Enhance Onsite Experience with Brand Concierge

Transcript

Hello, my name is Nina Giocano. I’m a senior software engineer in Brand Concierge. Today I want to talk about how to enhance on-site experience using Brand Concierge.

I want you to meet Marcus. Marcus is a student at university and he has to make a documentary film for his project. He’s looking for an application to help him build, create and bring his visions, creative visions to life.

So what do you think would be his next step? Obviously, he will go and search for best software to create or to edit a video. He will get on different blog posts. He will start reading reviews, what other people are using. He’ll get on Adobe.com. He’ll see over 20 applications and he’ll get overwhelmed by the options, but also he will get excited by the examples and potential. But still he is unsure which product is the best fit for his use case. So what if Marcus could just ask? So here we have Adobe.com website with Brand Concierge on it. And Marcus would ask something like, Hi, I need help to create a documentary film for my university project. And here is the answer from Brand Concierge. So what we can see is an answer with reach with Adobe.com information. It has links to the products.

It has product cards with images.

It has sources and citations.

And also follow-up questions for Marcus in case he needs to learn more about the product or maybe find another Let’s say Marcus would click on Adobe Premiere Pro and it will navigate to Adobe.com Premiere Pro product page. You can see that we have enriched the URL with the source so that later we understand that this product view and eventually purchase was initiated in Brand Concierge or as part of a conversational experience.

I want to go back to Brand Concierge initial view and this is the welcome template.

As you can see it has a welcome message and a few examples of questions. This can be customized by the customer and they can change the welcome message to sound differently for their customer or their audience and also the questions.

Let me show you a different answer, a different configuration which we have on a demo page where the customer can have instead of a single product view, they can have a carousel or multiple product views. So let’s say Marcus is happy. He found his tool to create his project for university and now he might ask a different type of questions to Brand Concierge like What is the meaning of life? And this is the answer. It’s an out-of-scope answer. While this was a joke, I want to prove a very important point that this might look trivial on the surface, but building a conversational experience requires a lot of preparation, discipline and effort and a big amount of effort from engineering.

A lot of engineers and customers underestimate the complexity of building a production-grade experience that will delight their customers. What we have seen is an example of a user journey that started with a navigation and then continued with conversational experience and ended up with a product view. In this case, a product view is considered a conversion. At Adobe, we have one of the best-in-class analytics tools, which is customer journey analytics.

And as part of Brand Concierge product, we have added support for conversational analytics. And here is a glimpse of what is possible today. Using analytics, customers can have the ability to see how traditional site navigation compares to conversational experience and they can decide where the conversational experience makes the most sense. In this particular example, we see a daily trend of number of conversations. For those familiar with web analytics, this is similar to web page views or clicks. However, a conversation provides more insights into what a user is trying to do. We can derive the sentiment and see how our communication with the user can be improved.

Besides sentiment, we can detect user intent. While in the past user intent could be loosely derived from user actions like page views, clicks, now we actually can see what the user is trying to accomplish. And we can provide a more useful and engaging experience.

Sentiment analysis and intent detection is just a few of the capabilities that we have seen. Lastly, I want to talk about how conversational experience can lead to product recommendations.

Here we can see how different here we can see different recommended products and how many recommended products were recommended in a conversation and how many of those recommendations led to purchases. So what is BrandCon Cieře? BrandCon Cieře is a conversational layer for your digital property, which is seamlessly integrated into one-site experience. It’s an AI-first application that transforms static website into personalized brand-guided conversations helping users discover products, qualify needs, and build loyalty. It provides personalized response. It understands site content, product catalog, user intent, context. It provides personalized responses, recommends product and suggestions, and also provides conversational analytics. So why BrandCon Cieře, you might ask? BrandCon Cieře gives a rich user experience. Most of the demos are simple. Text in, text out, but BrandCon Cieře allows rich user experience.

It also gives you integration with AEM and AEP ecosystem. It is easy to create a demo-like red pipeline, but doing it correctly and safely takes a lot of effort and Adobe handles that all behind the scenes.

Customer journey analytics. Customers get this out of the box. They don’t even have to set up datasets.

So obviously this is really cool and you might ask yourself, how much effort would it take to implement it on your page? It’s simple. As a prerequisite, you need to be a BrandCon Cieře customer and to have Web SDK on the page. Then as part of implementation details, you need to inject BrandCon Cieře client on the page and then initialize BrandCon Cieře client on when you want it to be visible for your users. If talking about high-level architecture, we have on the customers website the BrandCon Cieře client and Web SDK. BrandCon Cieře client is responsible for conversational experience while Web SDK handles identity management, context gathering and low-level communication.

Then we have BrandCon Cieře runtime, which handles which is responsible for routing session management and response orchestration. Agent Orchestrator handles chat history and it’s also responsible to pick the right agent for the right task. For example, a general question will be handled by the question answering agent. A support request will be handled by the support agent and any out-of-scope question like, what is the capital of France, will be handled by the out-of-scope agent. What you have seen in the demo are just a few basic use cases that are already bringing value to Adobe.com, but there is more ahead of us and in the future we are looking to add personalization based on AP profile, journey optimization and third-party data integration and also expand CGA integration. Key takeaways. The future is conversational.

Every digital property will eventually have its own conversational layer and eventually this will morph into an agentic experience.

AEM is the perfect foundation for conversational experience. It has structured content, metadata, APIs, which provides all the necessary blocks.

Start small and iterate fast. Don’t forget about safety and security, but also try to get a real feedback from your customers or users.

Conversations transform experience, increase engagement and drive conversions. Thank you.