[Music] [Scott Rigby] Hi, welcome to Adobe Summit. My name is Scott Rigby and I'm a Principal Product Marketing Manager for Adobe Commerce. And I'm super excited today to be able to present to you How AI is Transforming the Commerce Ecosystem.
I don't think any other topic has really got as much focus in the last couple of years than AI. And obviously we've had the benefit of things like ChatGPT that have really driven this rapid adoption of AI and in respect of the commerce landscape, we start to see some very fluid changes in respect of capabilities that are being made available to customers, and just customer expectation as a whole is changing around AI and what this can mean for them in respect of better commerce experiences. Hopefully, through today's presentation I'm going to walk you through some of the areas that we are starting to see AI development, but also talk to you about what's already available within the tool set for you to be able to leverage today, and then what does that mean in respect of once we start to put, for instance, Adobe Commerce together with other AI capabilities, and what's the better together story look like and what's available for you to be able to leverage as an Adobe customer.
So let's dive into it. The agenda that we're going to be looking at today is really going to be looking at, first of all, a very high level overview of AI and what that means within commerce. I'm going to talk about some trends, and I'm going to talk about some areas that we're seeing investment that I want to talk to you about Adobe Commerce specifically, what's available in the tool set today, what's expected for roadmap for this year, and then what's the better together story? How do we add capabilities on top of the Adobe eCommerce platform so you can branch it out into utilizing AI into not just tens of different ways, but potentially hundreds of different ways, and hopefully give you an understanding of how we have curated this platform to be able to extend AI capability on top of it. Let's dive into it.
So the first area we're going to be diving in here is really just a quick overview of AI and then get into the applications of Ai within Commerce.
So an overview of AI, right? What is AI? Effectively, AI is technology that enables machines to be able to learn from data and start to make decisions autonomously. We're also using AI here, and more recently you would've heard about generative AI where we've inputting data, and it's then curating that into either text or it could be into an image, or it could be into video. And in some cases also generating code that we can either use to create amazing customer interfaces or create integrations to other tool sets. And so we have this expanding use case of where AI could be used. Within AI, there is this thing called machine learning. And this is effectively a subset within AI where the algorithms have very defined parameters and to learn around programming and be able to output to a large degree, a very curated subset of information. We also have natural language processing, and this is our ability to be able to interpret what is being spoken and then be able to interpret that into human language. And so things like LLMs are using this to be able to deliver that, large language models. AI in eCommerce. We're utilizing this everywhere from identifying interesting new segments to able to target through to curating that experience to them actually coming to buy from us, making recommendations while they in the purchase funnel and then delivering autonomous responses back to the customers potentially after they purchased with information around their shipping order, etcetera. And so we're starting to utilize technology across the entire spectrum of the customer lifecycle.
Some of the areas and capabilities that we're starting to see AI being applied, so personalized shopping experiences. This has been around for a while, but we continue to expand on top of what this truly means. So this could be us, for instance, curating the search capabilities. So what they are searching for, we are identifying the segment, the types of products or merchant categories that they're potentially looking for, or we making, identifying the types of products that they might be interested in. We're testing out layouts that are specific to a particular customer segment. It might be predictive analytics. What's your propensity to buy or what's your propensity around being a lead for a business that from a B2B perspective it could be, looking at what is the likelihood of what's the lifetime value? What are our sales going to look like in the next quarter? And so we can start to use analytics build to predict forward what this looks like.
Fraud detection, identifying, certain anomalies outside of the bell curve, and that identify that this is probably a fraudulent purchase or the idea that they coming from this particular postcode or this particular area is indicative of being fraudulent. And so being able to help identify and mitigate those, we're enhancing our capability, right? Our ability to do augmented reality experiences. So be able to render products, for instance, in 3D to make them very lifelike and be able to understand, are they a good fit for myself or my business, or whatever it might be. And I'll show you some examples of that shortly.
Chatbots and virtual assistants are getting a huge amount of focus, and particularly because of all of the focus around LLMs and what this could potentially do in respect of curating that customer experience. So that could be chatting, you write into a chatbot, help me identify mountain bikes that have both shock absorbers in the rear and the front that have these types of accessories that have this componentry made by this particular manufacturer because I like that is robust componentry, etcetera. I don't want it in purple colors and they can then go and quickly index and search for products that fit their criteria and be able to surface that up very quickly. Dynamic pricing, we're starting to test this idea of, what is your propensity to spend and can we drive an incremental margin out of this segment versus this other segment and trying to increase margins and look at that from a customer segmentation perspective. We might be looking at supply chain optimization, right? Are we ordering the right amount of products in the right mix across the different fulfillment centers at the right time of the seasonality for the year to ensure that we have products available, but we are not over investing in products so that they're sitting on the shelves and we have to discount and sell them out the back door later. And then obviously voice commerce. And again, this ties into large language learning models where customers are coming because LLMs are getting so good that, we can start to then use voice to identify the information around products that we're looking for and be able to research that, but also then start to make Product Recommendations and even potentially purchase that for you. And so plugging your commerce experience into that because it creates a shortcut tools to be able to buy from you and be able to research, more customers are embracing voice as it gets better with LLMs.
Next around just the market trends and adoption, as you would expect demand for personalization and we're all dealing with the customers that are what I would call the instant gratification generation that have high expectations around personalized shopping experiences that could be products and product catalogs and pricing that is curated for their experience, it's Product Recommendations, it's helping them search, it's helping them, it's rendering the experience in the way that they like to digest information. And it's providing the right degree of appeal to that particular customer segment. And in order to be able to scale personalization you can't do this manually, you need technology, you need AI to be able to do this. Automation around operations, right? How do we start to automate some of our capability? This could be everything from the content that we have around our various products, and be able to create different iterations of that to be utilized across different social media channels or across different bandwidths whether it'd be a 3G with a lower quality rendering or 5G with a 4K definition, we need to think about how we start to scale operations. This could also be utilized in respect of our reps and their ability to be able to accelerate sales and streamline what they do. We are enhancing our customer interactions, right? We're looking at real-time, how do we interact with that customer and give them the information that they're looking for at the right point of time so that we're driving them to the next best path to purchase or path to action or service, right? And then using analysis to identify areas that we can drive improvements in customer behavior and decision-making, right? So again, the quantity of data that we collect day in and day out is increasing day in day. And so we need to be able to utilize, again, AI to be able to surface this up and increasingly so make recommendations for us to be able to execute an action because we just simply don't have the bandwidth to be able to do this from a manual perspective.
Why should we be using AI? Well, there's numerous benefits for why we should be using it, whether that be customer experience, whether that be helping customers search and find our products, and that could be both B2C and B2B. It could be driving operational efficiency and cost savings. It could be driving better inventory management and supply chain management.
We have data that shows that 50% of commerce businesses are already utilizing AI, and so if you're not already utilizing it, then obviously we encourage you to get started as soon as possible. The gap between those that are utilizing it and those that have not continuous to increase in respect, there are accelerated capability, and so if you're not adopting it, we encourage you to do that as soon as possible. We can see phenomenal uplift from using AI. 15-20% around Product Recommendations. We can see an increase in traffic. You can build out the content that you are delivering that can then be indexed and made searchable through SEO, and that can, in some cases, drive 10X increase in traffic. And even more recently, in November, 2024 with the Black Friday sales, we saw retailers that utilized AI saw 10% increase on those that didn't. And again, you want to be at the forefront of your competition, you want to be utilizing these technologies to develop to get these commercial benefits.
So we look at Adobe Commerce in our AI capabilities. So I'm going to give you a run through of what's available, what's expected on roadmap for this year, so you have some understanding of the capabilities that it's already available within the tool set.
So we have Product Recommendations. This is looking at products people that have looked this also looked at that and expanding customer's understandings of the products that you have available, but hopefully, also increasing your average order value. Live Search is looking at, how do we search for the right types of products, making sure that out of the millions of potential SKUs that you have, customers find the four or five that are relevant to what they're looking for. Image generation. Our ability to be able to create images and be able to deliver personalized content at scale to individual customers is an aspiration. And certainly with AI we have this ability to be able to wrap this up.
Performance recommendations. One of the things that we are looking at is, how can we improve both the performance from your site in respect of the speed of which it renders, but also areas that you can improve the code so that it delivers a better experience for customers. Intelligent merchandising. This is our ability to be able to identify categories and be able to then use different types of ranking techniques to be able to deliver different types of personalized experiences for customer segment.
Generative variations. Our ability to be able to create different types of variations and so that you have not just one, but you have multitude of potential experiences, and you can start to test whether the experience that you currently deliver to this customer segment is optimal, whether there's opportunities to improve upon it. Segmentation targeting. So to identify certain segments and be able to then target content against that and those segments can be identified and surfaced up through AI. And then lastly, predictive targeting. Working out what is the right content to put in front of the right customer at the right time.
When we start to look at AI applications within Commerce, we have a number of capabilities that's already available within the tool set today. We have these capabilities that are already embedded within the tool set, and then we've got some that are on roadmap or expected roadmap for this year. I going to dive into some of these so you have a better understanding of what they do and how they can benefit your business.
So let's start with Live Search, right? This is our ability to be able to deliver search results and help curate those search results for customers. And this is really focusing around simplicity and speed, and I'll talk a bit more about effectively what this does shortly. But essentially what we're doing here is managing multi-language, we're identifying categories and filter facets that help customers get from those millions of SKUs down to those four or five. Product Recommendations, 13 different algorithms to use here are both image based and both data based. To be able to work out what is the next best product to recommend to customers so that they understand what's available to them within a certain product set, but also to expand that cart size and average order value. And then category merchandising, five algorithms to choose for from here where we can push or pull certain results up and down, utilizing these different algorithms, and then test that out to see which delivers the best result for your customer.
So let's talk about Live Search, right? So Live Search is, first of all, it's used quite widely amongst a number of customers, and you can see there's a number of well-known brands here on the left-hand side. But as I alluded to, this is effectively the ability for us to be able to make search suggestions to auto complete. We're also making recommendations on filters to be able to narrow down your search results. We're also helping you push or pull your search results so that the right products are being pushed to the top for a certain search result and those that are not being pulled down. And so typically customers that are leveraging this are seeing 7% increase in conversion. But we've seen some customers have phenomenal success of up to 100% increase in conversion by utilizing this tool set. And it's a robust tool set that's had, in this case, as you can see, billions of requests put through it so that we're leveraging this data to be able to better understand how customers utilize this and be able to deliver better results.
Now recommendations. I talked about this in respect of the 13 different algorithms that you have, but you have both database algorithms and you have image-based algorithms. Database algorithms being ones that effectively look at, for instance, a recent customer behavior giving a high degree of waiting to the most recent or recency of data and those products then be surfaced up and recommended to your customer. And you can choose different algorithms and you use both the algorithm and you can still layer manual rules on top of that. And then we also have image-based recommendations. So here we can effectively use our image-based recognition technology to identify and add keywords effectively, meta tags to those images and start to look at similarities between those images. So an example here might be a blue hoodie. And when customers are looking at one blue hoodie, we used image-based algorithms to look through the catalog and identify other blue hoodies that are available within our catalog and surface that up. And typically, we are seeing customers here achieve 15% click-through rate and 20% lift in average order value by deploying Product Recommendations.
Next, we have intelligent category merchandising. And again, here we have five different rules to automatically re-rank products sequencing and searching on the page. And so here, whether it be training products or it could be, those that have added to cart or recommended by you or the most purchased, we can start to use these different algorithms to raise your categories and be able to then test this with customers to see which delivers the most beneficial result for customers and therefore your business.
Next, we have example here from Coca-Cola, which obviously is a substantial brand. They're well recognized within the market, and they're using both our Live Search, as well as our Product Recommendations to drive some very significant business impacts. For instance, on the Live Search side, they've seen increase of 18.6% within the conversion rate, which is obviously millions of dollars for them. They've had 60,000 unique searches indicating shopper interests, which helped them then better curate that experience. They've had 117% increase in click-throughs from new products that are being displayed, and 17% increase in click-throughs for frequently bought together, effectively increasing the number of products within a shopping cart, therefore average order value per customer. And again, some pretty phenomenal results from a well recognized brand.
Next, we are looking at GenAI within commerce, and again, we're leveraging here our ability to be able to personalize content for different products against different customer segments. And example here is a B2C one where we're swapping out the background for these different pieces of apparel. This could be equally represented within the B2C environment. Maybe you are selling hard flooring that is utilized within offices or warehouses or various size of the types of commercial premises. You could swap out the background here to be representative of those different customer segments and start to use generative AI to be able to very quickly create these different types of representations. On top of that, we can start to then create text-based content for your product description pages or titles around your products to engage audiences. Maybe you want to change your current product description to be more value-based. And again, we can start to use AI to be able to very quickly personalize this merchandise for your customers.
Next, we're looking at what we call generative variations, right? This ability to be able to create a number of generations of content. Now this is more around the ability to be able to create variance. Right. So we can obviously create experience adding and that would be typically your default content. And then you're creating numerous experiences on top of that by utilizing generate variants. And with this video, you can start to see that we can very quickly start to expand out the number of variants we can create and start the test, which resonates best with your customers based on in the KPI that you said. Mostly this is orders. It could be some other type of KPI might be revenue based or whatever it might be, applications, etcetera. But we can start to very quickly expand out using generative variations to be able to create these and test them into market to be able to get maximum impact.
Next, let's jump to expanding Adobe Commerce in our AI capabilities.
So one of the areas that you can start to really expand out Adobe commerce here is to be able to leverage our integrations that we have. So we have database integrations with our technology into what's called Adobe Experience Platform, where we are effectively sharing data from Commerce into Adobe Experience platform that can then be utilized within a number of other tool sets. For instance, within Adobe Experience Platform, we can start to use AI Assistant around this data to be able to curate it further. We can start to do look-alike modeling and the like, or we can start to pull that into our Real-Time CDP and start to do media Mix Modeling. How should we spend our media budget against the different segments that we identified from the Commerce data? Maybe we can do propensity scoring or lead scoring. We can also curate journeys through Journey Optimizer that we could create adaptive journeys. We can decide on which offers to deliver to different segments. We can start to enrich our analytics through Customer Journey Analytics and start to do predictive analytics or Journey AI. And then we also got integrations into our content tool sets, so Adobe Experience Manager and assets where we can start to do content personalization, we can start to do contextual targeting, or with assets we can start to create different types of various assets that are also plugged into our Creative Cloud capability, which I'll show you shortly.
So I want to give you a quick example of what this looks like. So from a data side, we have what we call our customer side data. So this could be added to cart, it could be checked out, it could be products that are being viewed. We also have our service-side data. So this is order status, order delivered, etcetera, inventory availability. And we are starting to then make that available through Adobe Experience Platform into our Real-Time CDP. And so we can effectively now do things like ad suppression. Quite often what happens from a retail perspective is a customer buys from us, and then from the next day or week, we continue to deliver them ads for the same product they just bought from us. And so what we want to do here is, ensure that we are maximizing our ad budget by suppressing ads like that for customers that have already bought from us. We can make those segments available into Adobe Commerce where we're then curating the experience that customer gets. That might be based off Cart Price Rules. It might be based off Next Best offer. Or we might be managing their marketing journey with this data and delivering different types of journeys for those customers that have abandoned versus those that are bought new customers versus loyal customers. We can also impact drive this into analytics and have an impact there by looking at campaign analysis of customers that are responded to campaigns and then actually bought from us, and be able to identify the downstream impact of campaigns that are delivering a result versus those that are not, or we might be looking at different types of channels, etcetera. And so we can start to leverage this and we can start to feed this data, it's a two-way integration back into Adobe Commerce.
Next is around our digital asset management from the content side, right? We have this ability to be able to then create different types of content. We can slot, deliver it, things like Adobe Express, where we're creating different types of content with different backgrounds or different types of elements within those products. Maybe you want to make a product more realistic by adding certain elements to it. That could be adding a water bottle to a backpack or adding various types of accessories to a mountain bike. And so we can start to really utilize our asset management system here to be able to bring these products to create life, be able to create different types of iterations of that for different channels. You might use it in social media, and so the type of asset that you use within Instagram will differ to the one that you might use within Twitter, or you wanted to use it across different portals. It might be mobile versus desktop. And so again, we're rendering different assets here. We're also wanting to personalize that. So again, we talked about this idea of different audiences from a B2B perspective, you might sell into a warehouse, as well as into commercial office. You want to be able to swap out those environments that they appeal to the right customer. And with our digital asset management integration, we can very quickly create these different types of variations and make that available to commerce.
Next is our ability to be able to add our Creative Cloud. And again, we have got integrations into Creative Cloud. Now Creative Cloud has a large number of AI capabilities that could be everything from 3D model generation, it could be video creation, it could be creative assistance. I'm going to give you a couple of quick examples of how we're using some of our integration capability into Creative Cloud.
So one of those areas is within a technology called Substance 3D, and this integration has been around for a while. And the amazing thing here is obviously we're using AI here to render lifelike products. Now you could be creating products in different colors. It could be apparel. And we're going to test out different colors for our range for the next season. But we're not quite sure what to actually manufacture, and you have the ability to be able to manufacture on demand. We can use this integration with Substance 3D to bring those products to life. Or maybe you're a B2B business and you've got a bobcat type digger. And it comes with different types of tool sets that go on the end, whether it be an oger or whether it be a barrel for being able to pick up, etcetera. You can show the different types of accessories that could fit on the end of it and bring that to life from a customer perspective. And so this tool is fantastic in bringing this 3D element to life. If we wanted to give you some examples on both the B2C and B2B side, the ability of augmented reality, you could do, in this case we are rendering here a carry bag and we're showing this. We could show this in different colors, we could show it with different fabrics. We can also show the quality of the product here by surfacing up the sheen within the brass zip that it has. Or maybe you're selling a B2B robotic arm that does, so it has certain capabilities and they want to see if this will actually fit within their warehouse. And so again, we can use this, be able to bring it to life. Place it within that warehouse environment and test the swing arm capacity or breadth to be able to see if it fits in that environment and we can start to see the different types of tools that could fit onto the end of it. And so we can start to really bring this entire commerce experience to life by our connection into Adobe Commerce Creative Cloud through our connection with Adobe Creative Cloud.
Next is our ability to be able to connect to our Developer Marketplace. So we have a Developer Marketplace that sits on the Adobe Commerce platform, and we actually have 4,700 extensions, applications that sit on Adobe Commerce Cloud. And these capabilities are in a lot of cases just for a nominal fee. But you have a huge number of AI capabilities that you can bring in very quickly into Adobe Commerce and be able to leverage these on top of what's already included within the product, whether this be from a content creation perspective, this could be text or images, video, etcetera. It could be from an SEO delivering more Meta keywords to be able to drive better SEO indexing within search engines. It could be around language translation. It could be around fraud. It could be around support assistance. And so using our app builder technology, which is an extension capability on Adobe Commerce, you can very quickly integrate and bring these capabilities through. And so an example here, looking at the Developer Marketplace, you can start to see some of the applications. And we've got roughly about 80 plus applications that are available around AI today that you can readily use. And as you can see from the pricing in the bottom right of each of those boxes, they're for a pretty nominal fee for you to be able to add this capability to your business. So again, something we encourage you to leverage.
Next, I want to talk about this ability of adding further cloud services or AI services through your cloud provider. Whether it be chatbots, whether it be code generation, whether it be an AI Agent Builder or an AI App Builder. We have again, App Builder allows you to be able to bring these services, whether it be from AWS or whether it be from Azure into Adobe Commerce using the App Builder extensibility capability or what we call our App Builder framework. And so very quickly you can start to bring these capabilities into your Adobe Commerce solution.
Next is our ability to be able to bring in LLMs, large language models into Adobe Commerce. And again, using our App Builder framework, you have the ability through ChatGPT or whether it be Anthropic, Perplexity, Llama, Gemini, etcetera, to be able to leverage their capabilities and the unique strengths for your business and be able to leverage them within Adobe Commerce. The App Builder framework allows you to be able to provide this two-way integration.
So hopefully, you've get an understanding of both the AI landscape for commerce, and then more importantly, what does this mean for Adobe Commerce existing today? The capability the expected roadmap together with the extended capability through Adobe Experience Cloud or Adobe Creative Cloud, the extension capability through our Developer Marketplace, and then if you want to go even further, whether it be through your cloud service provider or various types of LLMs, you can start to very easily bring this capability to the fore to deliver some amazing commerce experiences for your customers, but hopefully better operational efficiency for your business and opportunity to create new products and new markets that you can go-to-market in. AI is really enabling businesses to operate more effectively and efficiently. And Adobe is at the forefront here in respect of bring this capability to you and making it readily available within your hands, as well as be able to very quickly bring in external AI capability if you are willing to layer it on top of the Adobe Commerce platform. For those that are not currently utilizing, we're obviously, encourage you to start utilizing it today. Those that are already using it are learning amazingly quickly, and they building up core capability and the gap between those that are not utilizing it is getting great and great. We encourage you to start learning today, use what's out of the box available to you today, and then start to think about how you can expand on top of that.
Hopefully, today's session has been informative in respect of AI and the commerce ecosystem. Thank you for your time. I hope you have a fantastic Adobe Summit. [Music]
