16 Ideas for AI-powered E-Commerce With Our Friends at Burrow

15 min. read

Have you ever wondered, "How might generative AI revolutionize E-commerce?" Well, that's exactly the brain-tickling question we explored during our collaboration with the US-based furniture startup, Burrow. In this blog post, we're diving into a range of concepts, from immediately achievable ideas to future-forward dreams, all to give you a sneak peek into the potential transformations ahead in the world of commerce.

Generative AI is everywhere these days. But what is generative AI, really? We like to frame it as "Something that didn’t exist until a user expressed intent towards it". Generative AI is what powers tools like ChatGPT, Gemini, Midjourney, and others. These tools all rely on large language models, which have rapidly improved in quality over the past months.

And while tech giants are in a sprint to develop AI assistants that can do just about anything, our focus here is on Burrow’s specific business and user experience opportunities. We're excited about the possibilities generative AI brings and how it might help even out the playing field, giving smaller players the chance to compete with the giants.

When collaborating with a brand like Burrow, our goal is to ensure that the generative AI serves the specific needs of this business. That's why the ideas below are all based on the concept of what we are calling a Brand Language Model…

These specialized brand models take all the cool stuff the big models can do and tailor them with specific guardrails and training to meet the needs of a particular brand and business.

What if...

When working on innovation projects like this, we often find ourselves asking 'What if' or 'How might we…?' These questions broaden our minds to new ideas. Below, you'll find the questions - and answers - that sparked our interest. We hope you find them intriguing too!

(By the way, we're not just daydreaming here; As part of a cross DEPT® initiative, we're working with our AI Greenhouse team to make some of these concepts a reality)

What if you could access the AI assistant from multiple points in the customer journey?

It should always be easy to start using AI assistant tools. Here we offer suggestions on what AI can help you with, or you can begin with a custom prompt. This approach mixes traditional search with AI-powered features.

We could enhance the search dropdown, possibly within a larger menu, to support freeform searches and provide starting points for AI use. Chips or cards in product categories could highlight AI features, guiding users in adapting to these new tools.

What if the AI assistant integrated naturally in the current UI?

Imagine visiting the sofa product listings page and having it truly guide you to your ideal sofa. To overcome the initial blank canvas uncertainty users often feel when landing on a product page, we'll provide examples to guide your search. This allows you to search for specific needs, such as 'sofas suitable for a small one-room apartment,' rather than just standard options.

What if you had an ever-present shop assistant?

You could search for "furniture for my college dorm room" and find the results smoothly integrated into the current website layout, providing assistance akin to that of an in-store salesperson.

Additionally, you will find contextual follow-up questions in the UI's bottom right, asking about room size or budget—details. Filters that are not predefined, but generated based on relevance to your question.

Or what about adding a history of your interactions with the AI? This feature would give you a summary of your past searches, letting you go back, modify, and adjust, using a more traditional UI to interact with AI that isn't solely chat-based.

What if the AI assistant helped you navigate the page?

This could be achieved with minor updates to the current front end. For instance, we could add a space where you could ask the assistant questions like "What are the dimensions of this product?" The AI might then scroll to the relevant part of the page to answer your question.

When asking for smaller options the assistant could expand and show these options in a way that's clear and doesn't mess with the page layout, making it easy to learn more and navigate to other parts of the page.

What if the AI assistant had its own personalized space?

This idea would function like an add-on, easily integrated into different websites. When you begin interacting with the AI assistant, you enter a "shop assistant universe," a specialized area for personalized help and suggestions, without major changes to the rest of the site.

This feature might be simpler to implement as it's self-contained and minimally intrusive to the current website layout. It offers a steady, dedicated space for customers to seek assistance, pose questions, and get customized advice from the AI throughout their shopping experience.

What if you could ask questions about anything across the website?

Imagine if you could long-press on anything on the website, like the legs on a piece of furniture, and prompt the AI to answer related questions, such as "Are there other leg options for this furniture?"

This feature would greatly enhance the website's current capabilities. For example, long-pressing a sale label could let you ask, "How long is this sale going to last?" The AI could anticipate relevant questions based on the context of your actions and offer predefined options. This approach makes the site more dynamic, moving beyond the static sitemaps and predefined customer journeys.

For mobile shopping or any platform with voice input, you could long-press on a couch and ask aloud, "Does this come in an L-shaped version?" The AI would then respond in a conversational manner.

Additionally, you could ask the AI to display measurements in different units, such as converting inches to centimeters, and it would instantly update the content on the site.

What if you could add furniture to virtual spaces as you browse?

Imagine if you could add furniture to virtual spaces while shopping online. Instead of just using Pinterest mood boards, this feature lets you place furniture into customizable virtual rooms as you browse.

This simplifies the decision-making process and helps customers envision the furniture in their own homes more accurately. The space could even be collaborative, allowing others to join in and contribute to the design process.

In this interactive environment, you'd have the power to manipulate the room. For example, you could ask to "show more of the room," "remove the carpet," or "paint the walls green."

What if generative AI made images come to life?

Implementing subtle video animations could make product presentations more captivating, potentially increasing customer engagement and adding a modern, interactive touch to the shopping experience. It's another way AI can democratize high-quality content creation, enabling businesses of all sizes to compete more effectively in the digital marketplace.

What if you could easily customize what you share?

Imagine you're browsing a product that you want to share with your partner or a friend. Instead of just sending a link to the product page, an AI-enhanced sharing feature could enable you to customize how the product is presented. You could select an interior style that matches the recipient's taste or the context in which they might use the product.

What if you could view any product in 3D and AR?

What if you could view any product in 3D and use augmented reality to see it in your own space? While some brands offer this feature, it requires high quality 3D models and an external app. Generative AI could be used to make the 3D models and textures, allowing you to simply scan a QR code to preview any product in your home.

What if you could browse furniture in the context of your home?

After detecting the furniture in your picture, you can explore all relevant Burrow furniture options that would fit in that space and category. Taking your existing decor into account, the AI suggests pieces that would harmonize well with what you have. This feature could also extend to magazine images – snap a picture, and receive suggestions for similar items.

Now, envision generative AI replacing your current furniture with new options in the image. While browsing through the catalog, you'd see how each piece would appear in your space.

What if you could get interior design through a floor planner AI?

Imagine a future where you receive personalized interior design guidance through floor plan AI. You could upload an image of your room, sketch a floor plan yourself, or use your phone to scan the space. The AI would then analyze this input and offer tailored suggestions for your room.

It wouldn't just provide a standard list of recommendations; instead, it would adapt its advice based on the room's purpose and your needs. For instance, for a kid's room, the suggestions might focus on play areas and storage solutions, while for a lounge, advice might center around comfortable seating and lighting. If you're seeking to create a dedicated TV area, the AI could describe how to set this up in a way that suits your space.

The AI could integrate product suggestions seamlessly into this advice. For a company like Burrow, this means showing furniture options that fit the recommendations. This all happens within a user interface designed to feel like a natural extension of reading, with product suggestions woven into the narrative.

The AI could also generate a visual floor plan based on your room and overlay it with suggestions, giving you a concrete idea of where each item could go. This interactive and personalized approach could mimic the advice you might get in the stores or from an interior designer.

What if the site and copy changed seamlessly to fit you and your needs?

The AI could learn the brand's communication style and utilize it to craft product descriptions that are tailored to your searches and preferences. Instead of static pages that appear identical for everyone, you'd encounter a dynamically arranged version, featuring products and information precisely selected based on the site's understanding of you. It's a personalized shopping experience where the website functions almost like a personal shopper, describing the aspects of products that are most relevant to you and your needs.

While not strictly AI related, the page structure could tailor itself to you, changing its layout and the copy to match your interests and needs? Imagine the page structure adapting in real-time, highlighting dimensions if that's what you care about, or pushing up ratings if that's what matters the most to you.

What if you could easily see and tweak what the AI collects about your preferences?

If you purchased a piece of furniture with hairpin legs two months ago, the AI takes note and assumes that's your style preference. Alternatively, perhaps you recently selected a walnut finish for a shelf, prompting the AI to update your profile to favor walnut finishes. If the AI makes a mistake, such as assuming you love walnut when you don't, you could correct it. You might even have the option to dismiss certain preferences or deactivate the tracking feature entirely.

The goal is to provide you with control and transparency over the information collected. This way, you're empowered to manage your data and shape your shopping experience, resulting in a more comfortable and personalized journey.

What if, in addition to receiving personalized product recommendations, you could easily understand the "why" behind them? As part of a transparent AI-powered shopping experience, each recommendation could be accompanied by an explanation. This level of transparency would enable you to see beyond the fact that items are being suggested and understand their relevance based on your activity and preferences.

What if the site could visualize products based on your needs and preferences?

What if you could effortlessly customize how you view products, going beyond standard list or grid views? You could personalize your browsing experience based on what matters most to you.

For instance, you could instruct the site, "Show these sofas with fabric details," and it would emphasize materials and durability. Or you could say, "Show these in an industrial-style apartment setting," and the site would display the products within a virtual room in that style.
Perhaps you want to explore how modular furniture pieces can be configured. If customer reviews are essential to you, you could choose to "Show reviews in context," and the site would integrate user testimonials and ratings right next to the products.

This level of customization empowers you, allowing you to dictate how product information is displayed based on what you find most useful—whether it's material durability, how items fit into different interior styles, or the social proof from customer reviews. It's about tailoring the shopping experience to be as relevant and informative as possible for each individual user.

What if you could talk to a Burrow AI avatar shop assistant?

Imagine being greeted by an AI avatar as you enter the website, a persona designed to embody the brand's ethos and customer service values. This digital assistant is available to engage with you in real-time, using natural language processing to comprehend your requests and respond appropriately. For instance, you might say, "I'm looking for a neutral two-seater sofa," and the AI would not only display relevant products but also pose follow-up questions to refine the options, similar to a human store assistant. It might inquire about your preferred materials or budget to further tailor the search results.

The experience would resemble conversing with a salesperson, enriched by the AI's immediate access to extensive product data and customer preferences. Such AI avatars have the potential to transform the impersonal nature of online shopping into a lively and interactive encounter, rivaling the guidance and help you expect from a physical store.

And finally, what if we helped your business unleash the power of Generative AI?

We are amazed by the seemingly limitless possibilities emerging as AI inspires innovators of all kinds. And as we explore what the future might look like in this AI world, we look forward to collaborating with businesses to meet their users' needs and optimize their product experiences. We have more than a decade of experience of working with both startups and unicorns to discover the opportunities of tomorrow and seize the ones of today.

Learn more about our product offering — and please don't stay a stranger, say hello@hellomonday.com.