Episode 5 – Eva Dong

Filipe Santos
00:00: – 00:00:56

Consumer behavior is constantly evolving, and businesses are turning to AI powered solutions to streamline their marketing operations, enhance customer interactions, and stay ahead of the competition. Join us as we determine powerful strategies and methods to stay on top of AI changes. Today, we’re diving into the deep end with an episode you’re not going to want to miss. You’re listening to phraseology plus a I with your hosts and guides, Filipe Santos and Miguel Santos. Follow us as we uncover everything you need to dominate search rankings, and increase your online visibility. Get information bridging SEO and digital marketing with artificial intelligence. Staying updated on the latest tools, technologies, and strategies to make sure you’re always one step ahead. Phraseology is the podcast show to keep you in the know.

Filipe Santos
00:00:57 – 00:01:53

Without further delay, let’s get into the episode. We’re joined by a special guest, Eva Dong, the senior expert manager at McKinsey, an experienced AI marketer and also on the board of the library project. Eva has held transformative leadership positions at McKinsey and Company, as well as Visa, and focuses her technical expertise in data science to empower decision making and to create significant impact for companies. She’s also a global citizen, having worked across the United States, United Kingdom, Greater China, and South Africa. Today, she’ll provide some vision and tips on using AI effectively to achieve business goals. I’m so thrilled to have you on the show. Why don’t you take a few minutes just to talk a little bit about your background and your expertise for the folks that are listening?

Eva Dong
00:01:54 – 00:02:13

Sounds good. Alright. Hello, everyone. Really excited to be here today and share my experience with you. As just said, I’m Chinese. I moved to the states and studied financial engineering at University of Michigan Ann Arbor. And I started my career as a data scientist at Visa. At that time, cybersecurity was the hottest topic.

Eva Dong
00:02:13 – 00:03:02

So I become data scientist in cybersecurity, very luckily. And then soon after that, I realized even though I really, really love the data science and the data part, cybersecurity wasn’t really my thing. So from there, I pivoted my career to marketing, but I never forgot my data science and AI side. During my 8 years at McKinsey, under our marketing and sales practice, where I use a lot of big data, machine learning, deep learning, or even AI LM tools to drive marketing efficiency for my clients. Today in the episode, like, we will talk several aspect of that. In terms of my experiences, I changed several offices within McKinsey. I started from the San Francisco office, moved to New York office, then to move to London office. And the firm really generously offered me many short term rotations.

Eva Dong
00:03:02 – 00:03:47

I did a client in Cape Town. That was like such a wonderful experience, opened my eyes to the African e commerce market. Then I served another international clients in the Greater China area. Again, it gave me a lot of insights how a global brand need to localize in the local market not only on the marketing channel wise, but also on the wake of communication, the way of branding, and the way to communicate with their clients. I also served a global brand in Greater China area. It opened my eyes how the global brand need to localize for certain markets, not only in terms of the marketing channel, but also in terms of how to communicate with their client, what to communicate when and how to express the branding. Overall, really, really excited to be here.

Filipe Santos
00:03:48 – 00:04:17

That is amazing. I especially love the traveling aspect because, you know, I I myself also just absolutely love travel, and I think that we learn learn the best things when we’re out and about and seeing how different cultures operate. And in your case, you’re also doing work, so you’re learning how they operate in the in their economies. Really, really, really awesome. Great to have you on. Eva, the how are businesses utilizing AI to streamline marketing operations? Like, what’s your take on that? Have you seen anything in particular?

Eva Dong
00:04:18 – 00:04:50

Yeah. Yeah. I think the entire chain of marketing can use AI. And then I can talk about where it’s more mature and where it’s just beginning. But but really, I think it can be implemented on every single step along the marketing operation. So it usually starts from research. What is your brand? What is your target audience? How are they look like? Where are they? And then with that, you gain the data, you do the analysis, and then how you’re doing the analysis can also use AI. And from there, you develop a strategy.

Eva Dong
00:04:50 – 00:05:37

So that strategy part can be from people’s experience and or it can be combined with AI. And from there, you go to the creative part. That part is probably the most mature area for AI in marketing right now. So here, for example, I’m talking about how to use gen AI to create messagings, hashtags, different kind of images, or the same image with the same product that was different backgrounds. So I think AI can be implemented in every single step of market marketing operation. And then I can talk about each step is more mature and which step is less mature. So starting from the very beginning, usually start with research and analytics. So this part, you need to get data to know what is your brand, who is your target audience, how do they look like, what do they like.

Eva Dong
00:05:37 – 00:06:22

So this part, you can use a lot of AI to grab that information, or you can do it manually yourself, which is more traditional, like survey, that kind of stuff. So that’s just one area I think is less mature. People just start doing it. And from there, you move on to strategy. Based on everything you know, how do you develop a strategy, not only for your brand, but also for your marketing. So I think that way is chain this part is changing a lot. It’s it was pretty immature on the experience, your experience, your branding team, all the mark brand marketing experience combined with AI to develop the best strategy. Then after that, it come into the creative execution.

Eva Dong
00:06:22 – 00:07:00

I think this is probably the most mature part in in terms of AI adoption. So here am I talking about how to use AI to generate messaging, hashtags, images, and then lastly today, like, OpenAI released the Soarer so it can generate video for us in in the in the future. After that, then you can think about distribution and optimization. That’s all about, okay, on each platform, like Google platform, Meta platform, how do you use AI to increase our efficiency? And last but not least will be the customer care. Like, after people finish all this conversion journey, how do they like it? What do they dislike? How do you take care of them?

Filipe Santos
00:07:00 – 00:07:39

That is awesome. I mean, I I wanted to comment a little bit on the part that you mentioned about, like, creating assets and taking care of, you know, a customer the customer journey. One of the key things that I I’ve been, like, interested in in getting a take from from Yuan is do you like, I I it feels like an arms race in terms of everybody coming and getting better at producing video and images that are generated from AI. Like you mentioned, Soera, it’s it’s just like, Soera now is, like, starting to look old compared to some of the things that are coming out. So do you feel like it’s going to continue at that pace for a while? Or do you think it’s going to slow down? What’s your general take on that?

Eva Dong
00:07:39 – 00:08:19

Yeah, I think, well, maybe I wish. I wish it can keep growing rapidly as an as rapid as it is now. Yes. I agree. There are so many AI startup and each startup is doing things in their own way, along with the, like the more leading player, like OpenAI, Anthropic, Cohere. I think there’s like and then and then then you have the big auction nation. Like Microsoft is investing so much and in trying to catch up Amazon, not only they have partnership with like Anthropic and Coherent, they’re also developing their own AI model inside Amazon. So this is really like the landscape in, at least in the United States is very good.

Eva Dong
00:08:19 – 00:08:33

You have small players, medium player, large player. Everybody’s marching. My, my secret wish is like, I hope can, it can, things can keep going like this. I don’t know how long we can keep growing so fast, but I, the longer we last, the more advanced we get.

Filipe Santos
00:08:34 – 00:08:54

It’s amazing. I mean, it’s, like, it’s blown my mind. Even though I’m really comfortable with change, it just feels incredible. And I I agree with you. It would be really nice to get this to the point where AI is really our friend, and we can use it for the things that we really don’t wanna do. And it feels seamless, and we get it right the first time. So yeah. That’s cheers to that.

Filipe Santos
00:08:54 – 00:09:10

With that, let let’s let’s talk a little bit more about the integration of AI’s impact and its I I guess, its overall efficiency the the efficiency and the effectiveness in marketing campaigns from your perspective. Like, where do you see that going? Like, how do you how do you feel that disintegration is happening?

Eva Dong
00:09:10 – 00:09:53

Yeah. I think every company can be a little bit different, but my basic biggest concern on this is not really on the efficiency side. I feel like as long as you use the tools and use the technology correctly, it will increase the efficiency for sure. And we see a, like, lot of good examples from the market. I remember there was, like, one CEO talking about how AI replaced 90% of his customer service workforce. First of all, let’s not comment on whether that was an impact on human’s job. Just looking at the technology itself, it seems like a really good efficiency lift. However, my biggest concern is that the implementation part, because I there’s a stats I read.

Eva Dong
00:09:54 – 00:10:27

On average, the marketing tech stack is only used 33% of the time. And then 2024 is the 2nd year for the for the decline. So 2023, we’re already declining on that number, and 2024, we keep declining on that or we only reach 33%. That means each tool is only at is doing what it’s doing at 33% of the time. That’s just too little. You know? Like, that that means, like, we we only use AI 2 day per week. That’s just too little. That’s not supposed to it’s doesn’t really gave the investment the return is supposed to.

Eva Dong
00:10:27 – 00:11:14

So like my thinking is like, yes, we all want the tool and we all want a technology for its effectiveness. But first we really have to establish a use case, like know what you need, not just for the buzz and for the hype. 2 is established a guideline of using AI. You want to use AI responsibly and then coach all your employees on that. And 3, really adapt your strategy while using AI, because maybe when you are manually creating image, your strategy is a. While you’re using mid journey to create that image, the the strategy is a little bit different. So that like all this marketing strategy need to be adapted accordingly. And last but not least in order to reach that effectiveness, never forget the human judgment.

Eva Dong
00:11:14 – 00:11:25

So no matter how good and how like highly do you think of AI always looping human human judgment, judgment, always have human to review it before it go public. Perfect, segue into an aside. So you

Filipe Santos
00:11:25 – 00:12:02

mentioned that the employer that replaced, like, 90% of their customers, support staff or what whatnot. Now I’ve heard stories in the media, which I guess this could be pretty controversial, but they’ve there were, like, hacks where people were asking the AI to give it a discount of a dollar for, like, a truck or something ridiculous like that. And because there’s no human, because we’re we’re, like, relying on the effectiveness and the efficiency of AI alone without human intervention, Don’t you feel like do you feel like that’s a temporary thing while AI gets better? Or do you think that that’s something that is this is the this is the reason why we shouldn’t be scared of the technology?

Eva Dong
00:12:03 – 00:12:39

Yeah. That is a issue. Like, I definitely heard a similar news on that too. I think that comes to something like again, like, the human judgment should be in loop. I think a lot of company just like buy a chat box service and then use the chat box service and just say, everything is automatic. Instead of that, I think maybe a better way to use it is you have chat box copilot for your agent. So there’s a human there, but instead of him typing everything, maybe he has everything ready, generated by the AI, and then he can adjust on it or decide to send it or decide not to send it. I think that’s a better way to solve that problem.

Eva Dong
00:12:39 – 00:13:22

But at the end of the day, I think that just one of the drawbacks of the current LIM models for now. Because like when I was going through like trainings on LIM, there was like one class we specific do is called prompt engineering. And one part of that is, like, how do you surprise the LOM and make do something is not supposed to do? So that’s literally a training we go through to just find what kind of, like, weakness if there is an LOM. And obviously, since that’s too early, So there are weakness in LLM, which I believe a will be improving in the future, but well, nothing is perfect. No technology is a 100% perfect. So I think there will always be something like that. That’s why, like, human judgment is always should be always in the loop.

Filipe Santos
00:13:23 – 00:13:58

Love that. Love that. And I I I I really agree with that. I think we really should be using these technologies to get better and to be able to do what we are awesome at rather than just doing the tedious stuff because we’ve done it and because we’re we’re familiar with it being part of a role. And I think that’s where the the transformational shift is happening. And that’s where, you know, machines, if they can do it in split split seconds or, you know, much faster than we can, then it’s worth giving that job to them and, you know, taking on something a little bit better for us. What are the gen AI applications in customer online shopping? What what are your what’s your perspective on that?

Eva Dong
00:13:58 – 00:14:55

I think there’s several popular applications I have seen on the industry right now. So if we start from the beginning of the journey, it’s more like the chat box for conversion part. So that means if someone come to your website or someone to come to your app, how do you help them to purchase something and purchase something they will be satisfied with. And the second part will be after they complete their purchase journey, how do you do the customer service? So here, there’s a lot more applications. So for first one will be the chat box for customer care. Let’s say, if you’re not happy, you want to return, you can talk to the chat box automatically. The second one is that I think which is a better application which we just talked about is the agent copilot so instead of just automatically process their return, you can have an agent talk you to them, but the agent will have the AI generated messaging already type up, So that speed up and increase efficiency by a lot. And, another one will be like the knowledge based research.

Eva Dong
00:14:56 – 00:15:43

So instead of AI automatically generating the messaging for the agent, the agent can have, like, a chat gpt type of tool. So maybe the customer ask, okay, I bought this online. I wanna return it in store. Is that okay? Then the agent can have a GPT like thing, like type it in search and give get an answer. So this way, the agent has the most accurate answer per brand policy, which is always changing. So that is the 3rd way. The last way will be probably the easiest and probably has the longest history is like a post conversation insights. So instead of AI doing anything in the real time, AI can also summarize the conversation after the conversation ended and then to draw any insights or next step action or or for internal analysis purpose.

Filipe Santos
00:15:44 – 00:16:37

Yeah. You know, it’s it’s funny that you mentioned that. So I I do recall Amazon being one of those companies that uses this effectively. I’ve seen the AI being used as an initial, you know, I guess, returns agent where you’re returning an item and it’s it’s all being done by the AI initially, unless there’s something that requires a human agent. And then I’ve also noticed the summarization of product benefits and features, which I think is super helpful. And and when in particular you have an issue with a product that you need to know more about, if AI can dredge out all of those little use cases and make it so clear that you don’t have to wait for someone to to answer your question, but rather it can dive through all the data that’s available for the product and, you know, provide it whether it’s comes from video, text, text, content or whatnot. So I love that example. That is really that’s that’s pertinent to what we’re we’re seeing now, but, like, can only get better, I imagine.

Eva Dong
00:16:37 – 00:16:37

Yep. Yep.

Filipe Santos
00:16:38 – 00:16:54

Alright. So let let’s let’s talk a little bit more about then how AI powered chatbots or or, I guess, shop bots in some cases personalize the, customer interactions and enhance this user experience? I know you’ve mentioned some aspects, but is there any more that you would want to, like, elaborate on?

Eva Dong
00:16:55 – 00:17:47

I think first party data is really powerful in this. I mean, sometimes you guys say, oh, I’m a new brand. I don’t have much person like, personal data. But you you have more than you thought. So first of all, you usually have the product details. So regardless is the customer just come here to purchase something or does the customer already purchase something and want the customer care for it? You at least know the product detail. And based on that product detail, you can do a certain type of personalization for them. And if the customer has a history with your brand or with your website and that’s even better then you have the customer history like what have him or her purchased in the in the past have they talked to the customer care before have they used the chat box for conversion before maybe you know also know their personal information like which state do you live do they live in? What what is your demographics? So all of that information is really valuable.

Eva Dong
00:17:47 – 00:18:18

And just a little bit data can move the personalization so much because you see a customer come to buy, like, a male footwear. But if you know he he’s a she’s a mom, I’m buying for her son, that’s very different than you talk to a man buying a shoe for himself. You know, like the perspective are very, very different and the messaging should be very different. What they care is very different. So as long as you know a little bit, that can go a long way. So use all the first party data as much as you can.

Filipe Santos
00:18:19 – 00:18:48

Amazing. I I was gonna throw a pro tip in there too. Like, if you’re if you’re whether you’re on b to b or b to c, this might differ, but you could end up using manufacturer data to to help you kind of get faster along with your products and and the information that you may not think you have. And you might also be able to use competitor competitive information that’s been sanitized. That way, I I guess you could use that as your base of knowledge for your chat bots or your your shop bots. It can make things a lot better from the first party experience.

Eva Dong
00:18:49 – 00:18:50

Yeah. I love that.

Filipe Santos
00:18:51 – 00:19:08

Awesome. Let’s switch into a little bit of, I guess, the case study side. Like, have you are there any notable case studies or success stories that you’ve seen that highlight the effectiveness of these chatbots being used? I mean, we mentioned Amazon, but are there any other ones that, help increase conversion rates?

Eva Dong
00:19:08 – 00:19:40

There is one retailer I have in mind. We’ve done it online. So they have a chat box on their website for purchase, which I think chat box for customer care is much more common. There’s a lot of companies have that on their website, on, on their app. I think chat box for conversion is like new and is like more emerging. So my experience is like that one can 1, it can process, like, long text. So for example, there can be a customer because I’m looking for something I can really, like, type a lot something a lot. Hello.

Eva Dong
00:19:40 – 00:20:03

I’m looking for a, like, a cross country ski jacket. I’m a beginner, but I’m going to Utah for the first time, and I wanna try it. But I don’t want it to be too expensive. Because my first time, I don’t really know whether I wanna do this for a long time. Blah blah blah. The customer can keep talk on and on. In in this actually, right now, we have a lot of time audio enabled, so the customer can talk really long. And now the LOM need to go in the end and extract all the useful information.

Eva Dong
00:20:04 – 00:21:00

So this person’s going to Utah, which matters. This person’s doing cross country ski, which is very different from downhill ski, so that need to be recorded. This person is a beginner. That means they don’t want the most advanced jacket on the earth. They probably just need something they can use and it’s good quality. And then probably timing matters like, do I need this urgently? Do I need like fast shopping, like fast shipping or do I need just normal speed of shipping? So the LAMs will take all the keywords and then it will go to the product catalog and match everything it has in the product catalog and drag something And here, there’s the more the art because this chat box is supposed to mimic what a human shop in store experience. So if you imagine you go to a sports store and you you said everything I just said, the the sales agent is not gonna give you 10 different jacket and then tell you to choose 1 from them. The sales project probably give you, like, not 1, but 2 or 3 and for you choose from.

Eva Dong
00:21:00 – 00:21:34

And then probably tell you, oh, product a is like this, pros and cons, product b, pros and cons. This is probably the best experience instead of you can’t give too little. You you also can’t give too many. So chatbot should return maybe, like, 2 products. I’ll say, we have this from Patagonia, and we have this from ArtRx. ArtRx is a little bit more expensive, but it has an advantage on a and b. And then the customer say, oh, it’s a this is a little bit expensive. How do you have something like around $200? And then the the algorithm could go back to the product catalog and fetch something cheaper, but also fit in whatever the keyword we just extracted.

Eva Dong
00:21:35 – 00:22:03

So I think that one is a good example I have seen because one is very lifelike. It’s really mimicking your experience in a for indoor shopping experience. 2 is it doesn’t do too much. It doesn’t do too little. Because as you imagine, if you check with a chat box and give you one choice, this is the only product you have. Or if I give you 10 jackets, like, excuse me. I just told you I’m a beginner. Like, don’t give me 10 products for me to choose from.

Eva Dong
00:22:03 – 00:22:20

Like, what’s the point of talking to you? So that that is a good one, Absolutely. And also the back and forth, right? Like you might say, oh, this is too expensive. Or you might say, oh, I don’t like yellow. Okay. No, nothing yellow. So, like, the the back and forth and they keep going, I I think, like, is is is very magical.

Filipe Santos
00:22:21 – 00:22:39

I love that. And, especially with the amount of shopping I do. Yeah. It’s helpful to be able to whittle down just right to the products that you want. So, yeah, that that’s a great example. I appreciate that. I’m really excited about this section in particular. Just it’s, it’s near and dear to my heart.

Filipe Santos
00:22:39 – 00:22:53

So let’s talk about let’s get into this a little bit more. What do you think the future landscape is on Google, especially with them kinda jumping on the bandwagon with AI and trying to make themselves more relevant for today’s market?

Eva Dong
00:22:53 – 00:23:58

To how I see it, I think Google will be divided Google search will be divided into, like, 5 parts. And there’s 3 parts we’re already very familiar with as, like, digital marketers. We have the SEO, we have SEM, and then we have ShopIt, which is like the traditional three parts we always care about on Google. And then I’m not putting YouTube here because it’s kind of separate, and then we can talk about in a in a bit. And then the 2 new added and and I think the 2 new added and will be significant parts is 1 SGE, which is search engine experiences, and the second one is Google perspectives. For those people who are not super familiar with Google and Google’s new product, I’ll just explain a little bit so SGE Google search experiences that is something similar to what you see like on the very very top So let’s say if you ask a question, the Google can generate a block of text to answer your question. This way you don’t have to click into any blue links below. So that is the way of, like, Google is trying to move to more generative AI search experience.

Eva Dong
00:23:58 – 00:24:23

Because, historically, you Google something, yes, it give you, like, a 100 links. You still have to click on them and read on it. Maybe it doesn’t fit you. You come back. You click on another link. But with SGE, you should be able to get what you need on the very top. And this message, instead of just grabbing from one website, it should be a summarization or generated new text from what it can see on the websites.

Filipe Santos
00:24:24 – 00:24:24

So

Eva Dong
00:24:24 – 00:25:03

that’s SDE. And the second part is Google perspectives. I think this one is even more new and even less scaled. So maybe you don’t even see it on Google search right now. Maybe some people see it because they happen to lock you into the test group. So that is the part Google trying to bring to counter the SGE part. Because if you think about the SGE part, that’s all AI, AI grabbing this, AI grabbing that. And sometimes we know AI can be biased maybe it’s grabbing certain things BIM is grabbing not everything maybe the question you ask has multiple different answers and the ST only give you like 2 answers out of the the 10 answers.

Eva Dong
00:25:03 – 00:25:34

So perspective is another thing to counter it. So that part is we’ll be grabbing channels from, like, YouTube, Reddit, will be a very important one. It can be grabbing from x or all kind of social media on what people’s perspective on it. So that way is much more human. First of all, it’s human generated opinions instead of AI generated opinion is is like a good way to bring that human back to the search experience. So that’s the 5 parts I see in the future. So a CEO SEM shopping, as we already know, and then SGEM perspective as the newly added part.

Filipe Santos
00:25:34 – 00:26:25

I love that. You know, I’ve been reading a lot about perspectives, like, and also, like, the opinions of humans that are going through this, and a lot are are actually SEOs practicing SGE experience. Not always do they give proper credit to the the to the information that they’re delivering. And then on the other side, Reddit, while while a great resource, may not be ideal for every single result either. So I’m really curious to see how that pans out over time. But, like, right now, it makes sense. Right? That’s just they’re trying to, as you mentioned, counter automated ex experiences provided by AI and and that of the human side of things like we’ve traditionally experienced in the search engine. But with that, I I guess I would love to know a little bit from your experience and also from what you’re seeing what the impact might be.

Eva Dong
00:26:25 – 00:27:06

I think SGE will be very impactful. So several early trends we have seen because this Google is already testing it and slowly rolling it out in certain countries. The first thing that will be the decline in website traffic and ad revenue for Google too. So because you generated that text on the very top, the purpose is to reduce people’s time and energy to click on 10 different links, but that also means if I’m a brand website, I get less clicks into my website. So that can be very sum. So like historically, I have like 10,000,000 traffic to my website every day. From time to time, I might decline to 8,000,000. I might decline to 5,000,000.

Eva Dong
00:27:06 – 00:27:53

I like as a brand, I could be very concerned about it. It’s like, why then why are people not coming? And then same for Google because people are not clicking on that. Google also get less ad revenue from me. So I have less cost for a brand. I have less cost and less traffic. Is that a good thing is that a bad thing like I think people need to like rethink about it however I think there is a good part is that for the people who click into your website they probably really really want to buy or are really really interested in whatever you have on your website because they probably see the generative AI text. Let’s say I was researching, okay, what is the best restaurant in Rome? I wanna experience some, like, authentic pasta. And then the Google generated research and say, okay, I have 10 restaurant based on all kinds of Google reviews.

Eva Dong
00:27:53 – 00:28:22

Those are the 10 most best pasta restaurants. And then maybe I click into one of them. I look in their into their website, their story. Maybe I could go to their booking system. Then there, you know, like, I really want to convert. So the I think for a brand is you have less traffic, but the traffic you get should be much higher quality. Your conversion rate should be much higher. So instead of a traditional ecommerce, maybe 3% to 5% conversion rate, it could be 10%, 15%.

Eva Dong
00:28:22 – 00:29:05

So good or bad, like, that that that’s kind of a trade off. Also, I think, certainly, I think it’s a better user experience because sometimes I also feel like SEO obviously is the best thing of ours. Google, thanks for Google. I get information all day long for free. But also, I just I’m kinda tired from going from links, coming back, going to another link, coming back. And so if I can have everything on the on the top of obviously, that is a better experience. And I think last but not least, I don’t think every single industry will be impacted by this equally. There will be certain industries I can imagine more like b to c direct consumer e commerce might be impacted more than other industries.

Eva Dong
00:29:06 – 00:29:30

For example, if I’m doing if I’m a pharmaceutical company or I’m a life life science company, I don’t think people would just just, like, you know, read whatever AI generated on the very top for 3 sentence. And, and that’s it like, no, like my buyers, my enterprise buyers will still come to my website. Look at all the research, look at all the patterns I have. And then the top to me, so probably certain industry won’t be impacted that much.

Filipe Santos
00:29:30 – 00:30:13

Yeah. I love how you mentioned your money or your life concepts, like, you know, the fact that certain industries are just, you have to do due diligence and you have to dive into a little bit more. People want more information. They want clarity. They want, like, you know, a 180 or a 360 degree coverage of something. So they might go deeper or they might want more information. But on on the flip side, as you mentioned, like, it’s inevitable that behavior is going to change, and you did bring up the the branding side of things being impacted potentially. I’m curious to know from your perspective as well have do you do you think this is gonna be equal on the branded and non branded side? Or do you think there’s a little bit more of an impact to brand versus, you know, unbranded or vice versa? Yeah.

Filipe Santos
00:30:13 – 00:30:19

Specific specifically from the organic search side, like, do you think that that’s gonna hit 1 or the other more?

Eva Dong
00:30:19 – 00:30:44

That’s a very good question. I think maybe it will hurt non brand more, but this is but it’s a very interesting question. It’s my first time thinking about it. So I mean, just talking out loud. I think non brand might be hit more because if it’s brand keywords, I say, I’m going to REI, then I’m going to REI. Okay? So I’m good. So maybe I type in Google just because I don’t wanna type wwwrei.com. I’m still I’m going to REI.

Eva Dong
00:30:44 – 00:31:14

I’m going to search on REI what I wanna wanna do. So regardless what you give for me on the top, it doesn’t matter. And I think is that if I’m just searching, like, for a really generic one. Oh, I’m just looking for, sneakers for basketball. That probably can generate more. And, also, I don’t know because SGE will decline ad revenue for Google. I’m sure Google will come up new ways to make up for the ad revenue. So I’m not sure what it is.

Eva Dong
00:31:14 – 00:31:27

Maybe for the s g part on the very top, Google will also think about how to have sponsored ad and how to have the organic part. So if that is the case, non brand will be even even impacted even more.

Filipe Santos
00:31:28 – 00:32:02

Yeah. I imagine folks are gonna, you know, formulate their own opinions, but I I tend to agree with you because given the way that SG has been producing the the results, it seems to be hitting, you know, top of the funnel type content more and maybe a little bit on the of the middle. And that’s where I think SEO has focused a lot of their attention over the years, producing enough of a footprint across informational topics, but those are the easiest also to address given these LLMs and the responses to the data that exists in the search engine. So, yeah, I I love that perspective. Some fodder for the the group.

Eva Dong
00:32:03 – 00:32:15

Yeah. Yeah. For sure. Yeah. Like, I I think SEO is is like, I don’t know. It’s easier and more difficult nowadays. Easy because it’s so easy to generate blocks. You can just actually add to general 15 blocks for me at this second.

Eva Dong
00:32:15 – 00:32:27

And I, at the same time, because everybody can do that on at scale, then it’s difficult for me specifically as a brand. Like, how do I make sure my SEO stand out? Yeah. That that’s that’s exactly like you said.

Filipe Santos
00:32:28 – 00:32:58

Yeah. And the and the beauty well, I guess this is what we’re contending with right now. Yes. You can produce a ton of information, a ton of content, blog articles really quickly. But I I also know that there’s a quality aspect and the quality is not so great. And I think even in even as it gets better, it still will have to rely on new perspectives and and human writing. So I’m curious to see how people grapple with that and become better at gauging quality and producing quality given scale. But, well, you know, we can always talk about that at a later time.

Filipe Santos
00:32:58 – 00:32:59

I think that’s a bigger topic.

Eva Dong
00:33:00 – 00:33:01

Yes. Yes. For sure.

Filipe Santos
00:33:02 – 00:33:12

Well, given the material we’ve covered, I’d I’d love to get a little bit of predictive stance on what you see in the industry for, like, the next 6 months. I would love you to take a stab at what you think might happen.

Eva Dong
00:33:13 – 00:34:00

We talk a lot about Google. Like, personally, I follow Google very closely. And then something big that come to Chrome this year, and that will impact not only Google, but also pretty much the entire landscape of digital marketing is third party cookie. So Google and along with all the other web browsing web browsers has been reducing the focus on 3rd party cookie and enhance user privacy. So the plan Chrome give it to the public is that by the end of 2024, Google will remove all the third party cookies. The biggest influence on that will be retargeting, which is as digital marketers, we do that all the time and it’s super effective. So like, it’s always the best running campaign. Like the no brainer campaign would always, always do.

Eva Dong
00:34:00 – 00:34:45

And in the future, that’ll be much more difficult because we wouldn’t be able to know, oh, this client, he came to this website, was interested in our vacuum, then he come to our website. Obviously we should send him an email about this vacuum how good it is so third party cookie will change our landscape like so big But I don’t think that’s the end of the world. Right? Like, Google has changes. Google also have solutions. So I just like, in order to protect our user privacy more, we won’t have that third party cookie. However, we also have new technology to take care of that. So I know Google have something called privacy sandbox, which is something like will, like, encrypt the data. So you can do something similar to retargeting as we always do, going through that sandbox.

Eva Dong
00:34:45 – 00:35:42

So you wouldn’t know the you wouldn’t violate the user privacy, but still get something similar return as your typical retargeting. And also I think contextual display might be coming back. So, you know, like, in programming word or in the display word, we usually use a lot of 3rd party cookie to run those displays, run the display ads. And then contextual programming or contextual dis display is one of the things I think people usually do, but don’t do it at huge, huge scale. And I think where’s the 3rd party cookie being removed, so this part can become a new popular. So for the people who are not super familiar with contextual programming, that is not looking at your previous behavior, but it’s only looking at your current behavior on the current website. So for example, I’m looking at a website selling vacuums. It might show me, tell tell tell tell me about, like, what’s the best vacuum from all the major players.

Eva Dong
00:35:42 – 00:36:05

And that ad can just show on the right hand side on the website. So it doesn’t need to know who I am, what I did, what what I’m interested in. As long as I’m looking at a vacuum block now, I know I’m interested in vacuum now, So there’s a ad on the right hand side for me now. So I think that as a new old old classical concept may be coming back into the future with more dominance.

Filipe Santos
00:36:06 – 00:36:38

Wow. Okay. That that’s I’m I’m gonna be looking at vacuum cleaners in a whole different way from now on. But yeah. Oh, I I feel I feel personally smarter having spoken to you and having had you on the show today. I wanted to thank you, Yves, so much for your time and for kind of answering all these questions. I know we went in many different directions, but supplied so many good perspectives on this. I wanted to kinda give you the the opportunity to let everybody know how they can get in touch with you or follow you.

Filipe Santos
00:36:38 – 00:36:39

It’s the stage is all yours.

Eva Dong
00:36:40 – 00:36:59

Yeah. Yeah. First of all, thank you for inviting me as really happy to do this podcast with you. So for the listeners, if you wanna know more about me, I’m very active on LinkedIn. If you just search Eva Dong. I’m I’m McKinsey. If there if there’s multiple Eva Dong. And also, I run a newsletter on called Smart AI Marketing.

Eva Dong
00:36:59 – 00:37:34

I public weekly articles on how to implement AI in marketing. What’s the newest trend in AI and how we implement it with marketing. I know there’s a lot of things going on in the marketing world and in the AI world. So I really try to bring this to development together and try to give practitioners a little bit advice. And I really welcome like ideas and comments. And hopefully we can this can be a place for us to exchange ideas, learn from each other. It just my way of like getting my thoughts out. But I, of course I love to be here, like what other practitioners think about AI and marketing at the same time.

Filipe Santos
00:37:35 – 00:37:46

Excellent. Well, I know I’m gonna be following you and and seeing what’s coming out. Everybody, I suggest that you do as well because, you know, Eva’s been, like, on top of all this and, obviously, it was a fascinating conversation. Thank you, Eva.

Eva Dong
00:37:47 – 00:37:49

Thank you for having me.

Filipe Santos
00:37:51 – 00:38:12

Thank you for tuning in to phraseology plus ai. We hope that you enjoyed today’s episode as much as we enjoyed creating it. Don’t forget to sign up for our email list at phraseological.us. Give us a raving 5 star review anywhere you get your podcasts. Remember, stay curious, stay informed, and keep listening for more. Until next time, take care and happy listening.