#156: Enhance Your Website’s Conversion Rates by Tracking Visitor Patterns

Please support the show if you liked today's episode:

• If you love the podcast, please get someone else to listen, too!
• If you enjoyed this episode of the Ecommerce Coffee Break podcast, please head over to Apple Podcasts, leave a rating, write a review, and subscribe.
• Share the podcast with your family, friends, and co-workers.
• Tag the podcast on Instagram @clauslauter and let me know what you like about it.
• If you like the content and would like to support the podcast, you can buy me a coffee here.
• Become a guest on the show or sponsor an episode.

Claus Lauter: Hello and welcome to another episode of the E-Commerce Coffee Break. We wanna talk today about conversion rates and we want to talk about visitor behavior. Now, most of you have used any kind of paid ads platform like Meta Facebook or Google Ads in the past, and you might be well, , aware of any kind of audiences that you have that might be local like audiences that might be custom audience, any kind of audience that you use.

For instance, so retarget, , your visitor and then hopefully getting them back to buy from you. So this is in general the topic we'll talk about today. And with me, I have Tayib, he is the chief marketing officer of enhancer. He has been working there for five years, has started as an intern in 2017, and has been an integral part of the company's growth journey.

He degree in statistics and a master's degree in data science. His expertise and [00:02:00] dedication has been crucial to the success of the company, and he's proud to be part of Enhancer Vision of for the Future. So enhancer the name already tells it, but Enhancing your conversion will help us there. To see what we can find out on how to get more conversions.

So let's say hello to Tayib. How are you today?

Mutakabbir Ahmed Tayib: I'm fine, thank you and thanks for having me in the podcast. You're welcome.

Claus Lauter: Tell me a little bit of your journey into the world of Meta Ads, Facebook ads, Google Ads.

Mutakabbir Ahmed Tayib: as a company, we actually started as a project-based company back in like 2015 before I was joined.

After that, , we were doing some projects for, let's say big companies, insurance companies, , automobile companies. each of these projects were like data science based, but they were kind of dying out in the whole data science awakening era. The reason is, people started to do this in-house and there are a lot of tools that actually helping them do these kind of projects internally.

So as a dying industry, , we actually thought how can we take our [00:03:00] expertise and make it to a broader audience and scale it so that we can. Globally. So that was our purpose. after that we started to see that , in Turkey, e-commerce, there's a lot of gap between the, Audience optimization and campaign optimization.

And there is a lack of knowledge as well. So we thought our expertise of data science, we can integrate it into the e-commerce industry and see how we can actually, , benefit the whole industry and the companies. So that's how we actually started the whole e-commerce and enhance thing.

Before it wasn't even called enhancer.

Claus Lauter: if we talk about meta. Everyone knows the story with iOS updates since about like two years or so, things have begun a little more tricky. In the past, it was quite easy to have just creating audiences, local audiences, and then the third party cookies fell away, and now it's becoming more and more, , difficult to find right audiences.

You found a solution to help with. Tell me a little bit what the original problem is, why audiences were not that easy to find anymore, [00:04:00] and what you have developed to help with better targeting.

Mutakabbir Ahmed Tayib: So there is actually two ways to look into the situation. One is with the iOS restriction, one is without the restriction.

So let me first explain without the restriction, because it's much more simpler to explain. So what happens is in e-commerce companies, people actually try to advertise their product to their audiences. Like if it has a certain amount of traffic, it tries to retarget these a. That come to their website or application or let's say Instagram page.

So what happens is Facebook generally provides two or three basic audiences, like all website visitors, product views, add to cart, and previous purchases. in these three scenarios, what happens is if you go to your all website visitors, There are a lot of visitors that doesn't end up buying from your website because they're not simply interested in the product.

Some of them might have come accidentally from some links [00:05:00] on the outsource, and some of them are like, they find the price to be not suitable for their taste, And in these kind of scenarios, when you launch your campaign to , all website visitors, you will be wasting a lot of money on the people that actually doesn't buy anything.

Another issue is that in these Facebook and Google ads, , platforms, they charge you based on the reach and the impression you make. the broader the rich, let's say you are reaching 1000 people, it'll charge you differently if you reach 5,000 people. Regardless of the relevancy of the reach, because Facebook says like I showed 5,000 people, your ads five times give me this much money.

Doesn't matter if this 5,000 is relevant or not. So what happens in Product Viewer is just because the person actually looked at a product doesn't mean they're going to buy it from your website. They might find some competitive price elsewhere. They might find the comment section, they leave review to let's say, not positive.

So a lot of factors, into the decision of buying a product from the website and at Dukart is like, it's so small that Facebook [00:06:00] cannot even optimize it when it cannot optimize it, it kind of stucks into the learning, , limited. So that's where the problem lies. This is without the restriction.

with the iOS restriction, there is another step of difficulty because you are missing out on the, , people that come from iOS devices. That's actually crucial for a business because, , you are already going to miss out a large portion of your audience. Also, you will not be reaching your full potential.

If you're reaching the all website business. So that's where, , we have come up with the solution and hassle. So do I go into, , full technical details or just like high level explanation? That would be my first

Claus Lauter: question. We should start with a high level one, then we can go a little bit deeper into the technical.

So let, let me ask a question. That was a very good introduction on where the problem lies, and I think that gives a good overview for listener to understand that what meta Facebook and Google Ads offer out of the box might be suboptimal for the [00:07:00] potential that they can have in reaching the right customer.

if we go a little bit deeper into this, and now we're talking about creating the audiences and the right audiences, the ones that are really most likely to convert, how do you help with.

Mutakabbir Ahmed Tayib: this is where like we thought of the enhancer as a solution for these companies. So what actually happens is when a person comes to your website, let's say they looked into a certain product, looked into a certain product category, looked into a certain page, or let's say, went down to the bottom of the page to the reviews.

So these are all some behaviors of the visitors that come to the website. What Anesa does is take those behaviors and analyze what kind of behaviors in the website leads to a purchase what kind of behavior doesn't lead to a purchase. it's like primarily based on behavior analysis using artificial intelligence, , like algorithms.

in this scenario, when a person comes to your website and doesn't behave, , that interested. He or she will be [00:08:00] excluded from the enhancer audience. If another person comes to the website and they, looked at a certain t-shirt category product, but did not look at, jeans, pants product, so that person might have a higher score of purchasing something from the T-shirt category.

that gets included in the audience that you will be targeted using the campaign. This is all, , behavior based. So let me give another, , example using the demographic situation. Facebook also keeps tracks of the people who actually spend online and they're like profiles and everything. We don't know that, but Facebook definitely know it.

if you are, let's say, reach and you go into a website that sells computer, mouse, You have the capability of purchasing all 10,000 of them, but you don't actually need them. , at the same time as let's say high school guy who actually needs a mouse will go to the website. Their behavior will actually dictates whether that person is going to buy or not, whether, , they're interested in their product or not.

So this is where the behavior is actually more important than the demographic [00:09:00] itself. what Ansa does, if even that rich person comes to the website and doesn't actually show the behavior that leads to a purchase of the computer mouse, then he or she would not be included in the target audience at the end of the day.

so

Claus Lauter: I see some similarities through a good salesperson in a brick and bolter store. They probably will be able to tell if somebody comes into the shop and how they behave, how they walk around in the store, if they're likely to buy something or not.

And I think that's where you're going with the ai. , with the behavior. You can find the most valuable visitors, and then obviously you attend all your attention basically to these specific ones, and you don't spend money on the ones that probably will not buy it from you anyway. How long does the, self-learning time take before the AI kicks in?

Mutakabbir Ahmed Tayib: ,

so changes based on the traffic of the website itself. handicap of AI algorithm needs, like, it needs data. the bare minimum we actually accept is like 30,000 page views per month. When you have 30,000 page views per [00:10:00] month, the algorithm is able to learn it, let's say within two weeks of time and then send those audiences into your Facebook and Google ad, managers where you can use them in your campaigns.

So if your traffic is much more, it can learn within two days events. So it's completely depend on what that traffic. , let's say you have 30,000 page views, but after three months it went to a hundred thousand and like more than a hundred thousand page views, the algorithm doesn't stop learning.

So it keeps learning each like Updates, , let's say audience segments each week. , the reason is, a seasonal, , holiday came up and the trend in the industry changed. The behavior inside the website changed due to some kind of coupon or anything. And new product might have, , attracted a lot of interest.

due to this trend and seasonal changes, we decided to implement them weekly update. In the last couple of days, or let's say seven days, the website traffic had changed, the behavior had changed, the algorithm will [00:11:00] acknowledge that and it'll actually update the segments. As a result, it'll update the campaign, because it's using the n s R audience in its, target audience section.

[00:12:00]

Claus Lauter: so you said it's not just, , the plain atkar visitor or page visitor. You create segments that , can be much more in detail , on what it, , contains. Can you gimme, just an example on how a transparent, segment can look like? ,

Mutakabbir Ahmed Tayib: We provide these behaviors and the individual segments transparently.

, if you log into our product in the application itself, there is a dashboard that contains, let's say, 50 segments for your, whole website. So out of this 50 segments, there will be 50 different behavior, let's say The first set represents the behavior for the first segment, which is like, okay, , they have looked into a product.

Category under, let's say tshirt. They haven't looked at a product category under glasses. If a person comes back, let's say after seven days to your website, that is also another behavior that person came back to your website. I'm just giving three example from top of my head, but there are more than 200 [00:13:00] behavior that we actually look into before deciding what goes into a segment or whatnot.

This behavior will be like given to you in a transparent way. , and then when you select those segments and send it to Facebook, Facebook knows which person to target and which person to avoid. . But these behavior will actually also help you understand what works in the website.

So if you have a website structure that doesn't work, this kind of behavior actually highlights those issues as well.

Claus Lauter: Okay. No, that's a very important point and I think that's a very valuable, , result that you can basically go back to your website and start then doing conversion rate optimization.

Let's talk a little bit of about results. What are kind of average results that you see with your clients?

Mutakabbir Ahmed Tayib: we have been actually working with more than 250, , companies around the world. So the bare minimum we have seen is will return on ad spending go at least above double.

So in some cases we have seen five times , in some cases we have seen 10 times, but we have never seen, a written on ad spending that goes [00:14:00] below. In very rare cases where the price of the product is so low that even if you spend a lot of money on it, you won't get the return out of it because of the price of the product.

In general cases, we have seen the ROAS to go up like at least double or three times Okay.

Claus Lauter: from the technical implementation, understand, , you link into WooCommerce, you link into Shopify. Give me an overview of what kind of integrations you offer , for different platforms.

Mutakabbir Ahmed Tayib: if, , Shopify Martin actually wants to.

Take benefit of this service, they can simply go to the marketplace and search enhancer and it'll pop up. Once they install the app, the integration itself is done automatically. they don't have to do anything. have to do only two things. The first is integrate Google conversion.

I. Copy it from the Google Ads account and then paste it in enhancer, and the audience will be sent to, , Google as in that regard. Another, , manual, integration they have to do is. Create a custom [00:15:00] audience in those Facebook or Google app platforms that is linked to enhancer. And this is just a one time job.

So after that, , enhancer knows I have to send these audiences to this particular custom audience. So it'll be, , real time link. Let's see. And if you're not from Shopify or Commerce, , in that case, our developer team actually does the integration. For you. We generally ask for Google Tag Manager access, , and to our email addresses.

And then our team does the integration within 15 minutes or so. And then you're good to go. The data starts to flow. , the pixel is also integrated in those scenarios. Google conversion tag is also done in those 15 minutes. So it's, , the onboardings itself is like 15 minutes.

Claus Lauter: Okay. You said that you need to create data before the algorithm really can kick in, and it depends on how much, obviously you're driving traffic to your platform.

Now, on the upper side side, how much, , budget do you need to calculate from [00:16:00] then retargeting , and spending on meter to really make the most art of. .

Mutakabbir Ahmed Tayib: So it actually depends on the industry itself, to be honest. What happens is, if you are launching a campaign with a lot of budget, and if you have spare budget, let's say you have a hundred dollars, you wanna spend it, but your reach is, let's say only thousand people.

That only takes $20 to reach. So what Facebook will do is reach those people five times to spend the budget. So we don't say that you have to spend thousand dollars, you have to spend $500. What we say is if you used to spend a hundred dollars, you can keep spending the a hundred dollars. But since your, , reach has changed and the people are actually interested in your product, you won't need to show the ads five times.

You might have to show the ads to two or three times because they're already interested. We are pointing out those people for you. . , and in that case, the budget is kind of utilized better on the [00:17:00] target audience you are going to change. So in that way, we are not actually increasing or lowering your budget, but we are kind of optimizing the budget so it gets spent on the right people only.

Okay.

Claus Lauter: you're coming from the data science, , side of things, and AI is retaking over huge parts of the e-commerce side. What's your prediction? I wanna know, for this year, 2023, , how much will ai, , take over the parts or make the life easier for a e-commerce store? .

Mutakabbir Ahmed Tayib: So I have actually seen this question pop up a lot recently due to the Chad G P T explosion that we like lived.

Two weeks back. So my best guess for this year and onwards is that none of this AI will actually replace the human because I myself had masters in the data science. I have already worked in this kind of algorithms. So what it will help is, , for e-commerce industry, let's talk about this. , so for.

Performance [00:18:00] marketing experts, they're worried that their work will be taken over by ai, let's say, at creatives done by ai. , the targeting itself is also done by ai, but to be honest, it's not going to replace them. It's only going to make their work much, faster. So if there's a tedious work, so in this case, let me give an example for the target audience selection.

You can do this manually in Google Ads as well, but it's going to take a long time for you to arrange these behaviors, analyze them, and then create an audience for you. So it might take hours even days. So in that case, we're not actually, , taking away your whole perspective of the target audience.

We're just making it easier for you to access and target, , your campaign with. For the creatives part as well. It's not going to replace the, , content creators, but it's going to help them brainstorm and get ideas more faster and see if first, let's say [00:19:00] draft of the work they're going to produce.

And then this AI can create the first, say, skeleton, for these kind of scenarios. They can build on the skeletons using human. Then perfect it, but do it, , in, let's say five minutes where you could do it , in previous, let's say one hour. So that's the biggest impact we will see in terms of AI and the like industry takeover that everyone is scared about.

if you are a performance marketing expert or graphics designer, you don't have to worry. You can make your work like much faster with these kind of tools, , because it's never going to take over the human creativity.

Claus Lauter: Now I'm a hundred percent you. I totally agree what you just said. , AI tools, and I'm using them in my own business, they just help making things much, much faster.

You still have to have , the final check on it if you're happy with it. And I totally agree, the human factor will not go away. Now, where can people find out more about enhancer? And tell me also a little bit on the pricing. S.

Mutakabbir Ahmed Tayib: you can look up enhancer like in the website called [00:20:00] ww.enhancer.com, or you can go to Shopify Marketplace and Search Enhancer.

You'll be able to see the application there. And from there you can go to the website as well if you want to get in touch with us, you can go to our website and book a demo with us, and our team will get back to you with, addressing your questions and answers. And for the pricing structure, , we have a subscription based model, , it's like you can cancel it anytime if you're not happy with the product.

, the pricing structure is based on monthly page fee. The reason is, after the integration is done, the announcer algorithm collects those data in a real time, analyzes them in in real time, and sends them to Facebook and Google ads in real time. So this whole process is based on number of page views you have.

So that's what actually drives our server cost, to be honest. So that's why our pricing structure is based on monthly page view. So if you're coming from Shopify, if I'm not mistaken, , up to 25,000 page view, the pricing is $99. So [00:21:00] it goes up according to the page view. I don't know the whole breakdown of the pricing structure, but the highest one is probably At the moment we have three pricing structures, small, medium, and large. So if you are, , coming from Shopify, you will find four because Shopify adds another package. But on the website pricing, it's like three pricing packages. More or less it's the same. So it's based on monthly page view.

if your page view goes up after six months, then we will notify you that your page view is increased. That's a good thing actually, because it means you're gaining a lot of traffic and sales. It's a bittersweet moment. Your cost will go up, but your profit will also go.

So if they want to find more, , about enhancer, your audiences can actually search, enhance and Google as well. , also, they can go to www.nl.com.

Claus Lauter: I will put the links in the show notes and then just one click away. And then I would recommend that people just get in contact you who are a demo o when I saw on your website there's also a video which explains , the process and the benefits of using enhance.

Hey, thanks so much for diving into the world of [00:22:00] audiences and, , getting the right person to the store. , like the content, I think people have to listen twice to the episode because there's so much con good content in there. Thanks so much your time and have a great day.

Mutakabbir Ahmed Tayib: Have a good day. Bye.

[00:23:00]