Hive Consumer
ARTIFICIAL INTELLIGENCE FOR JOB SEEKERS
Overview
OBJECTIVE
To create a web and mobile platform that gives candidates more power over his / her career and job application journey.
ABOUT
Hive is a better way of matching people and jobs. It is a product for consumers to manage job searching to match candidates with positions using artificial intelligence. AI is utilized to match the candidate with the job beyond the words on a resume and understands the meaning behind the words in a resume and job post.
FOCUS
Focused on SaaS platform that integrates with competitive products to provide a more enhanced service and work with legacy products instead of trying to compete with them.
Unlike LinkedIn’s product which caters to recruiters, Hive also provides solutions for the job seeker.
OBSTACLES
Multi-features, large scale and familiarity with legacy products.
Establishing viability in a saturated $8+ B recruiting market.
6 month timeline including modeling and developing algorithms and models for a large industry.
The project was reviewed at 6 months and extended for another 6 months.
VIABILITY MODEL
Job searching tool used to search and manage jobs and applications for job seekers.. Integrates with - LinkedIn, Indeed, Glassdoor, Dice and other products.
Artificial Intelligence used in matching candidates to job posts, writing resumes and other functions. Can view and apply to job posts from all integrated systems at
PROOF OF VIABILITY
Hive was thought to be an enterprise product focusing on recruiting. After a viability study on a competitive model product - we found we needed job candidates - both active and passive for recruiters to use the site. We Pivoted our model to integrate with LinkedIn and other systems for the consumer / job seeker.
SOLUTION
FINAL DESIGN DASHBOARD
TEAM
CEO Stakeholder and product owner
Data Scientist
Backend Developer
Front End Developer
Nathalie Perez - UX / UI Designer
UX Researcher
UX Writer
TOOLS
Adobe XD
Figma
UserTesting.com
SurveyMonkey
METHODOLOGY
Agile Scrum with Lean UX methodologies.
DURATION
Phase Two
6 Months
STRUCTURE OF THIS PROJECT
Throughout the 6 months of Phase 2 (Consumer) we worked closely with the AI Data Scientists & User Researcher to create artifacts and requirements for the AI workflow and process relevant for job seekers. We worked, tested and reiterated at a very quick pace.
MONTH 1
DISCOVERY AND BUSINESS RESEARCH
In order to compete with the current market we needed to be compelling and different from LinkedIn. It’s social aspect is their biggest advantage and helps attracting passive and active candidates altogether as well as companies and recruiters.
WHAT LINKEDIN IS DOING WRONG?
Unwanted approaches from irrelevant recruiters or people wanting to join your network.
There is no way of verifying the skills endorsements.
Discovering which settings work to restrict fraudulent accounts from accessing user information is an uphill task.
There is no verification process in place to make sure that user profiles are genuine and authentic.
Irrelevant ads for users profession, skills and knowledge
OTHER COMPETITORS
MONTH 2
USER RESEARCH
IN SPECIFIC, WE OBSERVED JOB SEEKERS:
Log on and create an account on LinkedIn and other products
Create a candidate Profile
Search for a relevant job post
Apply to a job post
Contact the employer
Schedule phone interviews
HOW MIGHT WE ...
Define the primary persona?
Define needs of the job seeker?
PERSONA BUILDING
Wrote a survey of 10 questions about LI and sent it out to a list of 500 active job searchers and people employed within two years of the survey with the help of an agency and a focused database.
From these 2 data sets we came up with our personas Sarah, David and Jay our primary persona.
FINDINGS: USER RESEARCH
PAIN POINTS
Receiving phone calls and emails from irrelevant opportunities and recruiters.
Not knowing their status in the hiring process.
Keyword matching inaccuracy while searching.
Unrealistic requirements for the position.
NEEDS
Finding the perfect job, fast.
Improve salary, position and advance his career.
Easy commute.
Effective communication with hiring personnel.
ACTION:
We found the persona of JAY held the most points that were common with all the personas we developed and all the real people we observed.
JAY has learned through a learning curve where apply to jobs. Some sites cater to offshore recruiters that bombard him with irrelevant jobs. He is experienced in all levels of interviewing, recognizing how to talk and be presentable to a recruiter, a hiring manager and a stakeholder at a company. He well versed in all different career sites. He values company culture, traveling and flexibility to work remotely. His experience working at startups makes him knowledgeable in the volatile job market in the tech industry.
MONTH 3
IDEATION, SKETCHING, WIREFRAMING & SPRINTS
Hive for Job Seekers needed to have a way of effectively showing results beyond keywords. Most sites show irrelevant jobs and very little show how Applicant Tracking Systems parse resumes in order to pre select them for the recruiter. It also needed more support from the Artificial Intelligence in other tasks like creating candidate profiles and communicating with recruiters in an adequate language. It needed to integrate into current best practices already familiar with job seekers.
From previous research my team narrowed down the Job Seeker needs to focus on:
More relevant job posts and less searching through hundreds of results using Artificial Intelligence.
Ability to track interviews, interview process and receive feedback for profile improvement.
Redesigning the job search process using Artificial Intelligence
We needed Hive to be both competitive to LinkedIn as well as an appealing add-on enhancement to LinkedIn. We kept referring back to our competitive landscape and Engagement analysis to guide to our design of Hive Enterprise.
Taking in both business and user data, we first did a white board exercise categorizing the most important things we needed to fix about this process for Jay the job seeker. Here’s a sample of some of our ideation sessions:
GALLERY OF SKETCHES
FOCUSED SPRINTS (3, 2-WEEK SPRINTS)
On-boarding
Business requirements (Defining search engine, data sets and AI requirements):
Defining AI frameworks, visuals, messaging, and overall design patterns that make it more usable, more trustworthy, and better aligned with job seekers tasks and goals. Defining our data decision making system with coverage, salience, transparency, malleability and instrumentality for job seeker’s success. Uncover crucial patterns and edge cases.
Since AI is governed by complex relationships between data, computation, humans, and communities, we created an AI Persona to make communication effective and seamless.
Optimization of steps for the Job Seeker
Define rules that lead to patterns in recruiter behaviors and activities.
Set features that validate consensus and verification activities from job seekers and employers. Building templates that make job seeker’s work easier and faster.
Privacy frameworks
Defining what forms of data carry risks and regulations from job seekers.
Also defining our AI-driven identification, profiling, and automated decision-making that leads to fair anti-discriminatory, or unbiased outcomes. Implementing our AI is aligned with requirements that respect, promote, and protect international human rights standards.
Dashboard for the Job Seeker
Up to the minute updated info.
Job display: Most efficient way to improve visibility of large amounts of data to. How they would find jobs, in large quantities while skimming through a large amount of search data.
RESULTS OF SPRINTS
User Flows
Preliminary requirements for job searching
Site Map
Consumer User Flow After Sprint
Consumer Site Map After Sprint
MONTH 4.5
FINAL DESIGNS AND TESTING
After testing the comps from the Sprint, we realized testing failed because the Consumer / Job Seeker dashboard was too feature heavy. It also didn’t provide any novelty from our competitors. We did some more ideation and modified the designs and retested (images 1, 2, 3 below):
Job Match Display - was modified to display jobs as soon as candidate enters dashboard.
We eliminated a section of “Recommended jobs” that were being displayed in pages where they weren’t relevant. (Matches - image 4 below.)
Job Post - menu contains subcategories from integrations with Glassdoor and Jobscan to display company profile, reviews and how job seeker’s resume scores in an Applicant Tracking System.
Integrated a calendar to synch with user’s preferred system (Google Calendar or Calendly).
Initial Wireframes
Screens
Design: Job Results with direct access to job and other integrations.
Design: Job Scan integration which allows users to see how their resume is paired with the selected job according to keywords
(method used by most ATS systems).
Calendar integration that syncs with Google Calendar
MONTH 6
FINAL TESTING & HANDOFF TO DEVELOPERS
We designed and built-out the whole product, tested and handed off the files to the development team.
What the Developers received:
Annotated screens
Glassdoor, Jobscan and other implementation recommendations
AI requirements, justification, logic patterns and organizational groups of data
As the product was built, we tested it with job seekers and refined usability and design. We utilized moderated in-person and remote testing with both external and internal (inside a company) recruiters.
CONCLUSION & SUMMATION
+ FINAL THOUGHTS
What I learned from working on this project
I discovered the differences and possibilities of creating a product for consumer vs enterprise. The consumer interface was unique and required a lot of research and testing in order to create something viable in a saturated market that doesn’t really listen to the consumer needs because a profitable business model is mostly focused on enterprise.
What you would I have done differently if you knew then what I know now about the project
If we would’ve started with a more simplified approach without the assumption of a feature heavy dashboard, we would’ve advanced further on creating more necessary features. I also would have explored the possibility of customizing the integrations suggested and exploring what combinations of integrations were more suitable for the consumer.
What you would I have done next had you continued with the project.
I would’ve worked on the social aspect of this job market app to attract passive candidates. I would’ve also scaled by targeting candidates from specific industries, one at a time, trying to create a consolidated source of candidates for certain sectors.