How Serial 1 leveraged visual intelligence to increase web conversions

eBike manufacturer Serial 1 had lots of gorgeous product and lifestyle imagery at their disposal, but they didn’t know what specific shots would resonate with their customers. They also had limited time to wait for results from a traditional A/B testing approach. 

To accelerate their decision-making process, they implemented Vizit, a visual intelligence platform that uses AI and machine learning to identify the components of visual content and quantify its effectiveness. The platform’s AI-driven technology helped Serial 1 determine what images resonated most with consumers.

Serial 1, a standalone company that was initially launched from within Harley-Davidson, manufactures premium eBicycles. The bikes have been in development for years and were recently made available to consumers in the spring of 2021. Serial 1 wanted to quickly understand what visual content resonated with consumers, particularly in the current post-pandemic environment.

“As we began marketing our eBikes, we weren’t sure which specific shots would appeal most to our customers and potential customers,” said Rex Hamilton, Serial 1’s VP of Finance and Corporate Operations. “Do they like to see the eBikes in action? Do they like to see close-ups of the sleek design that shows off our integrated internal wiring, for example? Or would they rather see wide-angle shots of the entire eBike on display? Vizit offered us a way to use technology to figure this out rather than simply guessing or testing.”

Testing images at scale

Serial 1 knew that manual A/B testing would potentially take weeks to finalize versus using an AI-based solution like Vizit. The platform uses AI algorithms to assess and prioritize visual content by analyzing attributes such as surrounding scenery and environment, the presence of human models, and positioning of the product within the image. Selecting and onboarding the tool was a fast process.

“Several members of the Serial leadership team were involved in the buying process,” said Hamilton. “The decision was extremely easy when we started comparing the two options of manual A/B testing versus automation, especially when we were on a relatively tight timeframe.”

Serial 1 kicked off the initiative by outlining why they planned to use Vizit and how it could be beneficial to the brand. “We had completed a few photoshoots earlier in the year and had great images to select from, but other than that there was little to no effort to kick off the project,” said Hamilton.

Using AI for better decision-making

Using AI to improve efficiency and productivity is a key trend in 2021, according to PwC, who surveyed over 1000 executives — including hundreds of CEOs — about how they plan to use (or continue using) AI in their business. 

According to the PwC survey, AI is being used for everything from workforce planning to demand projection to simulation modeling. The use of AI is intersecting with advances in another key area—computer vision technology. Computer vision enables machines to “see” visual content. When combined with AI, it can automatically identify and classify visual elements such as objects and people within that content. This enables the technology to do what people once did manually—interpret and understand visual images. 

Vizit’s visual intelligence technology uses machine learning to analyze millions of images and measure them for effectiveness. The technology then applies a visual effectiveness score to help users understand what designs and images are more likely to resonate with an audience. 

“Vizit allowed us to use data to make informed decisions much more quickly. Our goal from the start was to connect with our customers on a more personal level. Subsequently, we believed that this would drive an increased conversion rate and sales,” Hamilton said.

Download Enterprise Digital Asset Management Platforms: A Marketer’s Guide

A vision of product accessibility 

After Serial 1 implemented the visual content recommended by Vizit’s software, their conversion rates increased by 98%. In addition, online shoppers spent 25% more time on their website, and their research and testing time decreased by 99%.

“One of the most prominent trends that we saw during the project was the importance of making the product feel accessible,” said Hamilton. “For example, automotive brands will sometimes show a vehicle traversing massive sand-dunes or plunging through a deep river. While this might happen to one or two customers, it is not how the majority of these vehicles will be used. We found this to be true with our eBikes as well.”

When Serial 1 showed their premium eBikes being utilized in the most likely use cases, they learned this resonated with customers the most. It helped the customer see themselves using Serial 1 eBikes in a similar manner. 

Using visual intelligence to inform future strategy

Serial 1 plans to use results from the Vizit campaigns to help inform future photoshoots and their overall asset creation process. 

“We know that shooting the product in some faraway tropical beach probably won’t connect as well as a couple riding their eBikes to their local farmers market or to a social gathering,” said Hamilton. “It doesn’t mean that there isn’t a need for those awe-inspiring photo shoots, it just means you have a better understanding of where to deploy said assets.”

Technology like Vizit’s visual intelligence performance platform can help brands refine how and what they present to customers. It can also help them make better decisions more quickly. “Visual intelligence is a great tool for brands who aim to disrupt an established market and want to use technology and human intuition to their advantage,” said Hamilton. “Thanks to technology, customers have access to products around the globe. Our goal is to make them feel more at ease when shopping on our website and to see themselves on our premium eBikes.” 

All images provided by Serial 1.