Element

Accelerate content consumers and creators' online sensemaking process.

Project Overview

Project Overview

Sensemaking tool for vast information online

Whether you’re researching your next blog post, lit review, or home remodel, the hard part isn’t finding information on the internet—it’s putting it all together. That’s why we built Element: an all-in-one place research tool that lives with you wherever you are on the web and helps you collect and organize your research.

Design Solution

Design Solution

01

A Single Point of Collection

Element collects your online sensemaking artifacts all in one place chronologically with filters and tags.

02

Flexible Templates to Reduce Cognitive Load

Element provides flexible templates for users to digest information while organizing it. Element helps you to connect the dots!

03

Annotate and Produce Artifacts

Ability to capture notes across all touch points of the content creation process. The annotation capacity mimics the current behavior of our target users, with the additional benefit of streamlining all the annotation in one place with a consistent workflow.

01

Background Research

Since we were aiming at adoption for the tool, we quickly chose a domain where there are potentially many users who conduct intensive research. We started researching users' needs to collect and synthesize information while shopping online to see if a sensemaking tool could help with it.

Initial research resulted in better understanding of the 'sensemaking process', a concept stemmed from the professor's research on how people make sense online. We identified the process specifically for online shoppers and their key pain points.


We conducted two rounds of research: 11 interviews focusing on tools and general processes and 18 directed storytelling sessions about the detailed process for a recent electronic purchase. Through these data points, we were able to get a holistic understanding of the user's journey and specific research behaviors.

After affinitizing these findings, we developed three behavioral models that depicts research processes and pain points across the participants. We are then able to zoom into the friction points and start to ideate for solutions that would help alleviate them.

Through the affinity diagram and behavior models, we learned that people's research behavior and time spent is influenced by: the importance of the purchase, prior knowledge about the domain, time available, and passion for the topic. There are also pain points around research along the way:

1. When people are not familiar with a product or domain, it is hard to translate intention into a search.
2. People rely on trusted sources like credible websites or experts to gather information quickly, but validation of sources is not always easy.
3. Contextualization of how the product fits into one’s life is a key theme throughout the research process, however, such information is not always available online.

02

Understand Users

Based on the pain points we have identified, we started our ideation process and came up with more than 30 ideas to address with those friction points. We ended up doing two rounds of testing with storyboards and solution sketches with 18 participants to see if users would respond to solutions that may help them speed up the process.

Contrary to our expectation that feedback from speed-dating sessions will point us to a clear direction, the results actually pushed us to think from a different perspective. Although the participants acknowledge the pain points we identified and some solutions gained traction, the overall takeaway was that majority of people would not spend a significant amount of time and effort to research for their purchases and thus, do not need a designated tool to help with that.

The linear model is a straightforward process characterized by a single round of research, followed by contextualization (how does this product fit into my life), then a decision & potential subsequent purchase.

The iterative process is a bit more complicated: information gathering and contextualization happens more than once in a looping pattern. Eventually, this leads to a decision and purchase, but this process may yet be repeated again if the purchase turns out to be unsatisfactory.

There was also a shorter, condensed model identified: the user goes into the shopping process already having contextualization, which abbreviates the process.

03

Iterative Design Process

Based on the design principles, we prototyped dozens of solutions and evaluated them with the users to extract highlight features and iterated based on user feedback at every stage.

Because we wanted our tool to both fit in and help streamline the workflow for content creators, we sought their thoughts and feedback at every stage in the design process. We started with parallel prototyping, sketching out different ways we could address the pain points around collection, contextualization, and synthesis we learned about from interview with 10 content creators.

Then we tested those point solutions with our target user or analogues users to arrive at a proposed workflow highlighting key features that would help them at each step of the process. We conducted cognitive walkthroughs with 4 more content creators, asking them to think-aloud when engaging with our high-fidelity design prototype and eliciting feedback on how this workflow is similar or different from their current processes.

These cognitive walkthrough sessions helped us to validate our intuitions, inform design iterations, and ensure that we were building a tool that fits with their mental model and would actually serve their needs. After several rounds of iterations, we developed the following features in our MVP.

04

Design Decisions

The key success metrics for the design of clips is if the information provided on each clip is sufficient and appropriate for users to identify its content and intended purpose of use after collecting.

All in One Collection Place

After two iterations, our final design for the clips aligns with the mental model of most content creators we tested with. Namely, all clips have source and media types indicated along with the annotations taken when clipped, and the ability to preview.

Additionally, multimedia clips would be visually represented, text-based links would stress the title and user-generated highlights, and social media links would include account handle. The participants also find the clips collection to be more well-organized and easier to navigate than their current tools (e.g. browser tabs and list of links).



Infinite Canvas Board

Our second critical feature is the infinite canvas board which supports organization and synthesis at various stages of the sensemaking processes. This is where we allow content creators to begin creating structures and organizing the information they've collected on a given topic.

Our key success metric for this features is the number of content creators who can fit their current content within our allowed organizational structure. We started with a single level of grouping and a mindmap-like organizational structure, but found it to be unscalable in supporting a more complicated thought process.

After two rounds of iteration, 3 out of the 4 content creators we tested with is able to fit their current content structure into the hierarchy we designed (Topic Board - Header - Groups of Clips). The other content creator simply does not post enough to make organizational structure a critical issue for her.



Ubiquitous Annotation

Our third and last critical feature is the ability to capture notes across all touch points of the content creation process. The annotation capacity mimics the current behavior of our target users, with the additional benefit of streamlining all the annotation in one place with a consistent workflow.

Our key success metric for this features is the number of content creators who can fit their current content within our allowed organizational structure. We started with a single level of grouping and a mindmap-like organizational structure, but found it to be unscalable in supporting a more complicated thought process.

After two rounds of iteration, 3 out of the 4 content creators we tested with is able to fit their current content structure into the hierarchy we designed (Topic Board - Header - Groups of Clips). The other content creator simply does not post enough to make organizational structure a critical issue for her.

MORE PROJECTS