Building authentic shopping experiences through interactive video

Weview (Techstars SAP.io 2018) was a product review video platform that helped shops and brands engage their customers by embedding video reviews on e-commerce websites. See our demo day pitch.

Update: Weview was acquired by DemoUp in 2020.

Featured projects:

Video Editor

Most video creators spent more than five minutes to upload a simple 10 minutes video on youtube. Imagine the time you spend on adding all information required and waiting until your video is processed. And then, you have to wait an additional 8 hours until your video is searchable. There was no doubt that it should be easier and faster for creatives to share their work with the world.

What we had to improve

Weview promised a quick and easy upload process for creators. In reality, it took up to 30 minutes to upload a video, around 20 minutes to fill in all the data, and sometimes up to an hour for the processing.
Now, you can imagine how motivated users felt every time they visited our platform knowing how much time it was going to take. Because they were frustrated they started leaving out important information or simply stopped using the platform altogether.
We had to fix this problem.

Product videos require a lot more metadata to be useful to viewers. Therefore the editing process proved much more challenging than for regular videos.

Research is formalized curiosity

We had two sessions with our users, one for interviewing them based on quantitative data we gathered and analyzed, and the other for watching them use our platform.
Once we interviewed and watched the most active users on our platform, we redefined personas to structure the user journey mapping which provide an in-depth understanding of our problem.
Our key finding was that some input fields were unclear and incorrectly labeled, and that the product tagging in- video was confusing and overly complicated.

The improved video editor greatly simplified the experience for our users by hiding complexity and allowing them to fill out more content in parallel.

Users talk to Alex the Alpacabot via a messenger chat interface that seamlessly transitions into custom webviews that display highly relevant information and search functions

Design. Test. Repeat

Based on our research and definition of personas, journey mapping, user flow, and the IA we started with whiteboard sketches, moving to paper once we agreed on two versions. We could cut down 2⁄3 of input fields and get rid of the step where creators had to make the product clickable. Once we had our first paper prototypes we tested them with non-users to see, if the process was understandable for everyone. For the very first time we got back to current users, to receive feedback of a low fidelity wireframes version. You could tell that they loved the idea of filling out less information and having most inputs already filled out, when uploading with their YouTube account. With this positive feedback, we created the high fidelity version and also improved the mobile web app for a smoother experience.

And we didn't stop there

Our uploads spiked shortly after the release. However, user centric design is not a linear journey but a cycle. Our next steps included analyzing how many creators get stuck in the upload and editing process and how many error messages pop up. We introduced the mean-time-to-publish as a business KPI and made it our goal to keep cutting it down to a few minutes.

We improved:
  • 4x more unique visits
  • 6x more uploads
  • 3x faster upload to publish time

Video player

While weview was serving videos on shop websites we also had our own video platform for users to browse review content. One of the main complaints was that the player was 'not easy to use on the phone', or 'not unsable'.
The main use case for our player was to show video reviews on e-commerce sites like Mediamarkt as captured in this animation.


I analyzed the data of user behaviour with tools i.e. Mixpanel, Hotjar, Google Analytics, Grafana, and user interviews. I found out that the controls of the player were too big. The video started with the maximum volume, and the ‘buy button’ was disturbing the experience. It was hard for users to change the volume or move to another timespan in the video.

Information Architecture

I created a list of issues users had while watching a video on YouTube, Vimeo, and our platform. I compared all elements used on all platforms and created a player that addresses all the issues.

Goals & Methodology

Goal: Users enjoying watching videos with our video player and having no problem in using it.
Methodology: Usability test consisted of 10 participants who had to complete a set of tasks.

User Testing

With high fidelity wireframes based on feedback gathered by testing lower fidelity wireframes, a clickable prototype was then created to further test and gain more accurate feedback.

On shop pages we would first show this thumbnail in order to 1) catch the users attention and 2) present multiple alternative reviews if present.

Tasks for the testers

Task 1: Click on a video & start watching
Task 2: Use all controls in the player
Task 3: Try to buy the product

Next Steps

1. Development for browsers and web app
2. Add tracking (e.g. upload completion, bounce rate,
completion time) to better understand user behavior and
to further help in optimizing the conversion funnel
3. Test & improve