Addressing the complex journey of career change with Career Compass

I am thrilled to share the release of Career Compass v0.1, a tool designed to help career changers navigate the maze of opportunities. Our objective was to provide personalised and relevant career recommendations, achieving a 26.2% conversion rate and 74% positive customer feedback.

Executive summary

  • Career Compass aims to revolutionise the way people approach career changes. In today"s dynamic job market, changing careers is almost an inevitability. This tool serves as an interactive guide that simplifies this complex process.

    People find themselves unsatisfied in their careers, unsure of their opens. Using LLMs to support Career Changers to Discover, Validate and Progress towards their next career. Unlike the transactional nature of most career quizzes and personality tests, our solution supports the candidates through their journey, building their confidence towards taking the plunge into their fulfilling new career.

    I led a SEEK Learning team to built a career discovery tool that uses OpenAI"s Large Language Model via their API and vector search to match the responses from a multiple choice quiz and a prompt to our detailed role page insights for career changers to discover new career opportunities. Through interviews and surveys we designed a solution that the recommendations were personal and relevant to candidates with 74% positive customer feedback and a 26.2% conversion rate.

Our team

    Product teamTheir role
    Jack RustSenior Product Manager
    Ahmed HakeemTechnical Delivery Lead
    Richard Simms (me)Principal Product Designer
  • We were a dynamic and agile product trio dedicated to accelerating innovation. In our discovery phase, we"re employing a rapid, data-driven approach to bring a new product to market.

    I, Richard, led the product discovery and design strategy to ensure our product met user needs while maintaining aesthetic excellence. As the Product Designer, I lead the ideation, concepts, and prototyping processes.

    Jack, our Product Manager, provided the strategic vision, aligning business goals with user requirements. He was responsible for articulating vision internally, and coordinating the execution met legal and Responsible AI requirements.

    Ahmed, our Engineer, transforms designs and plans into functional, scalable code. With an eye for detail and a knack for problem-solving, Ahmed ensures that the technical architecture aligns with the product"s long-term vision.

    Together, we leveraged our specialised skill sets to move swiftly, make data-informed decisions, and position ourselves for success to launch our new product.

Our approach

  • We employed Large Language Models (LLMs) to create personalised experiences, improving customer satisfaction and loyalty. LLMs have the potential to revolutionise the way businesses interact with their customers, by creating personalised experiences for improved customer satisfaction and loyalty. While there are well-identified challenges associated with using LLMs, techniques such as few-shot learning and connecting LLMs with embeddings can help mitigate these challenges, enabling businesses to create more personalised and efficient customer experiences. By leveraging these techniques, we improved customer satisfaction and loyalty, for those job seeking candidates who were looking to find a new career.Career competition

The Problem

  • Navigating the intricate maze of career possibilities

    People often became dissatisfied with their current careers for various reasons, such as hitting a perceived salary ceiling, experiencing burn-out, or due to changing lifestyle needs.

    1. Many sat in this situation for some time, open to new opportunities without actively pursuing them.
    2. Through speaking to friends and colleagues, they might have uncovered potential career directions that sparked their interest.
    3. They tried to evaluate their new options through further conversations, online research, and occasionally short online courses.
    4. Eventually, they were ready to take action, make a career change, or invest in education or qualifications.
    Changing careers is almost an inevitability in today"s dynamic job market. Statistics indicate that the average person will switch careers more than five times during their lifetime¹. Meanwhile, a staggering 78% of the workforce reports feeling disengaged with their current roles². Add to that the 40% of jobseekers who are actively contemplating a career switch³, and it"s evident we"re dealing with a pervasive issue. The challenge? Finding a new career that aligns with one"s unique combination of skills, passions, and goals.Information achitecture

The Solution

  • A compass for your career journey

    That"s where Career Compass steps in. Compass serves as an interactive guide that simplifies the complex process of career discovery. By answering a few career-related questions, users can quickly uncover job roles that resonate with their personal and professional aspirations. These insights aren"t just generic suggestions but tailored pathways drawn from a deep understanding of various industries and roles.

  • Information achitecture

The benefit

  • A compass for your career journey

    Career Compass isn"t just a one-off solution but a comprehensive platform designed to support you at every phase of your career journey. You start with introspection, identifying your skills, interests, and values. From there, our tool provides actionable advice to guide your first steps towards a new role. This all culminates in helping you muster the confidence to take the final plunge into your next career. By aligning your unique profile with relevant job opportunities, Career Compass increases engagement and satisfaction in your professional life.

  • Information achitecture
  • A level playing field with the power of AI

    It"s worth noting that while there may not be a “perfect“ career recommendation, the utilisation of large language models like chatGPT has democratised access to high-quality career advice. We"re excited about the synergies between our approaches and look forward to the transformational changes we can bring to the world of career transitions.

Design process

    I played a crucial and instrumental role in effectively facilitating and leading the design sprints, customer feedback sessions, and rapid iterations based on data-driven insights. This active involvement not only contributed significantly to achieving positive results but also had a direct and tangible impact on improving key metrics and enhancing the overall customer experience. Through my efforts, I was able to drive meaningful and impactful changes that resonated positively with both the organization"s goals and customer expectations.
  • How we built confidence

    I started to validate our assumptions with surveys on the site to better understand the needs of people who were looking to career change. I was able to broadly categorise the following themes:

    • Why are people were considering a career change
    • What steps have they taken towards making a career change
    • What challenges have they faced in their career change journey
    • What resources have they found helpful in their career change journey

    I started prototyping the solution ideas through Hotjar surveys on site to test the questions that we need to ask to provide relevant results.

    I created a wizard of oz experiment where I would convert the survey answers in to career recommendations, manually with ChatGPT. I then shared the results with participants in a customer interview, where I could capture the perception of the quality of results and gave them a PDF takeaway document.

    This led to the flow that was then used to shape the questions and the prompt that we went on to use in the product. This meant that we were able to test our assumptions of what a minimal amount of questions and iterate with customers very quickly.

    In parallel I was building out a visual prototype and working very closely with Ahmed on building the quiz and results. We knew that we needed to make the quiz playful and enjoyable to complete, so we invested more time of crafting a custom button style that felt delightful to use.

  • “The expression of my interest is more important as you"ve acknowledged my behalf. I"m grateful for what is to come and more, actually, task schedule for priority with table talk is much more reliable.“
    — Customer feedback


  • User feedback was systematically tracked and analysed using the HEART framework, which focuses on Happiness, Engagement, Adoption, Retention, and Task Success. For Happiness, we collected user satisfaction data through surveys and conducted in-depth interviews to understand emotional responses. Behavioural metrics like interaction depth and time spent on tasks were also monitored. To measure Engagement, we tracked user activities such as the number of role suggestion impressions and clicks, as well as the drop-off rates between quiz questions. Adoption was gauged by monitoring the number of monthly active users and the personalisation data collected. Retention metrics included tracking how often users returned to the platform within 30 days. Finally, Task Success was evaluated by measuring conversion rates and the time taken for users to receive their quiz results. This multi-faceted approach to gathering user feedback was integral in informing and refining the design process.

  • “You"ve touched on research methods and key insights, but you could elaborate more on how these insights led to design changes or innovations.“
    — Customer feedback
  • To evaluate the potential opportunity, I conducted a survey of our existing user base. The results revealed that over 32% of users were considering a career change. Among this group, satisfaction and work-life balance were identified as key motivating factors.

    Through continuous customer interviews, we have learned about a consistent story that people go through when validating new career directions. I have explored a way to visually depict this story using AI by creating a Customer Journey using ChatGPT, MidJourney, and Figma. To learn more about how this was achieved, visit this article.

    Cutsomer journey storyboard

    After identifying these key stages in the journey, I created a storymap of the experience. From there, I listed the opportunities that we could validate through small lean experiments.

    I documented our vision in an Opportunity Solution Tree. This tree mapped out all the solution ideas we gathered from interviews and ideation sessions. Its purpose was to communicate with stakeholders and prioritize our next steps in experimentation. Additionally, we created a Figma prototype to visually represent the core aspects of the experience we were focusing on.

    From our research we knew that there was a large portion of our user base that were open to career changing. We believed that we could target them on-site through there behaviour verses what is listed in their SEEK Profile. We could see that there were;

    • People just looking for a job right now and open to options
    • People exploring adjacent roles
    • People exploring different possible upward moves

  • To address the challenges of career discovery, we developed several solutions within Career Compass:

    1. Personalised Career Recommendations: By leveraging Large Language Models (LLMs) and vector search, we match user responses from a multiple-choice quiz and prompts to our detailed role page insights. This solution provides personalised career recommendations that resonate with users" personal and professional aspirations, helping them explore new opportunities.
    2. Comprehensive Platform: Career Compass isn"t just a one-off solution but integrated with a comprehensive insights in Career Advice which integrates tightly with SEEK core jobs. To support users at every phase of their career journey. It starts with introspection, helping users identify their skills, interests, and values, and then provides actionable advice to guide their first steps towards a new role. By aligning users" unique profiles with relevant job opportunities, Career Compass increases engagement and satisfaction in their professional lives.
    3. Integration of AI: By utilising large language models like chatGPT, we have democratised access to high-quality career advice. This integration of AI allows us to provide personalised career insights and recommendations based on users" responses, enhancing the accuracy and relevance of our solutions.

    Each of these solutions addresses a specific aspect of the problem of career discovery, providing users with personalised guidance, actionable advice, and relevant job opportunities. By combining these solutions, we offer a unique and comprehensive approach to career change, empowering users to navigate the intricate maze of career possibilities with confidence.

    Strategically, Career Compass positions itself as a tool that goes beyond traditional career quizzes and personality tests. It provides users with a personalised and interactive experience, supporting them through their career change journey. By leveraging AI and data-driven insights, we offer a more accurate and efficient way for users to discover and explore new career opportunities. The strategic value lies in our ability to simplify the complex process of career change and provide users with tailored pathways to their next fulfilling career.

    Through our innovative solutions and user-centric approach, Career Compass aims to revolutionise the way people approach career change, empowering them to make informed decisions and find meaningful and satisfying career paths.

How it performed

  • Conversion to role page Target 33%: Actual 26.2% 📈

    Because we were launching something new we took our base metrics from an existing tool that explored roles based on skills and experience, but only the indirect conversions. That is the number of people who left the tool and then went on to view a role-page. So direct conversions was always a tall order. I"m pretty happy with where we landed, 26.2% of people who start the quiz go on to click directly through form a recommendation. and 38.3% continue to role pages either directly or indirectly.

  • Positive Sentiment >35%: Actual 74% 😀

    Honestly we undersold ourselves here. I thought our first release would be mostly annoying and only occasionally helpful but the feedback has been extremely positive. It seems like people are itching for some support, any support. And most people seem to get something out of even this first version, 74% of feedback is positive. (17% negative, 9% neutral)

  • “You"ve been an eye-opener for me to find the perfect Career that I fit in.“
    — Customer feedback

Personal Reflection

  • From our experience, Career Compass has been successful in providing personalised and relevant career recommendations to users. The tool has achieved a conversion rate of 26.2% and received positive feedback from 74% of customers.

    Working on this project has taught us the importance of integrating personalised recommendations into the user experience. It has also highlighted the significance of early collaboration between designers and developers to ensure a seamless implementation.

    However, we have also encountered challenges along the way. One area for improvement is to further refine the user journey and enhance the accuracy of career recommendations. Additionally, we have learned the importance of continuously gathering user feedback and iterating on the product to meet evolving user needs.

    Moving forward, we plan to conduct further research and analysis to enhance the tool"s performance. We will continue to iterate on the design and functionality of Career Compass to provide even more accurate and personalised career recommendations.

    Overall, our journey with Career Compass has been a valuable learning experience, and we are excited about the future iterations and enhancements that will further empower career changers in their journey.

Conclusion & future iterations

  • Key Achievements:

    • Developed Career Compass v0.1, a tool to help career changers navigate opportunities
    • Achieved a 26.2% conversion rate and received 74% positive customer feedback
    • Provided personalised and relevant career recommendations to users
    • Simplified the complex process of career discovery through an interactive guide

  • Future Iterations:

    • Refine the user journey and enhance the accuracy of career recommendations
    • Continuously gather user feedback and iterate on the product to meet evolving needs
    • Conduct further research and analysis to enhance the tool"s performance
    • Improve the integration of personalised recommendations into the user experience
    • Enhance collaboration between designers and developers for seamless implementation
    The team is excited about the future iterations and enhancements that will empower career changers on their journey.