A Technology That Predicts What Your Customers Will Buy Next with Shanif Dhanani

Andy Splichal 0:02

Welcome to the Make Each Click Count podcast. This is your host, Andy Splichal. We are happy to welcome this week's guest to discuss today's topic which is A Technology That Predicts What Your Customers Will Buy Next. Today's guest is the co-founder and CEO of Apteo. Apteo helps ecommerce brands personalize their marketing campaigns by predicting what their customers will buy next. A big hello to Shanif Dhanani. Hi, Shanif.

Shanif Dhanani 1:16

And it's great to be here. Thank you so much for having me. It's a real pleasure.

Andy Splichal 1:19

Yeah, well, we're excited to have this topic. So let's get right into it. Now, what do you mean by predicting what a customer is going to buy next? How do you do it? And how can a company use that knowledge to grow their business?

Shanif Dhanani 1:33

Yeah, you know, it's always great to have a crystal ball, right. And so if we had that we'd be doing a lot of other things. But essentially, you know, when we talk about predicting what a customer is going to do next, really what we're trying to understand is what are the products that customers are likely to buy next each individual customer, or are they just not likely to come back, or they're just not likely to buy your product. And so brand marketers who sort of know this information can really tailor their marketing and customize our marketing. And so that's what we mean, when we're talking about, you know, predicting what a customer is going to buy. Next, just understanding the likelihood of what's the likelihood that they're going to buy product A versus product B or, versus no product at all. And the way we work at this is, you know, we use, you know, a lot of people call it AI. My background is in data science. So I call it data science, basically use some data science to kind of understand and find patterns across a brand's customers and their products, and then use those patterns to forecast what each person is likely to do next. And then, you know, group those group, those people by the products that they're likely to buy, basically just comes down to using a lot of math and statistics to find patterns and buying behavior to figure out what people are likely to do next.

Andy Splichal 2:49

So you can use your software, if you're a company, and you plug that into your software, and then it's going to split out your past buyer into different segments. Is that how it works?

Shanif Dhanani 3:01

Yep, that's, that's exactly right. So you might have a segment of people who are likely to buy, you know, a red dress, you might have a segment of the people who are likely to buy white shoes might also people who are, you know, a segment of people who are at risk of never buying anything again. And so that's exactly what our software does, takes your past buyers and spits them out into segments of people based on what they're likely to do next.

Andy Splichal 3:22

And then how do you go about reaching those different segments of customers? Is it through email? I assume it's email. But is it email exclusively? Or is there how are you doing it besides email?

Shanif Dhanani 3:34

Yeah, we like to plug into the tools that our customers are using. So I'm sure you've probably seen email has kind of been the the foundational channel but SMS has gotten really big and obviously paid social. A lot of folks argue that that has grown sort of the world of Ecommerce. And so we plug into all of these tools, essentially, what happens as you can, you know, once you've got the segments that our software creates, you can plug in your existing, let's say marketing tools, like an email provider or Facebook ads, and then we'll send the segments over to those tools. Now, if you've got a marketing team, you can then go in and use those segments your sells those audiences to either find customers who are likely to buy products, you can even use look alike audiences with our segments to create new customer segments. Or you can have our team do it for you. You know, we've got a lot of experience sort of using our audiences and our segments to drive new sales. So we try to get out of the way we try to help you kind of identify your audiences, send them over to your tools and get out of your way if you if you don't want our help or help you out if you do.

Andy Splichal 4:37

So if your team they specialize in using those with Facebook ads or you said Facebook ads, but is there anything else?

Shanif Dhanani 4:46

Yeah, so I'll kind of give you the quick rundown. We definitely plug into Facebook. So Facebook ads, Instagram because it's part of Facebook. We do a lot with SMS. So things like attentive and postscript email, which is kind of our bread and butter. We do a lot with Klaviyo, a couple of our folks have MailChimp, couple of folks have HubSpot. Those are some of the main technologies and tools that we plug into, because that's what a lot of our customers are using. But with our software, you also have the ability to just take any segment, download it to a file, instead of using one of our built in integrations, and then exporting it and importing it to whatever you want. And so we'll try to go wherever you are. But you're right, you know, Facebook ads is a big one. And then a couple of the other technologies that I mentioned, we also plug into.

Andy Splichal 5:31

So what what is through the process, if a new potential client calls you guys says, Hey, I heard about you, you know, I heard about you on the podcast. How can you help me? I mean, where do you even start?

Shanif Dhanani 5:46

We'd like to, you know, what we do is we like to have a quick chat with with customers, because we aren't using AI. And so what that means is, we may not be the best fit for a brand new brand that doesn't have any data. So what will generally do is it will schedule, you know, quick 20 - 30 minute chat, but the brand will tell them a little bit more about us. And we'll ask them hey, look, you know, we'd love to learn more about, you know, are you guys doing any marketing today? How much data do you have? How many historical customers do you have? What's worked well, what has customers that are a good fit for us, they tend to have, you know, one or two years of data, or maybe you know, a few thousand historical orders if they're a little bit younger and newer. And we say, hey, look, you know, we plug into a lot of the big tools like Shopify, WooCommerce, Bigcommerce, whatever, we make sure that we are able to integrate their data. And if all goes well, you know, if they're a good fit for us, and we're a good fit for them. First step is basically just connecting your store, which is signing up on our product, and then clicking a few buttons to get, for example, your Shopify store connected or your Bigcommerce store, it then takes a little bit of time for our system to pull down all of your data, and then sort of analyze your customers forecasts what they're going to do next. And then once that's done, we'll actually jump on another call with you will say, hey, look, here's what our system has found, it thinks that you should be targeting these people with an offer where it thinks you should be targeting these people with a particular product. And we'll walk through that information with you. And then we'll help you get set up. So you know, soup to nuts, it can take a couple of days. But you know, some brands move really quick, some brands move slower. What we try to do is be a helping hand along the way, because we know not a lot of folks are using sort of these data driven techniques today. And we want to be able to help educate folks. But within the first week or so of getting set up, we've had brands drive 30,000 of sales, sometimes a little bit more, just by using our segments to basically plug into their existing tools. So it can be pretty quick. And you can see results fairly quickly within a few days, if you want to move fast.

Andy Splichal 7:41

Now, do you have you had mentioned you know, a company but do you have a favorite sucess story of a client you could share?

Shanif Dhanani 7:48

It probably is one of those brands. So let's talk like let's say there's a coffee company, Canadian Coffee Company that are really sort of strong value prop, strong brand image. And what they did was they plugged in after you. And what we did was we came up with, there were probably several, you know, 20, or 30, different cross sell opportunities, along with maybe 10, or 15 different other segments that we found for them. We said, hey, look like there's a bunch of really interesting things we could do here. One, we can take a lot of these cross sell ideas and recommendations, and use them to launch email campaigns, you know, if there's a group of people who are likely to buy your, for example, five pound big bag of coffee, let's send them an email incentivizing them to buy this big bag of coffee. And we already know, or at least our AI thinks that they're more likely to buy this coffee than not. So that particular campaign did well. And we'd launched several other campaigns for them like that, not only on email, but also on Facebook. And so, you know, between the combination of what we did for them on email, which was, you know, targeting their at risk customers cross selling new products, incentivizing people who are AI thought would respond well to specific types of offers. You know, we drove about 30k in sales for them among all the channels. And even on Facebook, you know, we we started to see three to three to 5x row as on certain channels and certain campaigns, by email was our bread and butter with them drove about 15 to 20k. So, yeah, that Canadian coffee company did really well. We'll see similar results, you know, with other sorts of companies, we do a lot of stuff with fashion and cosmetics and things. But that's probably my favorite story so far.

Andy Splichal 9:28

What are some of the challenges that you struggle with and getting results for clients?

Shanif Dhanani 9:33

I think I mentioned that a little bit earlier. Sometimes we have really early clients who come to us maybe they just launched their stores

Andy Splichal 9:41

So lack of data.

Shanif Dhanani 9:42

Lack of data is a big one. That's a big one. And you know, I'll be the first to say, you know, my background is in data. I think data can help everybody but if you don't have it, it's really hard for us to use it. And then similarly if you're not collecting it, you know, even if you've got it but it's in like five different places. Or you know, One thing that is kind of that we've sort of had a challenge before is a brand who maybe had data on one platform, let's say WooCommerce. And then they migrated to Shopify, and now they're starting from scratch, they don't have that data anymore. Those brands are a little bit harder to work with. So definitely, we can work with them. But it requires the brands to put in a little bit more work in terms of getting us access to their data, getting us information. And then the other thing is, you know, brands that are really busy, maybe they don't have time to take our segments and our recommendations, but they don't want us to do it. In those cases, you know, the project kind of lags, it's it sits a little bit. And so it really depends on you know, A. how much data a brand has and B. what's their willingness to sort of put that data into play and experiment with different campaigns using it.

And I guess another challenge, I would assume, because you've been talking about WooCommerce, Shopify, so having your own but you can't really help somebody who's selling their stuff on shop or on like Amazon, right? They gotta be selling their stuff on their own website.

It sounds like you're very, yeah, you're very familiar with some of the data issues. So for those folks, I'm sure everyone knows what Amazon doesn't provide you, you know, data, it provides anonymized email addresses provides very limited information. So you're absolutely right. We don't do a lot of work with folks who are on Amazon, or exclusively on Amazon, we'd love to get into that down the line, because we know there's a lot of business there. But what's great is that folks who have built who've taken the time to build out their own domains and their own stores are very serious about their businesses. And so they generally have the good data infrastructure that we're likely to do well with.

Now, if you had a crystal ball, how do you see this industry changing over the next 12 to 18 months? It doesn't seem like a lot of people really use it right now.

It's really interesting. So one of the things that I keep getting asked about is, you know, what happened? Like, How bad was the the Apple iOS privacy updates? And next year, how bad is it going to be with Google's, you know, getting rid of third party cookies. And I think that the next 12 to 18 months are going to be really dictated by these sort of pushes to privacy where marketers now are no longer going to be able to advertise or market in the same effective way that they have been. And I think that we're going to start to see this big push to getting around the challenges that come with that. And it's already started to happen. You know, a lot of folks I've talked to have either shut off Facebook ads completely, some have shut down their stores, which is kind of sad. But others have started moving to new new channels like TikTok and affiliate marketing and influencers. As Google starts to remove third party cookies next year, I think it's I think the push towards new channels, and using your own customer data for your own marketing is really going to accelerate. So I think we're going to start to see, I do think we're going to start to see a few brands start to consolidate and shut down because maybe they're going to be affected overly. But for those folks that stick around, I think you're going to start to see a pretty big, a big push to get access to new channels, new distribution. And then we've already started to see folks who have not worried about their data before start to use it to do things like reengagement campaigns, lookalike audiences, and things like that. And I think there's gonna be a big push in that direction.

Now Apteo that's an interesting name. What's your story? What's the story behind that?

Yeah, this is going to come a little bit from left field. But when we first started the company, we were actually an AI research organization. And we were researching AI for finance. And one part of finance is options and options trading. So I took the word option and I kind of played around with some of the vowels. I looked to see if there are any sort of website and domain names still available. And you know, changing some of those to A's and E's I came up with the word Apteo now it's been a long time since we worked in finance. We're exclusively in the world of Ecommerce now but that's kind of the origin story of the name.

I was curious So if you chose the A's, so you came up first alphabetically on servers.

I, you know, I was I was definitely thinking about that when we were playing around with that word. And so that was one thing that helps, as every little thing helps. And so I was like, alright, let's, let's use the A for the first first letter, and we'll go from there.

Yeah, I know, Yahoo, small business renamed themselves Aabco. Obviously, it didn't help. But let me ask you personally, are there any business books out there that you can attribute to your journey as an entrepreneur?

Yeah, you know, I've been in in the world of startups for a while now, I read a couple of books that I wanted to mention. The first one was Lean Startup. And then the other one was Zero to One. Now, these are old books. And so they you know, me and these books grew up in a world where sort of the MVP, the lean startup methodology working, I'm not 100% sure that was applied today, given all the capital that's out there. Today, nobody wants an MVP, they want sort of a fully functioning solution that does what they need. But those books helped me understand how to be scrappy, and most importantly, how to talk to customers to figure out exactly what they need. And I would still recommend people read them because they are still valuable and good resources. So you know, Lean Startup and Zero to One very good resources to get started with. I'd also say there's one other one called Talking to Humans, which was written I think, someone Franklin Milewski, maybe by Diane, NYU, very short book, I think it's like 40 - 50 pages, talks all about how to get the product market fit by interviewing your customers.

Now, speaking to customers, what problem is Apteo solving for their customers? And how are you standing out from from the competitors? I assume there's others doing this?

Yeah, you know, it's, it's funny, you've kind of heard me talk a little bit about this. But just to make it, you know, painfully obvious ecommerce brands are, they're not necessarily hurting. But there's a lot of competition out there today. The number of Ecommerce brands that have come up in the past two years has grown exponentially. And at the same time, the advertising technique, Facebook ads that they used to use to build themselves is no longer working well. And so these brands are having trouble driving enough sales to show growth to get more investment from their investors to keep up with expenses. So at the highest level, just like any business, you know, we're helping brands drive sales, but what it comes down to for us is how do brands market their services and their products effectively, that's kind of the problem that we're solving. Today, you know, I think it's, I read something like the average person sees, like 5000 ads a day. And so if you're going to stand out from the crowd, you really have to make sure you're talking your customers language. And we help brands solve that problem, essentially getting through cutting through the noise. Now, there are a bunch of marketing companies out there who are doing this. Now you've got folks like nosto, who are helping people personalize their on site advertise their on site products. So you land on somebody's website, you'll see products that are tailored to you, you starting to see companies like a tentative and postscript start to come up with SMS ads that hit you directly where you look, which is your messaging app. You've got companies like you know, mailing tools who are trying to do things like send you the right mail at the right time, what we are doing, we're coming at it from a pure data perspective. You know, myself, my co founder, our background is an AI we used to work at Twitter, we learned how good data can be, we're coming at it from a pure data perspective, we're saying, Look, if you can find a group of people who are likely to respond to a particular message, then it doesn't, then you can then use that information across all your channels to reach those people. So what we focus on really is really, really good AI and really good data science, to help you identify what messages you should send to each customers. This lets you send fewer messages. It lets you spend less money on the messaging apps, and it still allows you to boost your for example, return on adspend if you're on Facebook, or your conversion rates on Klaviyo while sending fewer messages and kind of wasting less of your time. And that's kind of how we've approached the world. And we've seen pretty good results for our clients.

WhyKlaviyo? Seems the last four or five podcast guests that I have had on that have talked about, email marketing is been there esp. All of them say Klaviyo, Klaviyo, Klaviyo, why? What's going on there?

You know, I've always I would say 90% of our customers also use I call them Klaviyo, I hope that's the right way to pronounce it. I think that they've done something smart, which is they took they took everything about MailChimp that was working well. And then they put in everything that ecommerce marketers needed for ecommerce into their product. And by doing that and by sort of building up this feature set that people from E commerce can really use, for example, plugging into your product catalog super easily identifying, you know, a different way of doing product recommendations, keeping track of how much sales you're making from your emails, they build this product that sort of SMBs and mid sized marketing teams across the world of Ecommerce are able to use easily. It's a very nice tool. And so you know, I said most of our customers use Klaviyo, we plug in the Klaviyo, one of our best integrations is actually them, where we sort of supplement customer profiles on Klaviyo with new information, we create segments. And so it's just a it's one of those tools that sort of taking the cake when it comes to e commerce marketing.

So who is the perfect client for your service, if they're out there listening now they should absolutely get in contact with you listening to this.

We love working with folks who are sort of D2C retailers, you know, selling things in sort of industries where customers will come back and make a second or third purchase. So what we do really well with is like food and beverages, cosmetics, fashion. Even things like consumer electronics, we've also started working with companies who are doing medical devices. From an industry standpoint, some of our best clients and customers are folks who are selling to, you know, to the same customer over the course of their lifetime. From a technology perspective, brands that are plugged into Shopify Bigcommerce WooCommerce, who are using tools like Klaviyo who are using tools like Facebook ads are some of our best customers. And then from a data perspective, we do like to see you having at least like a year of historical data, at least like 1000 orders, 1000 customers in the past, so that our AI is able to learn. Now I could say anyone selling online can use Apteo. But the tools, the techniques, the industries I just mentioned, those are the ones who make the most out of our software.

How many different types of products do you need to be selling? I'm thinking of their coffee company example. If they're only selling two different flavors in small and large sizes? I mean, how many different buckets to put them in? Right?

That's absolutely right. And you know, at the same point, like if you're only selling one or two products, it's easy for you as a human to just go and promote those products in your ads and emails. In general, we'd like to work with brands who've got 10 or more products, you know, it's not a hard and fast rule. But we start to see that some of our cross sell recommendations work well, once you've got about 10 products, you could still use us like we still do well, for brands that have maybe five or six. But you're right, one or two is probably at that point, you're probably better off just doing the marketing by hand or finding new influencers. And so yeah, like, you know, that 1000 order minimum plus maybe 10, or 20 products is kind of where we start to start to do really well. And then for brands that have like one of our customers, I think 17,000 products, because they can print on demand, they can print clothes on demand, that's where the AI starts to really shine, because no human is able to sort of parse through that data as quickly as a computer can. And so it's one of those things, you've probably heard it before anywhere, the more data you have, the better the AI will be. And that's kind of the case with us as well.

Now, do the products need to be different? Like go back to that coffee company. If they had six different flavors? I mean, is that different enough to know like, you know, they try this flavor? They'll they'll know this one? Or is it really needed? They'd be certainly selling coffee? And are they selling tea, selling coffee mugs? I mean, do they have to, how similar can the product speak for it to work?

Yeah, um, our software basically works off of a SKU. And so if you've got a coffee flavor, maybe hazelnut, that's one SKU and Madagascar, vanilla roast, that's another one. That's perfect for us. Because all that matters to the system, you know, the system doesn't really know that they're both coffees, but what it'll do is it'll look through the buyer behavior, it'll look through the viewing behavior for each SKU, and it'll make its decisions based on that. Now, one thing you can do, which makes it even better is you can group your products into collections or categories. And then we'll can take we can take information about the categories that products are in, we can take that into account. But you know, short answer to your point to your question, as long as it's a SKU, it's considered a different product. And we can work with that.

So how does your fee structure work?

We, you know, what we do is we basically charge we're a monthly charge unless you want to do an annual subscription, and we charge based on the number of customer profiles that we are analyzing for you. So what does that mean? Now let's say you're, let's say you're a small store, you're starting out, you've got a couple thousand customers who have bought from you in the past, that will be you know, a pretty small monthly charge. And the reason is because our AI uses your customer data to learn and so it's got less data to process, there's less of a cost to us, and maybe there's less that you can do. But let's say you're a brand that's got 2 million customers, lots and lots of products, you know, we'll charge that brand a little bit more or a lot more, you know, depending on the case, because not only are we then, you know, spending more on our side to train the AI, but we're also producing a lot more predictions and recommendations and segments for you. And so smaller stores will pay less larger stores will pay more, starts with a couple $100. And then it kind of works its way up to there. Now in general, we we generally never ask brand to pay if we're not driving them money. So we offer a free trial where we can sort of evaluate how much we'll make them. And we've seen us make like 40x ROI on some brands and the brands that are paying us or at least making 2 - 3x ROI on a monthly basis for our subscription.

Now, what is a piece of actionable advice that you can give listeners on something that they heard in this episode that they can implement right away for under 50 bucks, it'll help grow their business?

For sure. I think if you're not collecting any data today, you got to do it. Like let's say you're a brand, you've got all your customer data in Shopify, if you've never looked at it, I would say plugin, some sort of analytics tool to help you there's a bunch of free ones out there, just plug in a tool to look at and understand your data. If you're not collecting email and phone number information, using something like pop up ad or a pop ups form on your website, got to start doing that right away. So that's something that you can really do. And then I would say for those who are more more advanced, they've already got data segmentation, segmentation, segmentation, don't send the same message to all of your customers, parsing them out into different groups, and then send those groups messages based on what they're likely to respond to.

Well, this has been great. How can an interested listener learn more about working with you?

We'd love to you know, we'd love to chat with your listeners, our website, feel free to get in touch our website is apteo.co. That's apteo.co. I'm also always happy to chat my email addresses is shanif@apteo.co. We'd be happy to extend the you no longer free trial to any of your customers, Andy

It's been great. Well, anything else you would like to add before we wrap it up today?

No, I think this has been really helpful. You know if anyone's ever interested in chatting, AI and data, I am a data geek, so feel free to reach out but otherwise, it's been a real pleasure.

Perfect, perfect. Well, thank you again for joining us today.

Likewise, thanks so much

For listeners. Remember if you liked this episode episode, please go to Apple podcasts and leave us an honest review. And if you're looking for more information regarding Apteo or connecting with Shanif you'll find the links in the show notes below. In addition, if you're looking for more information on growing your business check out our all new podcasts Resource Center available at www.makeeachclickcount.com we have compiled all the different past guests by show topic in included each have their contact information in case you would like more information on any of the services I have discussed during previous episodes. Well that's it for today. Remember to stay safe, keep healthy and happy marketing and I will talk to you in the next episode.