Whether you work at an established e-commerce business or a growing startup, our state of mobile shopping webinar had you covered. Follow the link to watch the recording.
We had a fantastic panel: Matthew Jones from Everything But The House, Yasmin Benatti from Shopify, Luciano Vázquez from Mercado Libre, and Raphael Leal Carletti from Farfetch.
Four great minds and some excellent audience questions: this is what we’re going to distill here into some choice learnings from the event. Consider this a TLDR, but we recommend you watch the full webinar back (and, of course, join us for a future webinar). And as always, the answers have been edited for clarity.
Since early 2020, mobile shopping has changed. A lot.
Ok, so maybe this comes as no surprise to people, but it’s worth calling out nonetheless. We also explored this topic more generally in our blog: The evolution of mobile apps in a post-pandemic world.
Matthew: “The road of mobile shopping has changed a lot. Especially when the pandemic really started picking up here in the States. It changed everything. Like, I still, very rarely, go out. I've got food and groceries and everything delivered to my door now. I think people's experiences with mobile shopping have changed a lot.”
Yasmin: “Yeah. If I think about myself and how I handled online shopping before, I would only go to online stores to see if I could get a deal on a very expensive thing, like buying a TV or buying a microwave, or something like that. But then it changed, and I started getting the smaller things or games or things delivered from other countries to my house. And I started getting groceries delivered to my house. I started getting food every single day delivered to my house. I started shopping from people who didn't have a store before. So when we see that change, from our perspective, we can see that the e-commerce platforms and companies had to change as well.”
When it comes to payments, give people options and listen to their needs
Yasmin: “Yeah. At Shopify, because we have different verticals and final users who are buying something (and we have the merchants and we have the partners who create third-party applications for the merchants), we have to listen to them and what they want to use. There is a partnership with Stripe as well, so we can accept more payment methods, credit card methods, and we have our own solution of payment with Shop Pay. And Shop Pay will allow you to use the same credit card, for example, on different platforms. So you're buying on Instagram with Shop Pay, you're using the same credit card.
But I think it's something that you have to pay attention to your customers and listen to what they are saying. So for example, our partners, they want to have direct access to their bank account, instead of receiving through PayPal, for example. So it's something that it's going to be very local from place to place. And keeping that relationship with the customer very close is going to give you most of the answers that you need.”
Matthew: “You bring up an interesting point, that some people want to pay directly through their bank accounts. That's a challenge we have at EBTH. Some of our items go for tens and tens of thousands of dollars. When I say we sell rare stuff or unique stuff, we sell really rare and unique stuff. And we even sell cars. You can't buy a car off a credit card. You can't buy a car with a debit card. And one of our challenges is we have to set up ACH transfers so that people can connect their bank accounts directly to us, and use that as a payment option as well. And I think that's a unique challenge that... Particularly like what you were describing, it depends on the items being sold and the type of marketplace that you're looking at. But it's definitely a challenge that I think a lot of us have to face.”
“Analytics is a huge thing because it's more important than ever to be data-driven”
Raphael: “I think that analytics is good everywhere. We have to have it on every page, in every part of our code. It's really a valuable output about our customers, and with that, we can make it an approach that makes sense for our rep in case of AI, preferred items, or features that we didn't imagine before but all the customers are looking for. I think that's a good approach. You know everything about your customers.”
Matthew: “For sure. It's one of those things that you don't necessarily think about upfront. And what gets measured, gets managed. And the more data that you have, the better decision-making you can make. Yasmin, What type of analytics am I able to pull if I'm a Shopify store owner?”
Yasmin: “A lot! So for a lot of reasons, we want merchants to be able to make smart decisions. So how many items you have left, how many items people were purchasing, what they're purchasing more, how much they pay for it. And if they enter your store, click on things, and leave the cart, if you start accepting a new payment method, do they purchase it more because you're accepting that? If you insert in your feature, like, a 3D modeling of your handbag, for example, does it convert more people to purchasing and not leaving your shopping cart?
And on the payments that we were talking about before, there's also this feature where you can have this Shopify account for your merchant, and you can split your finances from your personal account to your business account, so you can see how much you're spending with your business too. So it goes from understanding your customers to understanding your business and understanding your financials so that it's a combo.”
Luciano: “We use a lot and we have different approaches. So you were talking about, let's say, client-facing analytics, and we have something similar to Shopify, but we also have the analytics for us, right? So what everyone is doing on our platform, which item, which button, which space, everything is being stored and used to help us make informed decisions. We use a lot of experiments. We do everything as an experiment and we start iterating on that. We still work as a startup company, trying, failing, trying again with a different thing, and if it works, we iterate it, we create a new thing, and we keep on growing. So basically, analytics is everything for us. Every decision is based on numbers here.”
Matthew: “I am so glad that you mention that. Not just the customer-facing side, but also the backend side, because that's not necessarily the "sexy thing". You know, everyone wants to talk about, like, product recommendations and things like that. But if you can improve things about your internal operations, that can make or break a business.
So a great example, at EBTH, if you walk through our warehouse, we have a whole photography department, and that section, it feels like a rave because there's, like, flashbulbs going off all the time. But one of the things that we looked at was, because we take so many pictures, one of the things that we run into is the time it takes our photographer to take pictures. And if we can optimize that process and make it better so that they can take more pictures, it allows us to get more items up on the site faster.
And so that was one of the things that we looked at in the engineering department: can we increase the number of pictures that are being taken. And we were successful in that.
We found some kinks in the workflow and that we were able to use our analytics, and we were able to say, ‘We were taking this many pictures per hour before, and now we're taking this many pictures per hour.’ And not that it's just, like, raw, because obviously you want to make sure that you're taking good pictures, you want to make sure that you're accurately representing the product to the buyer. But at the same time, if we can fix those inefficiencies, then that's a huge gain to the business and a benefit to the folks that are selling products to us as well.”
Use the right tools and processes to prevent bad releases
Yasmin: “The bigger you get, the more solutions you have to create to avoid releasing any bugs to projection, because the pain point is when they are out there. Is like when you accidentally create something on the wrong range and you push, and then you have to revert your commit. and you're force pushing to the branch and the other person already merged that, and you're like, "Oh no." So you have to create tasks, you have to have a good CICD pipeline that will run any test that it needs to…”
Matthew: “Like Bitrise. Sorry, go ahead.”
Yasmin: “Yeah, so you have to have good, stable solutions that allow you to deploy your code with confidence that you're not doing anything wrong. But if you do something wrong, then you can have solutions for people on call, for example, on Black Friday. And I think and I hope that with all the apps, Apple is a little bit nicer during those periods, just because if you have a bug, you need them to revert things faster. So, you count on their empathy as well.”
Matthew: “No, for sure, for sure. And as you put it, the best thing is having systems in place to prevent the bugs from going out in the first place. I am a huge believer in unique testing and integration testing and having absolutely every pull request checked through a CI/CD system. And yes, a shameless plug for Bitrise here.”
“It’s about the right ecosystem and not specifically about the features”
Luciano: “For us in Mercado Libre, it’s about the right ecosystem and not specifically about the features. We are a Latam company and our context is unique. For us, it's important to have a complete toolset — an ecosystem that should include e-commerce, payment methods, personal grades or loans, really, really strong shipping capabilities, and logistics, all combined in order to create a seamless journey and robust experience to our users. So for us it’s a more integrated approach of the complete experience than specific features.
But there are a lot of challenges.
For example, creating shipping capabilities here in Latam is really difficult because our mail services are not as good as in other parts of the world, so we needed to create our own network of shipping them. And in Latam, we are like a couple of years behind on the technology side. Our standard devices are less top of the line — they have less memory, less storage. So we need to work on improving the upsides, the use of the network.”
The right solutions for the right situation... On using WebView
Luciano: “So, we are not using React Native, we are not using Flutter, we are using WebView. Why are we using WebView? WebView is something we have been working on for the last five years now. We've been iterating the process for a long, long time, and we think it's the right tool for us at this moment. And why? Because using React Native or Flutter, we would have to start from scratch in a lot of things. Our web stack, we have a really strong web stack that is React Node Express. We've been building that for a lot of years, so that's why we are focusing right now on WebView. It doesn't mean that in the next year or two years we are going to use other technologies.
We are trying to focus on a balance between quality and delivery time. We are a product company. So, even though on the technology side, maybe we want to test different things, we need to balance this. Think about it, how last year, we grew, like, five years in one, or maybe more, so we needed to be faster on delivering products. But the interesting thing is that we started this path saying that we will have a model, and that model is the one that doesn't feel like a WebView. Right? So, we are creating web use that doesn't feel like you are in a WebView. Of course, if you are a developer, you would probably know the difference. But we are working on simulating a lot of the transitions, a lot of the things that you could have in native development. So, we created something that we call a bridge. So, from JavaScript, you could use a lot of Native features.”
Matthew: “So I'll be honest with you. The first time that you mentioned this it completely blew my mind. But the more I thought about it, the more I saw some advantages to the approach. The first one is, in my world, if we push out something that, you know, escapes all of our unit tests, escapes our integration tests, and we get a bug in production, we have to release an update, get it through the Apple approval process, and then be at the mercy of people to update their phones. In your world, if you're using WebView and the bug’s in the WebView, you just update the WebView, and that's immediately in production.”
Luciano: “Exactly.”
Matthew: “The second thing that would be an advantage to that approach that I thought was really interesting is that before I got into mobile dev, I was a web developer. And with progressive web apps, there are so many things that you can do now that give you, like you were saying, access to the native hardware on the phones. So you have access to a lot of things, like an offline mode and things that people associate with native apps that you don't necessarily associate with web apps.”
Luciano: “That's exactly the point. We had everything working in place. And to start with React Native, for example, we needed to learn a new technology. We needed to create a lot of features to educate a lot of teams. And as I said before, with the analytics part, we are testing, we are experimenting a lot. So building that product with Native or whichever technology, it takes more time to hit the market and to iterate it. And with WebView, we have those capabilities of moving fast, right? We can test, we can iterate, we can create a lot of things faster and hit the market faster.”
