The Gist

  • Sneak peek: Adobe Sneaks is an event that showcases select projects across various areas that push the boundaries of what is possible with Adobe's creative tools and technologies.
  • Only half survive: At this year's Adobe Sneaks, the company showcased tech with the goal of securing their project's place in Adobe labs, but less than half of the projects presented typically end up making it into the world as future products.

Each year, Adobe Sneaks offers a sneak peek into the potential future of Adobe's products and services. Over the years, the highly anticipated Sneaks has featured innovative projects across various areas such as artificial intelligence, augmented reality, virtual reality, and more, that aim to push the boundaries of what is possible with Adobe's creative tools and technologies. From the hundreds of proposals they receive, Adobe selects just a few to showcase each year during Adobe Summit.

Last week’s Summit marked the company's first in-person event since 2019, and a lucky seven got to share their proposals, with the goal of securing their project's place in Adobe labs.

Still, even after making it this far, less than half of the projects presented will end up making it into Adobe’s labs and into the world as future products for marketers and customer experience professionals. Eric Matisoff, principal evangelist, analytics & data science at Adobe, hosted this year’s event, along with celebrity host, comedian, actress and author, Tig Notaro.

Matisoff started off by reminding the audience that these aren’t professional presenters trained in public speaking — they are engineers, research scientists, data scientists and UX designers — so, sometimes things tend go a little sideways. And with the array of not necessarily ready-for-primetime tech, a couple things did.

However, despite a few technical glitches, nothing burned down, and the show went on.

Without further adieu, here are the hottest highlights from Adobe Sneaks 2023:

Predict the Customer Journey: Project PathWise

The process of delivering personalized experiences to customers across various touchpoints is known as journey orchestration and involves analyzing data from multiple sources to understand customer behavior and preferences and then using this information to create a seamless customer journey. But the problem with journeys? You never quite know how well they're going to perform.

Nikaash Puri, a research scientist in the media and data science research lab at Adobe Center in Noida, India, started out the evening by presenting Project PathWise, powered by Adobe Journey Optimizer and Adobe Sensei GenAI. According to Puri, Project PathWise will show users approximately how well a journey will do — ahead of time — by leveraging machine learning to replicate traffic at each customer touchpoint in the journey. Then Sensei GenAI is utilized to predict users' reactions.

“I can't predict everything, but I can predict the future for this journey,” Puri said during his demo.

By selecting the “simulation” option users can replicate pathways using data points from past journeys to evaluate how users are going to flow through a journey (how well it will work) before launch, including metrics like customer lifetime value (CLV) and email deliverability. The tech also suggests actionable ways to improve the journey. For example, with 85% confidence in a 3% lift in CLV, PathWise will suggest how to raise it to 6%.

Related Article: Adobe Unveils Generative AI Tech and Major Partnerships at 2023 Summit

Targeted Marketing Videos in No Time: Project Custom Clips

For those among us who are not professional video editors — or simply don’t have the time it takes to produce quality video content — Stefano Petrangeli, a senior research scientist in the REAL Lab at Adobe Research, said his project, Custom Clips, offers a way to quickly create high impact videos, without the editing experience.

“As marketers, we know how hard it is to make impactful videos that can resonate well with specific channels and audience,” Petrangeli said. “And what I wanted to show you how Project Custom Clips can make this process very, very simple for us in just a few clicks.”

Using marketing materials within Adobe Journey Optimizer, he shared two video campaigns, targeting two different audiences: city lovers and outdoor enthusiasts. Using AI, he said, not only can the tech predict whether a video will be impactful, but it can also craft a new optimized version of the video. Further, it can quickly trim a 60-second commercial into 15-second spot that is solely focused on the desired audience — say outdoor enthusiasts — and remove all video content related to the city, etc.

Petrangeli said Custom Clips can also modify and edit new videos automatically based on analytics and past performance data to help advertisers align videos with their target audience and publishing platform.

One Image, Unlimited Marketing Options: Project Unlimited Options

As a machine learning senior manager at Adobe, Deepak Pai said he understands the enormous effort and resources marketing teams invest into personalizing campaigns and content. By integrating Limitless Options powered by Adobe Firefly's generative AI technology (that provides a boost to content velocity and helps target the right audience) and Adobe Journey Optimizer, Pai said marketers can create, improve and deliver customized content at scale, quickly.

During his presentation, Pai demonstrated how it’s possible to choose from several segments to create new, accurate and visually appealing content. By incorporating all a brand’s specific elements from its landing page, like a mascot, colors and products, the tech blends it with elements likely to resonate with a specific audience segment and generates an appropriate image. From there, additional images can be easily produced that appeal to different audiences and brands — and the options are “limitless.”

Learning Opportunities

Editor's note: We asked ChatGPT-4 which innovation it liked most, and it said, "Out of the various innovations presented in the article, I find Project Unlimited Options particularly interesting. As a machine learning senior manager at Adobe, Deepak Pai understands the significant effort and resources marketing teams invest in personalizing campaigns and content. By integrating Limitless Options powered by Adobe Firefly's generative AI technology and Adobe Journey Optimizer, Pai said marketers can create, improve, and deliver customized content at scale, quickly. ... The ability to generate customized content quickly and at scale could revolutionize the way marketers engage with their audiences and create personalized experiences."

I See Your ... Project True Colors, for Better Customer Experiences

Dej Mejia, a senior UX designer at Adobe, says there’s no need for a personal stylist with True Colors. Because, whether you already know which colors suite you best, or not, this AI-powered tool aims to make it easier to find apparel and color palettes that suit each shopper’s personal looks.

After analyzing a single, shopper-submitted image, True Colors uses AI to analyze a user’s personal coloring, undertones and facial features to arrive at the colors that are going to work best for them across the color spectrum, based on extensive color theory research. Then, the tool filters website inventories to show which products are available in these colors from retailers. With its predictive abilities, the tech reduces abandonment, increases conversion and provides more satisfying customer experience that results in less product returns, according to Mejia.

“Overall this experience really reduces the likelihood of return products because shoppers are buying colors that are going to work really well for their coloring,” Mejia said.

Related Article: What to Expect at Adobe Summit Digital Experience Conference 2023

Whittle Down Those Ecommerce Search Results: Project Fast Filtered

Everyone loves spending hours scouring pages and pages of product to find just the right one — said nobody ever. So, Ronald Oribio, a senior software designer at Adobe Commerce, created Fast Filtered, incorporating Adobe Sensei, that works to whittle down search results, from 6,000 to 60 to six, in a matter of moments.

Instead of reading the fine print for each option or figuring out complicated filtering settings while shopping, a custom dimension filter allows users to search for specific must-haves like height, width, style, building material and more. Users can also eliminate products they don’t want immediately from their view.

Smart, Personalized Content: Project Segment Smarts

Jennifer Qian, a solutions consultant at Adobe, believes that with the right tools, massive amounts of unused data and assets like images, videos, articles, blog posts, and more, can be leveraged for analytics that drive content.

“With the personalization explosion marketers are pressured to create more and more content,” Qian said. “The challenge now becomes, what content to use.”

With Segment Smarts, Qian said users have the ability to natively connect audiences and content in Adobe Experience Manager and use Adobe Sensei-powered machine learning to recommend the best content for any audience. Marketers are also able to take full control over the model and define exactly how the content is recommended.

Your Side, My Side: Project Side by Side for Marketing Collaboration

Adobe Senior Software Engineer Richie Preece introduced Project Side by Side to enable marketing, product and data teams the ability to collaborate on the same datasets. The "Side by Side" feature creates a collaborative analytics experience that facilitates cross-functional team collaboration and enables Adobe Analytics access.

According to Preece, it empowers teams across the enterprise to view identical datasets simultaneously from multiple devices and facilitates synchronized report updates. Moreover, the project offers transparency in terms of who is accessing the report and/or viewing the project. Multiple cursors appear in real-time alongside a chat box, which enables quick and efficient communication, thereby reducing the time and resources required for one-off requests.

TBD: The Big Winners 

Adobe's Kevin Fu said there is no magical formula to selecting what will work as an actual product, but the common thread is that it needs to solve a pressing pain point. In 2021, Adobe Sneaks presented something called Project Catchy Content.

“The underlying technology uses AI to measure content at the attribute level (colors, copy, objects) to determine what resonates for different target audiences,” Fu said. “Last week, we announced the capabilities in Adobe Experience Manager, available as an early beta.”

Going further back, anomaly detection (to identify unusual events in a brand’s data, so that teams can pinpoint broken customer experiences or uncover opportunities) was also a Summit Sneak before it became a feature in Adobe Analytics.

As far as “who wins,” each presenter asks the audience to show support for their individual venture by tweeting their product’s hashtag following their demo, but according to Adobe, the audience votes are not something they release publicly, or base their entire decision on.

“We don’t want to pick favorites,” said Fu. “Each presenter puts a lot of time and effort into their Sneaks and the audience votes are one of several signals we leverage to fine tune these projects.”