example. Okay, great. Um,
Lauren Livak Gilbert (08:08):
So - Yes, please. We love examples. <laugh>
Spencer Millerberg (08:11):
Okay. So think about it this way. So if I'm shopping for, uh, I don't know, say a grill that I have here, it's, uh, it's, uh, tailgate season that we have. And SEO is gonna be, it's a three-burner grill. It's, uh, provides even cooking, black cast iron grates, whatever the case may be that you have here. That's about what the product is. The GEO is that personalization where if they're talking about perfect, for tailgating, the image has it on a camping, uh, place. It's perfect for apartment dwellers, holiday cookouts, Fitz12 hamburgers, right? Like, that's w - about that. Where am I gonna use it? Who's gonna use it? And then AEO is, that's the test that you're doing. You're indexing to see if that's working. So, like, <foreign>, like, about 10% of queries are now in Spanish on the English site that you have here, and oftentimes that linkage is missing there, or maybe portable grill or tailgate grill.
(09:08):
Those things don't link because for some reason, they haven't quite linked those things together. So SEO, GEO, AEO, you have to have all three of them working in concert with one another in order to be that top recommendation, because remember, you only have one to three recommendation slots now, not the 30 or 40 that you used.
Peter Crosby (09:29):
So why don't we start at the sort of go, go back kind of in a way to basics of the, the PDP, like, where a lot of... I mean, obviously, there's so many places now where content needs to be across so many places that now the, the machine shopper will, is looking to see what's, what's true, what's real, and, and what is it the right context. But the PDP still has a lot of power in that, in that argument. Uh, and so I would love to just start, have you walk us through one, one, is that true? Like, is the PDP still... I don't know what royal title you wanna give it, but, um - <laugh>... Is it still up there in the royal family of content? And, and if so, uh, what are the things that, that our listeners should be thinking about, um, in their PDP work over the next couple years?
(10:19):
Well, probably all we know is the next six months, but, you know, give us a sense.
Spencer Millerberg (10:23):
Yeah, no, PDP is reigning supreme on anything that is discovery related, right? If it's, "I already have chosen the brand, I've already chosen the product," that's where you start getting the brand sites and brand pages coming into play. But we're really finding for anything about, "Hey, I need some running shoes, I need a new morning moisturizer, I need any of those things," the PDP ends up becoming the most important part because they're going to the retailer site over 80% of the site, of the times, the citations from ChatGPT at Gemini are going to a retailer site and a retailer feed that you have there. And you think about why, like, it's all about the money money, right, right? Uh, we should sing a song about that, I, I think. Uh, but, you know, it's all about the money because you don't want to go and scrape a hundred different sites for this.
(11:12):
It's really, really, really expensive. Instead, what they're doing is they're taking that retailer feed and they're making it toward that's a cheap way in order to get that information, and so now we're, we're gathering that, that data together. Does that answer your question first, uh, your first part of the question of -
Peter Crosby (11:27):
Yes. Does PDP matter? It, it matters, and, and there's a reason why it matters, which is awesome.
Spencer Millerberg (11:31):
Yeah.
Peter Crosby (11:32):
So now that we can, you know, confirm that for the listeners, then what do they do about it?
Spencer Millerberg (11:38):
Yeah. So, you know, we talked about a magic quadrant beforehand. Now let's talk about a magic triangle that you have from a PDP. Like, think about, uh, that triangle being inverse, that the top of the PDP is the most important part of that triangle, the wider part. That is where, um, you're gonna have the higher level of indexing and higher level relevancy that you have at the very top, right? And then as you go down further into the, that, that's gonna matter a little bit less. And so what that title, that highlights, those are number one and two, bullets number three, description four. And then Q&A, this is major shifts happening right now. That one stay tuned because we're seeing Q&A, uh, within Walmart be taking, like, one of those number one and two slots that you have, Amazon's still at four or five slots that you have down there.
(12:27):
And so it's a little bit different depending upon where it is, um, uh, where you have on the page from a location perspective. Does that make sense so far, Peter?
Peter Crosby (12:35):
Absolutely.
Spencer Millerberg (12:37):
Yeah. And so, like, again, like, let's just hammer it home with an, uh, with an example that you have here. So, like, I was just looking at a just ingredients, uh, protein powder before we got on, and it's winning on joint comfort, skin elasticity, and protein baking, right? The reason why it wins on those things is because that content is really high up on page. Remember, we talked about what the product is, who uses it, where they use it, right? They're winning on these matches because it's high up content, exact match. It matches to a little bit older demographic that you have there. But then they're losing on low sugar, post-workout, fitness crowd, anything about, uh, that put fitness crowd. Why? Because zero sugar is only in the ingredient panel. Like, the post-workout is not really mentioned that you have there. So they lose to some of these native, uh, digitally native brands, like Legion Way is the one that's showing up within, uh, really, really high because it's in the item highlights, it's in the top bullets, it's in the titles, it's in the images.
(13:44):
They're reinforcing that for that zero or low sugar, where just ingredients is losing because of that, um, because they're not matching on those top spots with the top, uh, content.
Lauren Livak Gilbert (13:54):
Did you see this change more when Amazon changed their title requirements to have that highlight section, that brands need to focus way more on what's put in that kind of second layer of the title versus when it was just one field?
Spencer Millerberg (14:08):
Yeah, that's where we're really starting to see the GEO take effect, um, because the SEO about what the product is, is really within the title, that GEO of the benefits and product highlights, those really start getting to that match, to those Cosmo search dimensions within Amazon. And, um, a lot of people don't realize that Walmart changed their title requirements at a very similar time as well, right? And so then take it back to the very where, place we started, Peter, like, when you talk about that CMO and how they're doing this, like, a lot of CMOs are realizing that that content within that title and highlight area, that actually is the exact content that is now appearing within their advertising. It's the same characters, the same character limit, all the same information that you have here, so you have to win the match and even get into the paid spots now, and it's influencing the organic search as well.
Peter Crosby (15:05):
It's just, you know, uh, Spencer, you, you have been generous enough to do a few webinars with us and, uh, and all of these things. The, the sort of common, the common thing that comes away with me is that this is going to continually change, well, forever, <laugh> but certainly for the foreseeable. There, the rules can, can be changed almost at any time, depending on, on what they're seeing. And it's, you know, the, the sort of idea that we've been talking about a lot is, is the idea of needing to work to do commerce at machine speed.
Spencer Millerberg (15:41):
Mm-hmm.
Peter Crosby (15:42):
That there's just no human way to keep up with all this and find the growth opportunity that's available in the moment where you have it. And I'm just wondering how you think about that sort of speed of change. Is it as sort of almost impossible for a human to do and win, uh, and how you, how you feel about that?
Spencer Millerberg (16:04):
Yeah, no, like, it's, uh, it, uh, sometimes people ask, "Can I go and do this on myself, on my own?" I'm like, "Sure, you wanna try this one?" It's like trying to take an MRI, uh, by yourself, right? You gotta have some specialized equipment in order to make this sort of thing happen. Um, you think about the speed at which it changes, um, but then also the speed at which the algorithm is changing, the customer queries are changing that you have, and then making sure you apply those things into the content that you have. So you really need to start with the AI machines in order to, uh, change your AI-related content in order to get picked by AI that you have here. You just can't do it at the speed of humans anymore.
Lauren Livak Gilbert (16:49):
And do you suggest to brands that they start small and then wait and see what changes? Because we know most large organizations can't make the amount of changes needed, and they don't have the system to move at machine speed quite yet. That's something they need to build.
Spencer Millerberg (17:04):
Yeah.
Lauren Livak Gilbert (17:05):
So like - Yeah. Do they start with a couple of skews and then see what happens? Do they do their whole portfolio? Like, what's your guidance?
Spencer Millerberg (17:11):
Yeah, what we do is we, um, when clients come in, what we do is we run what's called an ROI dashboard. I'll just tell you the metrics. You can run it yourself, you can run it with us, I don't care. Like, um, but what we do is we're gonna start by, uh, taking the entire portfolio of products that you have here, and then you figure out how much, how many units of sales that you have, and then measure the gaps that you have. Because many times, a small tweak on a really large item will be something that will work really well, but more common is that middle tier of your, uh, product portfolio that you have, items three, four, five, six, seven, eight, 10, right? Like, those have pretty much been neglected for a long, long time. And so taking and doing a big catch-up on that mid-portion of the portfolio, more often than not, is the part that we're seeing for many, many brands.
(18:01):
Now, one thing to add to that, many digitally native brands already are doing this intuitively. And so what we're seeing is this, uh, democratization of the shelf that has been going on for a long time. Um, those digitally native brands are, are a little bit closer and more in, uh, able to make fast changes. And so they're updating content on a quarterly basis to be exactly what the customers are looking for. And because they're moving at the speed of AI and at the speed of e-commerce, they're gonna outpace these brands who don't adapt and don't move at that speed and that pace.
Peter Crosby (18:39):
That actually gave me a little bit of relief that you said quarterly. Like, part of me sometimes though, you know, something happens on a Tuesday, you should be doing it by Wednesday so that you're getting it. But, but quarterly is, is sort of a reasonable timeframe to be deciding what to do about the data that you're getting. Is that true or...
Spencer Millerberg (18:57):
Yeah, we, um, it's interesting because when we test this in, this information, um, what we're finding is in retail, it takes between four and 12 weeks to index. And within AI, it's actually taking much, much longer. It's taking about 12 to 20 weeks in order to index that you have here. And so if you were to try and index every single week or every three weeks or every four weeks, think about it like a seed that was planted in the ground. You pull it up and go, "Hmm, is it growing yet? Nope. Stick it back in the ground." You end up killing what happened there because, uh, the way that it works, um, again, I was talking with Amazon at Harvard a few days ago, and, uh, what happens is they put it in, it tests, and it sees if there is adoption by the customer.
(19:44):
And as that gets adopted, then it pushes it a little bit higher, and a little bit higher, and a little bit higher. And that takes that eight to 20 weeks that you have, depending upon if you're talking about retail or AI, in order to see that linkage start to take effect. We typically, within Amazon and Walmart, start to see it level off around 12 weeks, is where you see. And then within ChatGPT and Gemini, we see it level off around 20 weeks is where we start to see, see that level off. So make it quarterly. That's why quarterly really ends up becoming recommendation.
Lauren Livak Gilbert (20:18):
So question to that though, because I know some data just came out that, like, the half-life of a citation is 11 days. And so, like, the citations outside of the PDP think, like, brand.com and all of the other places is, is happening, happening on a more frequent basis. So is this recommendation in terms of content changes just for the
Spencer Millerberg (20:38):
PDP,
Lauren Livak Gilbert (20:39):
Or does it also extend to all of the other places outside of the PDP that content needs to be updated?
Spencer Millerberg (20:45):
Yeah, good question. Editorial, uh, editorial content is largely what we'd be talking about. We don't see the brand reference, um, there, or we don't see the, uh, AIs referencing the brand page very often. Only about 5% of citations are going back to the brand.com site that you have here. A year ago, that was different. It was much, much higher. But today, we're seeing the brand.com site really not being referenced that you have. So the editorial sites, right, like, uh, camper, uh, like campinggoods.com or whatever the case may be, um, we're seeing that those, um, do have a pretty short shelf life and they do need to continue, uh, continue being updated. But what's interesting on that site, on the editorial site, is how, um, the, uh, you have to watch which ones are important to that, uh, AI. So Reddit was a really interesting example of this.
(21:37):
Reddit was, like, all the rage, all of last year. We really saw that thing really, really, really high. But then I think it was, uh, around August of this year, on all of our trackers, we saw this thing just fall completely off of a cliff that you have here. And, uh, Reddit really fell out of favor with ChatGPT, Gemini, and Meta, all those others. They just fell right off, and those citations continued to stop. So then, what does that mean for you as a brand? You have to know not only that we have to keep our content fresh on these editorial sites, but we need to know which editorial sites the brands are actually looking at, or the, uh, AI is looking at in order to make those decisions and make those recommendations that you have here.
Peter Crosby (22:19):
Do you think that drop has something to do with the reason that Reddit just said they're just gonna start a reviews sort of dedicated section of their experience?
Spencer Millerberg (22:29):
There was, um, a pretty significant backlash that, uh, was... I, I think that there was a credibility gap that was going on there. Let me see here. Yeah. It was, uh, August 2nd was the exact, uh, date that you had. And ChatGPT, I think, was having a credibility gap there, saying all that you're really doing is grabbing some, uh, some chat, uh, Reddit reviews and pushing things up. And so you, we, I'm looking at one right here on my screen, and Outdoor Gear Lab, boom, shot up really, really high. Clever Hiker, Travel Leisure, Outdoor Life, all those started to go up, and they're taking more credible, uh, editorial sites to take real, uh, basis that they have there as opposed to Reddit, and it just dropped off. Or expertise. It was just expertise.
Peter Crosby (23:13):
Fascinating. Yeah, fascinating.
Lauren Livak Gilbert (23:15):
Gosh, we could talk about this all day. All right. Let's go back to the PDP. So we talked about the titles - Sure. The written content. Let's talk about images. I know there's a lot of questions around are images being viewed by AI agents. We know that Alexa for shopping looks at them. Uh, so tell us a little bit about images and kind of where brands should think about that.
Spencer Millerberg (23:35):
Okay. So images, let me give you a couple of data points just to start with. So on Amazon, uh, over 95% of customers who make a purchase will look at all available above the fold images, right? Customers are not reading bullets. Only 4% of customers even open the bullets, let alone read those bullets that you have here.
Lauren Livak Gilbert (23:54):
Okay? Humans like pictures.
Spencer Millerberg (23:55):
<laugh> It's a perfect way to say it. Like, humans like pictures. But on the opposite side, AI really likes text, right? So text gets you to the page, images help you convert. That has historically always been the case. Now, on, in March, uh, 26th of this year, ChatGPT started indexing images. And on April 26th, a month later, Amazon, uh, started listing images within their Alexa results as well. And so we're seeing those move up funnel into discovery for images that you have here, right? Does that make sense so far?
Lauren Livak Gilbert (24:33):
Yeah, that makes sense.
Spencer Millerberg (24:34):
Okay. So, like, uh, like, the criteria is really, really simple when it comes to images. We, we had this example of a drill, and it had, uh, about 16 different specs that were there, and it was pointing to the drill and showing all these pieces. AI actually, uh, just skipped right over those, because what it's preferring is a couple of things. First of all, what it wants is a text overlay, but it has to be a single-minded message that you have here. It doesn't want 16 things going around there. Instead, it wants to know this is exactly what, uh, what it wants. And then the lifestyle image that goes along with it too. And that lifestyle image is really fascinating. Go back to the persona. It wants to match to the persona of who it is, right? And so, like, I'll give you an example.
(25:21):
Like, if you search for school snacks, and on the bottom, it says, "Great for school," that you have here, then AI wants a pat on the back. It's like, "Yes, that's the one." If you're searching for, "Hey, what's something good for bachelorette party?" And ring pops pops up and it says, "Great for bachelorette parties," boom, right? It just nails it. Nails exactly who it is that's searching, what it is that they're searching for. And I could go on and on, like routers, battery packs, et cetera. We see it constantly that the AI wants that pat on the back, it wants that match. And the easiest way to show it is to show that customer, see this image, boom, this is exactly what you were searching for.
Lauren Livak Gilbert (25:58):
And the agents are inherently lazy, right? Like, they d - they don't wanna, like, dig through... Uh, correct me if I'm wrong, but, like, all of that content, they want, like, a very clear, "This is what it is, this is my match, we're good."
Spencer Millerberg (26:09):
Yeah. And they're gonna utilize more of the written text to get you within the beauty lineup, but then, uh, the one that gets selected within that beauty lineup oftentimes is because that image is the one that's showing exactly what the customer is caring about, right? And so, get you there from the written text and then get chosen because of the image that you have.
Peter Crosby (26:31):
And what about enhanced content? So, obviously, a, a lot of our listeners have invested - Yeah. Uh, I think officially a crap ton of money in creating -
Spencer Millerberg (26:41):
<laugh> The technical term.
Peter Crosby (26:43):
Exactly. And, uh, how do you advise brands contin - do they continue that investment? Do they shift it elsewhere? Uh, is it helping them across the spectrum?
Spencer Millerberg (26:53):
Yeah. Yeah, so on Amazon, you'd call this A+ content. On Walmart, it's enhanced content is the word, the phrase that they use. It used to be that there was all absolutely, like, no benefit to doing this stuff. In fact, I remember when we were at Amazon, when we had to make a budget, we'd be sitting around and be like, "Oh, who can we sell A+ content to?" Right? And, uh, we knew it didn't do anything, but we had to make our budget. Now, um, there's just one small difference that's there. Very recently, about, um, it was October, uh, or maybe it was end of September, we found we, we are, we index and scrape fields, like, millions upon millions every single day. And what we found is there's one nuance that is starting to get in there. It's the A+ premium Q&A section. That piece is now starting to index as of that late September of 26, and it's free to all Seller Centrals.
(27:42):
It should be free to Vendor Central, but it's a bit of a mixed bag. Sometimes your vendor manager, uh, wants to try to charge you for it still, you know? Uh, but what's happening here is, uh, again, go back to the money. It's all about the money, money, money. This is what can be included in the retail feed. If it's standard, it's really, really easy to include in the retail feed. If it's not a standard field, they don't include it in the retail feed. So ChatGPT does not get enhanced content from Walmart. It's zero value that you have there. Um, same thing with Alexa. It's now only indexing that Q&A. Everything else, non-standard field, doesn't get indexed.
Lauren Livak Gilbert (28:20):
But does it get indexed in Sparky on Walmart? Nope.
Spencer Millerberg (28:24):
It's not.
Lauren Livak Gilbert (28:24):
Okay.
Spencer Millerberg (28:25):
No. We're, it's one of those, uh, we're not finding... W- what I should say, Lauren, is we're not finding current evidence that it's getting - Today. In today, right? Like, the, catch me next week, this might change that you have. But today, we're not finding any evidence that Sparky's indexing the, uh, enhanced content. Also, when it comes to Sparky, we're also not finding as much consumer adoption on Sparky as we are on Alexa for shopping as well. The consumer on Walmart still is continuing to use that standard, uh, uh, top search bar that you have there, um, which is AI enhanced and AI enabled as well.
Lauren Livak Gilbert (29:00):
If we think about Q&A for a second, you mentioned brand.com and how it might not be indexed as much. What about adding Q&A to your brand.com and, like, taking that same content and maybe even expanding it more there?
Spencer Millerberg (29:15):
Yeah. Okay. So again, think about the consumer journey. If they're thinking about discovery, that's really not gonna be taken into effect, uh, that you have when they're trying to discover the new water bottle, the new headphones, the new whatever it is, right? But once they are down and they're researching a specific product or p - specific brand that they have there, that's where that Q&A and that's where that brand.com, uh, content starts to come into play really, really strong. Like, what is the battery life? Uh, can I charge it with, uh, uh, wireless charging? All those different pieces, those come in, but only after the customer is already down the funnel.
Peter Crosby (29:49):
Yeah, I think we're, we're asking these question because I, I keep looking for something that our listeners can, like, stop doing or stop in- invest less in because it... And Brand.com is a perfect example. Like for the past, I don't know, year and a half, we've been saying that's where your truth lives. That's where you gotta make sure that that... And, and so I, I don't wanna over index on your, not to overuse that word, uh - Yeah. On your brand.com, they don't care about anymore. But if it's, if it's that big a shift and do you think that shift is permanent or could they decide... They could, right? Like a year from now or six months from now, they could go, "Oh man, wherever we were looking is not working out as well as brand.com used to." Or what's your - Yeah. What's your, what, what are you telling your clients to do with their brand.com sites if they're asking even?
Spencer Millerberg (30:39):
Yeah. It's possible versus probable, right? Like is it possible they go to do - uh, brand.com? Yeah, it's possible, but it's not probable. Like you think about the consumer journey where you are. I mean, Peter, when was the last time you went to a brand's website, you personally?
Peter Crosby (30:53):
Well, I just mean about the agent going there. I'm, I'm thinking of brand.com as a machine customer.
Spencer Millerberg (30:59):
Absolutely.
Lauren Livak Gilbert (30:59):
And an authoritative source, right? Because that's like the brand voice.
Spencer Millerberg (31:05):
Absolutely. But think about what we just said there. An agent should be somebody who represents you. And what we're finding is the agents are acting much more similar to humans in their shopping behavior now. And humans are not going to the brand.com sites in order to do shopping. And then the agent isn't going to that brand.com site as well. Well, we want it to be true. We want there to be this brand authoritative site and we want it to, to have all this content. We're finding that the only time we ever would go to that brand.com site is if we had a real question about that product that you have here. And most of the time that's afterwards.
Peter Crosby (31:41):
Very far down the funnel. Yeah.
Spencer Millerberg (31:43):
I've already selected into that. And so you talk about like, uh, prioritization, where do I need to go? The retailer content is driving almost everything now. It... AI is cutting through all of the crap and it's going in and it's saying, "Okay, I'm gonna compare this product versus this product versus this product, and I'm going to do that the way that a human would, which is the content on a retailer site." Then that content drives the advertising. And then only after we've selected that does that product end up going to that brand.com site. The two,
Lauren Livak Gilbert (32:19):
The two examples I've seen like working with brands around brand.com is to dispute something incorrect about their brand. Would you agree with that? Like someone was saying something wrong and they used the brand.com to influence like a ChatGPT around the correct content and what they should be saying. And there's been proof points of that. So that's something that I've seen it to be strong for. Would you agree?
Spencer Millerberg (32:42):
Yes. Um, but again, like is it strong there? Yes. But where are more effective places to do that? I would argue that going into editorial citations, we're seeing editorial citations, uh, coming in right behind that retailer content that's there. And, uh, what they call those are proof points, right? And so what they're looking for is proof points and independent proof points are going to have a higher authoritative weight than something on the brand site does itself. And so I would move that into editorial or into that retailer content because that's just the way that the, the agents are shopping.
Peter Crosby (33:19):
That really reinforces the importance of what, what a, a lot of people have been talking about, but certainly on the podcast we have that collaboration between PR brand comms tightly with the e-commerce digital folks so that you're covering all those bases. And it sounds like even more importantly now to get the, the editorial third party site strategy underway or reinforced.
Spencer Millerberg (33:43):
It is. Yeah. Yeah. Having that, uh, we started with AEV, that, uh, that, uh, voice that we're really looking for. Uh, that is now a little bit more expansive than has been traditional within that retailer space. It used to be we only had to control our retailer and our brand content. Now we're moving further into that PR space.
Lauren Livak Gilbert (34:06):
So Spencer, how the heck do brands actually measure any of this? Because we're talking about all the things you need to change and what matters, but at the end of the day, if you can't measure it, you don't know what the impact is. So how are you suggesting that brands focus on that?
Spencer Millerberg (34:19):
Yeah. So, um, uh, the share of voice ends up becoming the number one most important measure tool, uh, measurement that you have right now. And, uh, we have a great, uh, share of voice measurement tool. You can come to us, you can come to other people, but I don't care who you do, just make sure you're measuring that share of voice. And then the second piece, you need to make sure, uh, that you, uh, are going to take action on that. This has to have prioritized action on what's gonna move that share of voice the most. It may be... Like we were just looking at one the other day, um, and we were, uh, we were looking at it and, uh, for toilet seats, it was actually more important for them to update their target content than it was for them to update their Home Depot content, because Home Depot was already hi - killing it.
(35:06):
They were already well above their, uh, share of voice on Home Depot. Um, but Target, their content had been neglected for a long, long period of time, so they needed to fix that. We were looking at another one for, um, tents for Coleman, and when we were looking at that one, that was editorial content they really needed to fix. And they needed to fix that on Outdoor Gear Lab, right? Like they needed to fix that content there. And so you have to have, um, the share of voice, but it's only as good as the actions that you take from that.
Peter Crosby (35:36):
Well, Spencer, this is, as always, and it seems like will always be an incredibly fast-moving field that I know e- everyone's trying to figure out, uh, what to do about it. A lot are on, on the path. And so when you, when, if, you know, our listeners are all probably in some stage of figuring out how they're gonna invest in this area, and what's kind of... I- if you contrast with even six months or a year from now, what, what have you seen the, the people that come to you today, the new, the new clients, are they further along? Are they more sophisticated? Like, are you feeling like now everyone knows they need to do it and now it's, "Okay, what do we do?" Or where do you th - feel like, uh, the best folks are, are... When they come to you, like, how ready are they to move and what's that current state like?
Spencer Millerberg (36:30):
Yeah, uh, I think that there's a... The, the difference that I'm seeing is everybody knows they need to do it. Not everybody knows, number one, what to do, and then number two, has an organization that is aligned to making things happen fast enough that you have, right? That, uh, those two things, uh, really become critical, is aligning the organization behind what to do, and then the speed at which they can do things. Like, let me just give you a very simple example. Um, we have one client, it is going to take them, I'm not kidding, seven months to approve content. They have to push it through legal, they have to push it through brand, regulatory. They have 18 steps along their content approval process that they have here. And if you approve content seven months ago's content, you're getting seven months ago's results that you have here, as opposed to, uh, we have another organization, very, very, very large, Fortune 500 company, they have realigned, uh, their content optimization process, uh, to where they can approve things in about four days that they have, right?
(37:36):
That's the speed that you have to go. It's totally possible. And if you think about that from a framework perspective, what is it gonna take me to get this content approved in four days? That's the speed at which they have to move.
Peter Crosby (37:48):
What a great takeaway. I feel like that, that conversation needs to urgently happen <laugh> at places that are struggling with multi-day approval cycles. And often that... The roadblocks are not within kind of the people who are looking to make the money. They're the people who are trying to lower risk. And so having that conversation and where real, you know, where does... Is that, is that your experience?
Spencer Millerberg (38:18):
It absolutely is. And not only that, Peter, but, like, we have great tools now that can do this, right? Like, um, uh, when we push forward content to our customers, what we do is we're actually prioritizing it based upon risk, right? Have we enhanced or created any new claims that are there? That would be number one. If so, that's a higher risk and those things should be approved and are reviewed really, really tightly. Anything that we have just modified wording or said it in a slightly different way, that's bucket number two that you have. And then everything else that we've just taken the content and re-shifted it or re-jiggered it or put it in different places or, uh, added a persona or where they're using it or so forth, that's really a low risk bucket of, of, uh... And it doesn't need very much approval.
(39:02):
That is stuff that you can really push through really quick. So if you use the technology to speed up that process, it goes a lot, lot faster.
Peter Crosby (39:12):
That's... Spencer, as always, uh, we are so grateful for you supporting this community in the way that you do. And also, I know always you have made you and your company generously available to folks that are trying to figure out those steps they wanna take, whether they do it with you or not. And so for those that are interested in sort of getting a pulse on where, where they are at, et cetera, what's the best way, what are the best ways to reach you?
Spencer Millerberg (39:37):
Sure. Um, uh, LinkedIn, pretty active there, or just email me at spencer@detailpage.com. Um, happy to help. We're just here to help, right? Like, I do this because I love it, and, um, we'll give you free reports, we'll give you whatever it takes. Like, we wanna make sure that it's valuable to you when you contact us. So, yeah, feel free to reach out, spencer@detailpage.com.
Peter Crosby (39:56):
Thank you so much, Spencer.
Lauren Livak Gilbert (39:59):
Thanks so much, Spencer.
Spencer Millerberg (39:59):
Thank you. Thanks for having me, guys. Real appreciate it.
Peter Crosby (40:02):
Thanks again to Spencer for bringing the data and recommendations. For more of that, become a member at digitalshelfinstitute.org. Thanks for being part of our community.