Welcome back everyone to the e-commerce marketing podcast. I am Arlen Robinson your host and today we have a very special guest that is going to be with us today, Yang Han. Yang is the co-founder and chief technology at Stack Adapt which is an award-winning advertising technology platform Stack Adapt is one of the most Innovative and fast growing companies in Canada, and so we all like to warmly welcome Yang to the podcast today.
How's it going in I'm doing great? How are you? I'm doing? Well. Well. I appreciate you coming on the podcast that we're very excited to have you here, and why don't you tell me a little bit? Little bit more about uh stack adapt, and what it is that you guys do how you got started in a little bit more about your background prior to joining the company.
Yeah, sounds good, so we started stack adapts four years ago, so what we do is we are an advertising platform. We work with Brands and advertising agencies, or across the globe we are essentially a platform that targets users across the web, so if you run an advertising campaign. You want to show your targeting the right user we're integrated with over 35,000 websites and apps all across the globe and we're it's done through a system.
What's called real-time bidding, so every time a user goes to a website a real-time auction occurs and we submit a bid for this user to try to show an ad to the right user so let's say for example if for is one of our clients then they run a campaign with us we want. And sure we're targeting users online that are interested in cars or buying a car, and we may look at the historical data.
We do not read about cars before or we are not they purchased a car, so we utilize a lot of that data in sure you're showing the ads the right people online okay, so it's really just kind of smart smart targeted marketing and understanding who the the audience is yeah, and we deal with a lot of volume of data because you know we.
All the traffic of over 35,000 site so our system actually handles requests around 200,000 times every single second we have data on over a billion users online wow yeah, that's that's a lot of data a lot of data that to manage and be able to uh to mine through their interesting and now you one of the founders of the company.
Correct, yeah, so I'm one of the co-founders. We have two other co-founders of Italian okay. Yeah, we started around four years ago. You know we at first when we started we didn't really have an exact idea of what we wanted to do they came from an advertising background. I built a lot of different software before I started a couple other companies my background is more building Financial software and the real-time trading environment, okay, so I also handled a lot of volume of data back in that area.
But advertising is so much similar, but also a bit different as well. Okay, so we combined that expertise and we looked at you know. What are the problems out there right now. You know there's more and more data online that isn't being leveraged and our focus is mostly on Native advertising as well, which is like the type of unity typically see on Facebook.
Four years ago that was just starting out, but it wasn't prominent yet right right so if you wanted to make sure those type of ads that were more engaging you know we're Distributing more everywhere across. You know all the different sites gotcha gotcha. Yeah, that definitely makes sense yeah, I definitely see where you guys were kind of thinking when you founded the company and as you saw where everything is really headed and with with marketing digital marketing in general and that kind of brings me to my initial question.
We hear a lot about AI of these days artificial intelligence apply to a lot of different aspects of technology, so now how really does a benefit? You know digital marketing efforts specifically yeah, that's a good question well digital marketing is fairly broad you can apply a in a lot of different areas when it comes to digital marketing one of the most common ways that most ad platforms utilize.
You know machine learning and AI. Is for example optimizing clicks so making sure that you're getting as many clicks to your as possible you can go a step deeper and try to learn in order to make sure that people are actually spending time on your site, so then you want to optimize the time on site spent on your site, and then finally deeper the final you want to you know use to protect conversions.
You actually make sure people after they see the ad. And spending time on your site are actually making the purchase you know order signing up because at the end of the day why do we you know why do we use online advertising? We want to get an Roi from it right? We want to make sure we're actually getting the right users onto our site that want to engage with us and want to use our product right right.
Yeah, yeah, that makes sense. You know the our audience of our active listeners of the podcasts are primarily. You know eCommerce businesses, and I know everyone out there. That's running an e-commerce website. You mentioned a couple key things there as far as predicting. You know the conversions.
How exactly does that apply like what what exactly would that do for a business running an e-commerce website and the fact that they can predict a certain conversion. What does that knowledge given the ability to do? Yeah? Well, let's start from the basics, so you know let's say there isn't any prediction there you have.
Over a billion users on the internet right and you're running an ad campaign, and you're trying to Target people in line. There's so many people out there. Who do you pick right traditionally marketers? What they did is they would select their targeting manually so what they would do is. They would say I want people you know who are.
Let's say they've read about certain keywords on a site or let's say they're in the certain location or I want to spend more money in this time of day. It's great, but that's all done manually right right, but you have a lot of different data points. You can look at now everything you do online is tracked.
When you go to a site when you purchase something that data is being stored somewhere and that data is being funneled to a certain system and at that form that can get ahold of as much data as possible will be able to utilize all these different pretty much what we call features into the system in order in order for us to figure out patterns right because that's what does it runs.
You know statistical models. Across Mass amounts of data in order to find relationships from there so the more data. It has on the actions you have online the better automate decision making that it can make essentially we want to connect and understand what type of users. Typically go to your site or view your product.
You know so typically for example Facebook has what's called the look like audience right? That's something. We're building as well, so you put this code on your page Facebook will then read about the types of users that go to your site? They will look at your social media profile look at other sites.
They've been to so every time like a Facebook like button is on the site. They actually know you went to that site. So you know you're being tracked even outside of Facebook. If you combine all these data points together the users that have been to your site. We can see what size they've been to you what the media profile is like then Facebook will take this data, and they'll come up with a model, and they'll Target similar users all across the internet that describe yourself.
Yeah, yeah, that's that's very powerful. You know. We've done a little bit of. Advertising through Facebook internally here. I know a lot of our audience members have as well, and you know familiar with the look like look like audience and so like you said when you open that up and apply that to you know to everyone online and all of the different areas not only the social media, but everywhere the in customer can possibly be at it's yeah, it's really a tremendous opportunity to hone in on on those potential customers in areas where you may not necessarily have have.
About now, I know when you were talking a little bit about you know what you guys do and part of what you just said has to do with these tailored and the kind of targeted marketing efforts, and I've recently heard more and more about this term, but what specifically is or what it specifically are smart ads actually and how did they?
How do they benefit in e-commerce business I wouldn't have everyday business utilize a smart ad right. That's a good question. Well Smart in my view obviously they have to be highly targeted like I mentioned. They can't just Target anyone online. They have to use the right. They have to use your site's data, and they have to use have access to the data that comes into the system to be able to predict that you're targeting the right person smart ads on another component that we haven't talked.
About is also how the message is delivered if you think about most ads you know people come up with a fixed copy you know a fixed headline a fixed image. These are static right and you're showing the same message to everybody online, but if you think about it. Everyone is a bit different Dave had different experiences online you know and so if you could tailor that message specifically to that user that's a much more powerful message so one thing.
Do at stack adapt as we have a capability what's called Dynamic retargeting, so we take what a user has done on your site for example we work with Airlines, and we will know if a certain user browsed a flight ticket from let's say Toronto to New York right then when we show an ad to them we will stand say hey are you flying from New York?
We have a new discount now. At you know 80 dollars round-trip. Do you want to take this right now this very smart because in you exactly where you did before and now I can take advantage of that right yeah, that's yeah, that's that's extremely powerful and like you said that's really kind of taking it to the next level on what you mentioned initially was the retargeting which I know most of us are familiar with like if you go to a site.
You know let's say you go to Expedia and all of a sudden you go to see it in and you start seeing Expedia ads everywhere for different. Things but a lot of the retargeting systems these days is more. Just general retargeting of that website that you previously were at it's not as targeted as what you said you know for instance like what you said, I recently put the trip on Expedia, and uh you know from Orlando to Chicago, and so what you're saying is that type of targeting would basically show me like let's say, I went to seeing in the next day after booking the trip.
It would say you know. Or maybe I didn't book the trip, but I was that was my search parameters of the particular trip when I checked in I made the ad that says you know flights as low as you know $200 on such-and-such website discounted site, and it would be for Expedia whatever other site is in their in their network.
Is that kind of the gist of it. Yeah exactly, and it works really well because like like you said people are used to the regular retarding already. It's been done. Over the place you know you go to a site, and you expect it to follow you right, and it works for sure right, but people get bothered by this as well because it's so plain.
You know it's not intelligent enough to really benefit you yeah, and also like a lot of the current platforms like Facebook have supported this too, but it's actually manual. It's not really smart you have to actually manually upload all the combinations and the. Onto Facebook for example so what we do we take it a step deeper is we you don't have to set up anything at all utilize what the user is actually doing on your site or specific signals that they've done.
Okay, which is good, but obviously you know. There's a lot of progression in the industry of we can see getting even smart as well. We're just scratching the surface. The potential here. Yeah, the potential is definitely huge. It's just really wide open, and you know one of the things that I was I was thinking because there's so much out there.
There's so many ways and you guys are really deep into it as far as gathering this data. Using all of these analytics to to allow businesses to make smarter decisions with regards to marketing and honing in on certain audiences. You know let's say we have a fairly new e-commerce. Uh business they you know they're new to marketing.
They haven't done a whole lot of marketing efforts. What do you think they should start as far as analytics because you know like you said it's there's so much out there. It's really kind of hard to know what to do, but I know the main thing is if they starting to get a fair amount of traffic initially.
What do they really need to focus on with regards to analytics? What what do you recommend obviously you want to start with the most basic type of analytics, which is like a Google analytics that can basically tell you you know the people that I've been to your site. You can segment it. You can run ad campaigns into your different performance time on site and conversions however in my opinion Google analytics is just very basic vanilla that you want.
It's a bit flawed because it's completely cookie-based. Or more people now are the same person as using multiple devices. They have their own. They have their iPad. They have browsers. They clear their cookies once in a while. They get more sophisticated like that so I recommend using a specialized analytics platform, and there's a lot of options out there.
I'm not an expert in the differentials of them, so I won't give specific names, but use across device analytics provider. They can give you better insights on the same user typically with. Analytics they may also assign the wrong conversion value to the wrong platform because it's cookie base and because let's say I saw a product through an ad I clicked on the I wanted to add, but later on you know.
I went away, but I decided to go back there. I just type it in Google, and then I back to that site well. Google's last attribution is going to say well, actually the search causes you know conversion to happen, and it doesn't give enough. Any weight essentially to the original ads so also use a conversion tracking provider that is more sophisticated and can break down the different ad platforms that went through the whole that the user went through the whole funnel, so you understand where you're getting a value from right right yeah, that makes sense as far as what you said.
I know you know company would want. Started with the basics and of course I can mention at Google analytics is a great place to start, but you are a little limited with the fact that it's primarily cookie-based, and that's really what the challenge is and you know maybe you can lend some insight into it.
I know a lot of companies that are doing at oolitic. There is a challenge because like you said it's it's a it's a new day these days because you have people with all these devices whereas you know. Five ten years ago people had you know before there was such an explosion of mobile devices in the internet people just had a home computer in a work computer, and that's really it and so it was fairly easy to track someone's online activity and what they're doing because it's usually between two devices and primarily.
Their home device you know a lot of times at businesses because of work networks and restrictions. You know people aren't doing a whole lot of personal things on their work computer, so it's just you know one computer. You know typically per household, so there's a lot easier, but you know these days you've got people with a you know cell phone.
They have tablets they have laptops, and you know multiple devices, so it really does become very tricky to really track someone's activity, and so I I really am. ESS to the direction you know the whole tracking is going to go like what types of tools they're going to need to use or what types of ways are we going to be able to pin down a user across all these devices you know because you have these networks as well and people's I appease are going to switch, so I'm really curious where it's going to go.
Do you guys have any insight as to what are going to be some key ways of tracking people. Yeah, well, it's obviously a huge problem and a lot of companies. Are coming together to try to solve this problem obviously the best data for cross-device on the user is deterministic based on logins now Google and Facebook have you know this data more than anyone else for the rest of us.
We have to leverage a lot of the the more data. You have the better like for us. We predict this to the ad Network system like I mentioned we have two hundred thousand requests coming in every second so we have continuous data on users. We would do more statistical based based on this history so we have actually a universal ID and we have the history of each IP each cookie right we would use machine learning to look at you know behaviors.
We've seen before and then we would say is it likely that this IP is connected to this mobile app device ID if you're likely the same person based on machine based on a lot of Statistics right obviously. There is margin for error for that, but the more data. You have the more accurate you're going to be and then you can mix that in with Partnerships with companies that have login data, but it's very fragmented there.
Yeah, that is it is and it's it is a challenge, but like you said The more data the better and then having these relationships with a lot of these ad Partners your you're able to. Have different algorithms like you said where it can make a prediction based on somebody's activity and based on their IP may not be 100% certain.
You know it can definitely help really you know pin down what you know who that particular person is to make to make sure these advertisers are really getting them the most kind of bang for their for their book yeah, and it's definitely any more accurate than not using at all if you're just using cookies you know.
Getting lesson as accurate Apple has completely you know changed their stance with cookies on mobile devices. It's hard to track their now so yeah these methods were starting to see a lot more improvements in his industry in making an accurate so it's definitely getting there. Yes. Yeah, definitely for sure well.
Um you know before we close things out here. I always ask one question of our guest on the podcast here, and that is what do you think is the one thing that an e-commerce business can do right? Now you know to help their business grow get traffic and get sales the question. Well. I think it really depends on what stage you're at.
I am gonna. You know cheat a bit and say it's not one thing, but if you don't have anything at all start with the most simplest type of advertising, which is you know retargeting or dynamic retargeting? If that works well, and you have a positive Roi move on to the look like targeting from Facebook and Google the intent-based targeting based.
Their search data based on product. No history if that works. Well, then you can go one level up two more Leslie targeted. You know methods, and if that if even the basic Dynamic retargeting doesn't work, then the probably that advertising probably isn't the right time right now you probably have to change the positioning you know of your product or your site in how its presented on marketed right because you have to differentiate you know.
How you sell and what your product is to really make it be able to scale in the first place right once you have something great. Then it's time to scale sure sure definitely. I think what you said up front is key. It really does depend on the stage that you're in with your business. You know because things that a start-up would do is definitely going to vary from you know business sets you know three and five years into it so that really is it really is key really have to see.
Really, just not a one-size-fits-all when you're thinking about marketing. It's uh you really got to look at a lot of different factors, and that that is really the main thing well. I really appreciate you joining us today on the Commerce marketing podcast. It was definitely a pleasure if any of our audience listeners our members are interested in getting in touch with you.
How do they go about doing that? Yeah, they can send me an email at Yang at stacked after thanks a lot. It was great being on here in talkin about this. Okay great likewise. Uh you take care, and we appreciate you joining us sounds good. Thank you for listening to the e-commerce marketing podcast.
To access Ecommerce videos and other resources to help your business grow please visit get forward slash videos subscribe to us on iTunes by searching for e-commerce marketing podcast and please leave a rating and a review. Thanks for listening. See you next time.
Podcast Guest Info
Co-Founder, Stack Adapt