Impostor Blocker

A B2B SaaS tool that improved website phishing detection to 96% using computer vision.

Why was this software made?

This software was made to help enterprises detect phishing websites before their customers.

The problem solved:

Enable enterprises to track phishing websites so that they can cut financial losses and regain brand trust.

Design Process

Product Research
Due to this project being a part of the Information School at UW, we were allotted two weeks to conduct our research. We wanted to understand the pain points of businesses in detecting phishing websites and also wanted to know how they dealt with it. We also wanted to look at other companies trying to solve the same problems and understand their solution.
User Interviews

Semi-structured user interviews were designed to understand the end to end process an enterprise takes from detecting a phishing source to taking it down, and note the pain points.

Take-aways:
  • Usually, the companies were notified about phishing websites by their customers
  • There is no direct way to check whether a website is a phish.
  • Analysis of whether a website is phish or not is often inaccurate and unreliable.
  • There are no/limited action items on how to take down a live phish website.
Competitive Analysis
Objective

Explore what kind of features have been implemented already and to assess the scope of improvement for those features.

    Design Guidelines
    After cumulating the results of our research, we came up with certain design guidelines that would help guide our design process.

    01

    Accurate

    The results should be accurate and reliable.
    The client should also be understand the results

    02

    Easy access

    Detecting phishing websites should be a one-step process for users

    03

    Secure

    Users should know & understand the process we take to detect phishing websites
    Ideation

    In our ideation phase, we focused on integrating our design guidelines into the first version of our user dashboard. We started with multiple sessions of brainstorming to figure out the features we can include. Given the strict timeframe, we started adjusting the scope of our features through user testing creating multiple iterations.

    Idea 1 :


    The feedback from this accordion was.. pretty bad. Even though it felt like a dashboard, there were not many elements (visualizations, tables, filters) that reflected the same. The label names were confusing to 7 out of the 10 people we talked to, and the whole dashboard was "hard to maneuver".


    So we took a step back and started from scratch. We dove deep into the feedback we received by re-examining some of the interview notes. We also looked at 100s of B2B dashboards fro inspiration so as to make sure we could identify some key components that were easy to understand.

    Idea 2 : Information Architecture


    This time around, we decided to narrow down to the goals of this project. Phishing detection and increasing proactive action items to fight phishing were evaluated as central themes to the design solution. After finalizing this new direction, an information architecture was made that detailed how the functions were spread across the dashboard. This helped bring the team on the same page for our design solution would entail.

    Idea 2 : Task Flows
    Task 1: Check if a website is a phish
    Task 2: Evaluate a particular previous URL search item
    Task 3: Find monitoring sites
    Idea 2 : Wireframes


    Once we identified key components from the task flows, we started working on wire framing. Creating wireframes helped get a better feel of the dashboard and ideate more ways the user could use it.

    Visual Language

    High-Fidelity Prototypes

    01

    Log In

    02

    Home

    03

    Emergency

    04

    Fraud History

    05

    Search

    06

    Search Loading

    07

    Search Results

    08

    Manage Data

    Next Steps

    At par with some of the suggestions we received, the next step includes introducing an onboarding experience for users. Having a modal pop up or a help chat box would help the users navigate the dashboard comfortably and reach its full potential.