How Google’s Conversational Commerce Is Changing Everything — Lucas Tieleman | Why AI Changes Search, How Conversational Commerce Works, What Product Feeds Hide, How AI Landing Pages Convert, Why Bad Data Fails (#485)

In this episode, we dive into the shift happening in online search and how e-commerce brands can stay visible as search engines become more conversational. 

Lucas Tieleman, Co-Founder of Optiversal, shares how his platform helps brands improve visibility across AI search and shopping tools through better product data. 

He also reveals strategies for optimizing product feeds, using AI to automatically generate high-performing use-case landing pages, and getting ready for holiday shopping spikes. 

Topics discussed in this episode:  

• How conversational AI is changing search.   
• What gaps make brands invisible to AI.
• How natural search queries differ now. 
• What attributes to add to product feeds.
• How editorial content builds search context.
• What an automated data audit reveals.
• How AI builds high-volume landing pages.
• Why click rates drop while impressions rise.
• What rising ad cost per click signals.
• Why structured data wins holiday sales. 

Links & Resources

Website: https://www.optiversal.com/
LinkedIn: https://www.linkedin.com/in/michele-nieberding/

Get access to more free resources by visiting the show notes at https://tinyurl.com/hb2matcy

______________________________________________________

LOVE THE SHOW? HERE ARE THE NEXT STEPS!

Follow the podcast to get every bonus episode. Tap follow now and don’t miss out!   

Rate & Review: Help others discover the show by rating the show on Apple Podcasts at https://tinyurl.com/ecb-apple-podcasts   

Join our Free Newsletter: https://newsletter.ecommercecoffeebreak.com/   

Support The Show On Patreon: https://www.patreon.com/EcommerceCoffeeBreak   

Partner with us: https://ecommercecoffeebreak.com/partner-with-us/

Transcript

00:00:00:05 - 00:00:23:02
Unknown
SEO is just evolving. The thing that's, you know, evolving is the search piece of it. And as a result of that, the optimization is evolving. There's like a new acronym every 24 hours or so. Why are many e-commerce brands becoming invisible right now? Yeah. So one big thing that's happening is that.

00:00:23:04 - 00:00:47:20
Unknown
Hello, and welcome to another episode of the E-commerce Coffee Break podcast. A lot of changing right now in online search. Google is moving beyond simple keyword search and becoming more conversational. Shoppers are now using tools like Google AI overview and to discover products, compare options, and make business decisions in natural language. But while search behavior is changing fast, many e-commerce brands are still relying on outdated product catalogs.

00:00:47:21 - 00:01:17:19
Unknown
Static product pages that AI systems struggle to understand. So today, we're tackling a big question how do e-commerce brands stay visible in the age of conversational e-commerce? To break this down, I'm joined by Lucas Tieleman. He is the co-founder of Optiversal. helping e-commerce brands improve visibility across Google, AI, search and shopping platforms through better product data. Before launching optical, Lucas was chief product officer at Bazaar Voice and has years of experience building and scaling commerce technology businesses.

00:01:17:20 - 00:01:45:11
Unknown
So let's get started. Lukas, great to have you on the show. Great to be here. Class. Let's dive into Google search. Just changing fast. Why are many e-commerce brands becoming invisible right now? Yes, one big thing that's happening is that people are starting to trust search engines more and more because of the proliferation of AI tools. And so consumers are sort of getting used to not asking a computer dumb questions anymore.

00:01:45:15 - 00:02:11:03
Unknown
So they're using more and more words to describe what they are looking for. And the engines are getting really good at answering those questions, but they are better at answering those questions if they have all sorts of data. Right. And so what we see is there sort of a gap between how consumers are searching for products and how e-commerce brands talk about products.

00:02:11:04 - 00:02:34:16
Unknown
And so natural language attributes, use case based searching is all sort of new upcoming. And there's a gap. Now we talk about conversational conversion. So some of our listeners might not know what that is. Can you give an overview what conversational commerce is in simple words? Let's say I'm going on a camping trip right with my kids here.

00:02:34:16 - 00:02:59:21
Unknown
It's almost summer here. Or it is. It's pretty warm here in Austin, Texas, so we only have a few days left to go camping before it gets too hot. So let's say I'm going on a camping trip and I need to buy a bunch of camping gear. In the past, I might have searched for tent and maybe tent for sleeping with two kids.

00:02:59:23 - 00:03:21:08
Unknown
But now, you know, with all these AI tools that are available, I can actually extend that query and make it, you know, all sorts more relevant. So I can say I'm going on a camping trip in the Hill country in Texas with my two daughters who are seven and five, and I'm going this weekend where there's going to be some bugs and it's going to be pretty warm.

00:03:21:08 - 00:03:48:01
Unknown
What's a tent that has, you know, good, good mosquito nets and great airflow? And so that, you know, query is so much different than I'm looking for a tent, you know. And so the way that consumers and then the AI system is going to respond with a bunch of things, and then I'm going to look at them and say like, well, actually this looks like a good option and I don't really want to do and talk about or think about this or it's going to ask me about price range.

00:03:48:01 - 00:04:10:15
Unknown
And so now I'm having a conversation with, with the search engine versus me just hitting the keyword, looking at ten links, trying to figure it out, going back and forth, back and forth and then landing on a purchase decision. It was a great overview of how complex things has become. And right in the past, you were searching for a blue tent and you were optimizing for keywords.

00:04:10:15 - 00:04:32:13
Unknown
And that was, in a nutshell, SEO. But that has changed now when it comes to the product data. Obviously, you can't put all this information that you're searching for into the product data. What need brands to do to be still visible in search? Well, there's a few places where you can actually can give a lot of context and give a lot of data, right.

00:04:32:15 - 00:04:59:09
Unknown
So typically what we look at is like there's your product feeds that you send to advertising platform. So every e-commerce brand has a product feed and they'll send them to e-commerce platforms. And on average you can fill out about 65 to 70 attributes in that feed per product, right. So that's actually like a decent amount of data. What we see is that people fill out about 10 to 15.

00:04:59:10 - 00:05:22:08
Unknown
So that's like first very big gap. And then what we also see is that people really focus on factual data. So is the t shirt black. But the subjective data that you know that's grounded in natural search oftentimes get missed. So is this a comfortable t shirt that can you where you know is a sweat resistant. Who knows.

00:05:22:09 - 00:05:41:08
Unknown
Right. There's all of these things that are, you know, more natural language based that you can put into your data feed. That's their space for it. And most of you know, brands are not optimizing for it. So that's number one in your advertising feeds you can add a bunch of data. Then secondarily you can add data to your site.

00:05:41:10 - 00:06:12:21
Unknown
Right. So you can describe your product in a different way. You can add editorial content to your site. You can generate buying guides. You can generate landing pages, all of which that can all interlink to different products, all of which helps a search engine get gain context on, you know, what is the use case of this product? How would you search for this product, etc. now how does AI understand actually if it's a use case or if it's a product wipe?

00:06:12:21 - 00:06:39:10
Unknown
Or is it just a blog post about something? How does it really pick the data that it needs to serve? The one who's searching, who knows? And it's probabilistic models, right? So like search historically has been very deterministic. Right. You look for it's like vector based. You look for certain attributes and you present results. Whereas large language models search is more probabilistic.

00:06:39:10 - 00:06:55:00
Unknown
And so, you know, it's trying to find what you are looking for. And if you use more complicated words, the result is going to be different. So you are probably also going to get very different results than I am, even if we would write the same exact same sentence because it has context on you and as context on me.

00:06:55:01 - 00:07:25:16
Unknown
And so but what what we know is important is that there's high quality, structured data that is valuable for consumers and gives context about the product. And so those are sort of the the most important things to make sure that your product content, your site sort of gets a check mark on. Now, if I'm a brand and I have my product catalog and my data is not as structured, obviously you're helping with getting this right.

00:07:25:18 - 00:07:45:03
Unknown
What's the process to to work your way through that? At the end of the day, you will be searchable in Google in AI. Yes. So we typically start off with an audit where we look at your product data and determine, you know, how good or bad it is, where it is. Where is it very complete. Where does it lack complete list?

00:07:45:04 - 00:08:10:24
Unknown
Where does it you know, where is it unclear. What is it you know, have clarity. So we start with an audit to identify the gaps. And then we try to get the gaps as complete as possible. Right. You can very hardly get to 100% because there's always things you can't really figure out for all products. But you can get very close.

00:08:11:01 - 00:08:34:16
Unknown
And so sort of the mark that we hit for is 8,590% of data completeness. And so that's on the completeness side. And then there's on the helpfulness of clarity side. So the way you write the description we might want to, you know, rewrite the description etc.. And then you can use that data and you can either use it on your website or you can use it in your advertising feeds.

00:08:34:18 - 00:09:01:20
Unknown
And what you see is better performance, which is nice. You get more people to your site and they convert higher because they have more context. Now, for our listeners, probably a lot of them are used to the classic Sao. Now this is a very different approach. Can you tell me the difference between old school SEO and what you're doing?

00:09:01:21 - 00:09:28:05
Unknown
I don't think there is a difference, to be honest. So I think, you know, SEO is just evolving, right? Search engine optimization is involving the, the the thing that's, you know, evolving is the search piece of it. And as a result of that, the optimization is evolving. But, you know, there's like a new acronym every 24 hours or so and it's hard to keep up with them.

00:09:28:07 - 00:09:58:14
Unknown
But at the end of the day, it's just good search optimization. In fact, like Google published a thought piece a couple days back where they said like, hey, there's there's nothing special about AOE or Geo. It's just like good search optimization. You should build helpful content for consumers. That gives them context on the products you sell. And if you do that and you deliver it to us in a very structured and clean way, your chances of getting picked up and results increase.

00:09:58:16 - 00:10:25:20
Unknown
So if you give helpful, unique content and a unique perspective on a product for a consumer, that will bode well for you. So it's changing for sure. But it's not like this is a totally new thing. It's just search optimization. And so having structured data, giving helpful content, having a unique perspective and helping consumers, you know, choose which products to buy, I think is is most helpful.

00:10:25:22 - 00:10:52:18
Unknown
I think that was always the credo of Google. Helpful content, helpful results and sort of their circling back to that, where were they coming from now with Optimus obviously helping brands getting better? Can you give some examples and you need to tell the brand on how it helped your clients and what kind of results they saw. So we work with a wide variety of clients, mostly in the sort of the enterprise space.

00:10:52:20 - 00:11:16:03
Unknown
We do two things for them, so we help them with optimizing their product content, and then we also help them build landing pages that are that answer consumers questions. So or are very specified to a use case. A good example of that is like what are Pans that chefs use is a landing page we have for a home goods retailer that does very well.

00:11:16:05 - 00:11:36:14
Unknown
So we've created a landing page with pans, chef shoes and AI came up with the keyword, and I came up with the landing page and built the landing page. And now, you know, that retailer gets a bunch of traffic from from that page. And so and similarly, we can then take the chef's data and put it into the product catalog.

00:11:36:14 - 00:12:05:07
Unknown
And that works too. And so in the case of that retailer I believe there they're sort of feed their advertising performance is up somewhere around 12%. So they get 12% increase in conversions for the same ads band. And we drive our, you know, AI landing pages drive approximately 16% of their site traffic. And so that is pretty substantial because it's all kind of incremental.

00:12:05:09 - 00:12:30:20
Unknown
I mean, that's that's really substantial. Now, you said you mentioned that you're using AI to create landing pages. What kind of homework needs emergent or brand to do before they come to you? Because obviously you need to have something that you can build on. How does that work? Very little homework actually. It's a full circle system. So the way that we structure it is we do keyword research automatically based on the product catalog.

00:12:30:20 - 00:12:49:19
Unknown
So we look at the product catalog and we basically ask the questions, what are all the ways that consumers could search for this product. And then we go to the search engines and we validate whether that's a way that consumers actually search for the product. So we validate, hey to people on a monthly basis, search this way for this set of products.

00:12:49:19 - 00:13:07:07
Unknown
And if there's validation and the retailer or the brand does not show up in the search results, that means there's an opportunity to create a landing page. And so then the system creates a landing page, and then we'll see if it performs and if it performs will keep it around. And if it doesn't perform we'll toss it out.

00:13:07:09 - 00:13:32:23
Unknown
But the what happens is it's like a full loop of keyword research validation that there's volume page creation and then performance analytics and optimization. And so it goes around and around. And because we use AI we can build thousands of these. I think we're coming up to a million landing pages. So which is like, you know, totally awesome.

00:13:32:24 - 00:13:53:22
Unknown
And so you can create a lot not all of them are going to work, but that's totally awesome too. So, you know, you get rid of the ones that don't work and you double down on the ones that do work. Now, for our listeners that are probably just getting started with this, how can they check if Google understands their product right now or understands it correctly?

00:13:53:23 - 00:14:15:21
Unknown
Are there ways to check that? For sure. You can call me in that or us and we can help you with an audit of that. I did, but there's other signals, right? So one of the things to look at is you can log in to your Google Search Console, and you can, for example, look at the overall search over your at your overall search trends.

00:14:15:21 - 00:14:43:18
Unknown
And if Google is not really understanding you in the age of AI search, what you'll see is that your impressions are actually up because people are searching more. But what you'll probably see is your click rates going down year on year. And so if that's the case, there's probably a swath of keywords that you used to be found for, but that have now turned into different keywords that you're no longer found for.

00:14:43:18 - 00:15:10:17
Unknown
And there's an opportunity for you to get after it. The other signal is if you log into your Google Ads account and you see your cost per click go up, that means that you're bidding on a narrow set of keywords where there's high competition. And keyword volumes in general are exploding, right? So Google had the most queries ever last last quarter in the history of the company, most keywords ever in a non-holiday quarter.

00:15:10:18 - 00:15:33:21
Unknown
And so if your cost per click is up, that means that there's a whole world of keywords out there that you're not going after. So those are two signals. If you look at those two signals and you see that your cost per click is going up and your organic clicks are going down, it means like probably you're not being found as much as you used to, and it's probably because of your optimization strategy.

00:15:33:22 - 00:15:59:20
Unknown
Yeah, I think it's a very timely matter because, I mean, the smart businesses, they are starting to optimize or have started optimizing for that quite some time ago. And if you're not on that bandwagon, you definitely will feel it in your business. Now what's the typical onboarding process? What steps are involved? Yeah, if you work with us like for the on the on the feed side or the product data side, the onboarding is typically a week or two.

00:15:59:21 - 00:16:20:20
Unknown
So it can be really quick because we can take your existing feeds, do the audits, find the gaps optimizing them. And then once you're going we will continue to optimize, so to speak. If we're building landing pages for you, that's a slightly more involved process. And typically we have that up and running in like 4 to 6 weeks or so.

00:16:20:22 - 00:16:47:15
Unknown
And then with everything search engine related, it takes time, right. So it's not there's no magic bullets. So it takes time. But it is compounding. So you won't see results like the day after you turn it on in most cases. Right. It takes some time to build up the results because people the engines now need to pick you up and people now need to be able to find you, etc., etc..

00:16:47:15 - 00:17:09:17
Unknown
And so it's a it's a compounding asset. So we currently are working very hard with our clients on like holiday season, which is crazy, but that's like what we're starting to prepare for. Like how are we going to show up for Black Friday, Cyber Monday, Christmas shopping? Okay, now you already gave some examples of customers using your system.

00:17:09:17 - 00:17:39:00
Unknown
You said it's bigger companies, but who's your perfect customer? What kind of verticals industries do you work most with? Verticals. Doesn't really matter where you have successful businesses and successful clients in almost all verticals. So we don't. We're vertical agnostic. It applies for most verticals. And you need to you will see the most benefit if you have at least 250 SKUs, because variety matters, right?

00:17:39:01 - 00:18:02:07
Unknown
If you have five SKUs, I mean, you can just use ChatGPT to optimize your product data. You can just take your Google Catalog, put it, put it into ChatGPT or Gemini or Claude and say, optimize this data, make sure it's all filled out, make sure it's optimized for natural language search, and it will do a great job. The trick is like when the data sets gets bigger to do it well.

00:18:02:08 - 00:18:25:12
Unknown
So at least 250 products and at least, you know, $50 million in gross merchandizing value. So you need to have revenues of at least 50 million bucks. But if you have those two things, more than 250 products and a decent amount of sales, then we're a great partner for you. Okay. How does your pricing structure work? How do you charge for your services?

00:18:25:13 - 00:18:45:05
Unknown
The pricing generally works that it's a great ROI. So we price based on skew volumes. We have lower tier and higher tier. Okay. Makes sense. I think there's a lot to learn in getting it right with the product data right now. I think it's a it's a race. And I think the one who is the fastest will win the race.

00:18:45:05 - 00:19:05:14
Unknown
And so that's definitely interesting times that's on there. Look for our coffee break comes to an end today. Is there anything you want to share with our listeners that we haven't covered yet? I don't think so. Don't sleep on it. Sure. Engines are changing very rapidly. Yesterday was Google I o or I don't know when this will be broadcast, but when we're recording it.

00:19:05:15 - 00:19:31:18
Unknown
Yesterday it was I o and they announced like a whole slew of new features. You know, AI is getting deeply embedded into the search experience on Google where there's like a billion shopping queries per day. Chat GPT is still, you know, growing in terms of volume. Also on the shopping side. And so there's like a lot more people are searching for or not even a lot more people.

00:19:31:19 - 00:19:55:04
Unknown
People are searching very differently today than they used to be. And if you do not have the right structured data, the chances of you're going to be found are like just very low, because you're asking the models to do too many gymnastics. And if your competitor just livers very clean helpful data, they're going to win. So don't sleep on it.

00:19:55:05 - 00:20:16:11
Unknown
Don't wait for it. You know holidays coming up. So you want to get ready, you know pretty early to rebuild the benefits. Yeah that's that's a perfect summary of what's happening right now in the market. Where can people go and find out more about you opt. Cool. Very straightforward. I will put the link in the show notes and you just want to click away.

00:20:16:13 - 00:20:36:00
Unknown
Lukas, thanks so much for giving an overview of what's happening in the search market. I think it's amazing times what's happening there. I'm in this business for a very, very long time and there's always something new. But I think right now the changes are faster than they were ever before. So for all listeners, check it out. Get in touch with Lucas and get your product data right.

00:20:36:01 - 00:20:37:23
Unknown
Thanks so much, Lukas. Thank you much.