Would You Rather Be …

Phil Hewinson
15 min readOct 31, 2019

--

Over the past 5 months I’ve been working on a web app to help people find a satisfying career — it’s called “Would You Rather Be”. The app asks you to choose which of two careers you would prefer to do. For example:

After 100 of these questions, the app shows the top qualities you look for in a career and the top careers that match those qualities. It then asks you if you would like to chat with someone doing one of your top careers to explore further.

So what would you rather be? 😉 Try it for yourself at www.wouldyouratherbe.com!

Why Employment?

Since starting my journey into social entrepreneurship just over a year ago, I’ve given a lot of thought to the best ways software can make a difference to people’s lives. I explored the many ways that software is making a difference today. These cover broad categories like education, employment, healthcare and finance. They also cover narrower categories like refugees, homeless people, victims of trafficking and disabled people.

One of the great properties of software is how it scales. You can write some code and the only limit to how many people it can impact is the size of the market. You write it once and it runs everywhere with next to no extra effort or costs. So it seems to me that software can be best leveraged to address broader categories with the largest addressable market, rather than narrower ones.

When it comes to employment, most of the world’s population need to work. The only times when you typically don’t is when you are too young and haven’t yet developed the necessary skills or when you are too old and your health may prevent you from working. The majority of a person’s life and their best energy is usually spent on their career.

Work gives people independence, dignity and purpose. It can enable communities in developing countries be sustainable rather than rely on foreign aid.

Vision

But what problems can be solved within employment that software can be leveraged to address? And what problems am I passionate about?

The question that I settled on is:

How can I help people find meaningful work?

If people spend most of their energy on their careers, it would be a shame if they didn’t enjoy it. And it would be a loss for employers too as they wouldn’t get the best out of their staff. But how many people really consider all career options when choosing their path? Could there be a better career that they would find more satisfying than the one they picked?

Prototyping Your Career

Last year I went to a talk by Dave Evans, a Professor at Stanford University. He led the design of Apple’s first mouse and co-founded the games company Electronic Arts. He came to London and gave a talk on “Designing Your Life”. This was based on his lecture series at Stanford, which is the most popular at the university. The essence of his talk was the idea of “prototyping” your career.

A prototype is a light-weight experiment. This is best explained with an example. Imagine if you spent 6 years training to be an architect. Then 6 months into the job, you realised you hated it. You would have wasted 6 ½ years chasing after the wrong job. Instead, it would be a lot more efficient to talk with an architect for 45 minutes and explore whether it sounds like a job you would enjoy. Maybe shadow an architect for a day or do a week’s work experience.

Connecting People to Explore Careers through Conversation

While I was on a walk in the early summer this year, I came up with the idea of building a platform to connect people who are interested in certain careers with other people doing those careers so they can have a conversation. They could pay to have these conversations. This means that anyone doing a job could get paid to talk about their job with someone else. And of course the platform would keep a cut so it could scale and be sustainable.

An ideal initial target would be 2nd and 3rd year university students. This is often the first time in a person’s life when they are about to step into the unknown and have to make a big, difficult and complex decision about which of the thousands of possible careers they want to pursue. Every step in their life up to this point has been a series of relatively simple decisions about what to study. But this decision is orders of magnitude more complex.

I reasoned that soon-to-be graduates may not have considered many career possibilities or may not know yet what they want to do. So how would they know who they are interested in talking with? I decided that I had to first try to solve the problem of getting people from not knowing what they want to do to having a list of possible careers that interest them.

Approach

Deciding Between Two Careers

My first step was to get a pretty comprehensive list of careers that people can pursue in the UK. There are many websites that helped to form this list. I came up with about 400 careers.

I then came up with the idea of showing people pairs of careers and asking them to choose their favourite. This is based on the same idea as my last project, Compare and Learn. During that project, I learnt that people are really good at choosing between two options. Given the complexity of choosing what career to pursue, I thought this approach would work well given its simplicity.

But with 400 possible careers, it yields 80,000 possible pairs. I couldn’t possibly show someone all 80,000 pairs and ask them to choose. I figured that I could only show at most 100 pairs to someone. But if I chose 100 pairs at random from the list of 80,000 possibilities, it would only yield a very weak signal on what career would suit someone.

Commonality Between Careers

In order to solve this problem, I needed to first figure out attributes that different careers have in common. As people made choices between pairs of careers, I could then get a signal on what attributes matter most to someone. That would then infer the careers that would suit them.

But what should these attributes be? There are so many different ways of describing careers. Most online tests that ask a series of questions to help you figure out the right career are basically personality tests, often based on Myers Briggs or a variant. But from trying these tests myself and talking to others who have tried these, they don’t seem to often yield really relevant results.

Other approaches could be location, salary, industry, lifestyle, challenge, skills or expertise.

Qualities that Best Describe Careers

My goal though is to help people find careers that are most satisfying to them. And I wanted to think through the problem from first principles and come up with attributes that match really well to jobs. So I went through each of the 400 careers and tried to pick one word that best describes what each career is about and why someone might find it satisfying.

Here are some examples. Accountant is money. Solicitor is law. Nurse is improve wellbeing. Secondary school teacher is teaching.

This yielded a list of about 50 qualities that best describe each career, which meant there was a nice overlap between the 400 careers. This formed the basis for the model I then built to map people’s choices when shown pairs of careers to a final list of qualities and careers that should suit them best.

Crafting an Algorithm

The most complex part was figuring out what pairs of careers to show a person. Picking them at random would yield a weak signal, so they had to be selected intelligently based on their previous choices. Then as people answer the questions, it hones in on the qualities that they find most satisfying in a career.

I won’t go into detail as to how the algorithm works as it’s proprietary and I may seek a patent at some point.

Building Would You Rather Be

I built the model using the approach described above by carefully reading through the job descriptions for all 400 careers and mapping them into my model based on the qualities I defined.

I then went about building the web app. For my last project I used Angular, but it turns out that the most popular web framework is React. I had never used React before, so I had a learning curve to get up to speed with how it worked. It’s based on Javascript, which I’m familiar with, so the learning curve wasn’t too steep.

I used Node for the backend and MongoDB for the database, similar to my last two projects.

Testing

Friends & Family

Once I had finished building the first version, I tested it with myself a few times and then with friends and family. This yielded a lot of useful learnings.

My original approach was to show the top five careers suited to someone, selecting one from each of the person’s top five qualities. This seemed to work quite well for about half of the people I tested it with. But it missed a few careers that would have been a good fit that didn’t quite make the top five cut. Also, the qualities seemed to fit really well with people, but I didn’t highlight these in the results.

So I changed the final results by showing the list of the top 8 qualities that match the person at the end. I then gave them the option of selecting the qualities that sound most appealing to them. I then select the top 15 careers that match these selected qualities.

Next, I asked the person if the results are missing any careers that they are interested in and adding them to their results. I also asked them if they would like to chat with someone doing one of their preferred careers to test the next part of my product vision. I asked them for their email to send them their results and as a way to followup to setup conversations if they requested them or send them updates as the product evolves. And finally I showed them their full results on the final screen. Here are what my results look like:

Target Market

Where to Find People

I then needed to test my product with people in my target market — ideally soon-to-be or recent graduates. To start with, I spent a day researching many possibilities for how I could reach these people. I looked at local universities, colleges and schools and got in touch with local people who run student groups and job clubs. I researched job fairs and got in touch with three big upcoming job fairs in London. And I researched a few careers advisors and coaches.

Finding a Way Into a London Job Fair

My best bet were the three big upcoming job fairs in London. I called each one and they all offered me the opportunity to exhibit there, but at a cost of at least £3,000. As I only needed about 6–8 user tests, I figured it wouldn’t be worth paying more than £300. So I tried negotiating and pitching myself as a hot new tech startup that would add value to their conference. Unfortunately it didn’t work.

So I tried to go in from the top instead. I discovered that the same company runs two of the three exhibitions I was interested in. So I searched around to try and find who runs the company. I found a lady who was a Director and seemed to have founded it, so I got in touch by sending her a LinkedIn request.

To my surprise she replied and said her Head of Marketing would be in touch. She didn’t get in touch, so I found her email online and emailed directly to say that we had been put in touch. Thankfully she replied and invited me to the exhibition focused on graduates as a careers advisor. They wouldn’t pay me, but wouldn’t charge me either. It seemed like a perfect fit!

Preparing For and Acting as a Careers Advisor for Two Days

But now I had about 3 working days to prepare and I had never been a careers advisor before! I went along to another exhibition for school leavers and sat in on a talk from a careers advisor there and took notes. I used the platform Fiverr to have a designer create a logo and business cards for me for £50.

And I also had a t-shirt made with the logo on the front.

The first of the two days was a little crazy. I was the only careers advisor to turn up until 2pm, so I had a lot of people to meet. I handed out business cards to each person who arrived and they completed the questions on their own smartphone in the queue. Then I saw people after they had reached the final results screen to discuss their results and get feedback on their experience.

In total, I conducted about 35 user tests over two days.

Learnings about the Product

What Went Well

The app seemed to work well for about 70–80% of people. It tended to work best for people who were more self-aware and understood what they enjoyed. This is because the app is predicated on the person being able to make good choices on which careers they prefer. It yielded a good set of jobs for most people and almost always yielded more ideas than that person already had.

The app also exposed people to a lot of career possibilities, which broadened their horizons.

The results are quite consistent too when I tested it on myself a number of times. This was an important and encouraging sign.

Food for Thought

Some small improvements that people voiced were that it was a little too long. And in some cases, people didn’t want to do either career, so they wanted a “neither” option. But that didn’t stop people from completing the app.

Finally, many people expressed interest in having paid conversations, which was a promising sign. But since the exhibition, it has been a challenge setting these up. And I don’t think it will scale too well either. Some of the challenges around this model are:

  • People (often students) may not want to pay for these
  • Some people doing certain careers, such as those in very busy jobs, may not have the time or incentive to have these conversations
  • There is a logistical cost to setup these conversations, although software may be able to help with this
  • It is quite a low income opportunity as the most I think I could keep per conversation is £5–10

I’m not writing this opportunity off, but I think a pivot in a different direction may make more sense. So far I’ve tried setting up 8 conversations and only one person has responded saying they are not interested, but do want support in getting a job.

Learnings about Myself

I’ve also learnt a few things about myself during this project. I realise that I really value working with other people in a team. Working on my own for a long period of time doesn’t suit me.

I’ve also realised that it is much healthier for me to focus on just one project at a time.

During the past 6 months I’ve also been working part-time for a startup called Soultime, which is a meditation app that helps people with their spiritual and emotional health. It’s a paid role, so that’s how I’ve been paying the bills. And it’s actually a really great project! I’ll talk more about them in another blog post.

But switching from Soultime to this project and back literally every day has been quite stressful. It would be much healthier if I was just focused on one project at a time.

Next Steps

Immediate Next Steps

There are a couple of things I want to pursue with this project over the next couple of months or so.

I will do some more user testing. I have a user testing session lined up tomorrow morning at the local job club run by Christians Against Poverty. I’ve also lined up some user testing with the student group at my local church in the next couple of months.

I also want to see if I can generate an income from the product as it stands today. I will need to give this more thought, but maybe showing some simple results for free and charging £10 for more comprehensive results could work. Making money at this stage would yield far more opportunities for me, so it’s worth a go.

Raising Money and Finding Cofounders

At some point I may also try and raise some funding. There is a scheme called SEIS whereby it is relatively easy to raise up to £150,000 from wealthy individuals as it is extremely tax efficient for them to invest. Raising money would then give me far more options to explore focusing on this fulltime and finding a cofounder or two to join me on the journey. Given what I have learnt about myself during this project, I think that is an important step for me to take if I want to take this project much further.

If I can get funding and find a cofounder or two, there are so many opportunities for where I could take this startup.

A Much Bigger Product Vision

There are a few more steps I can help solve to get people into a satisfying job. I’ve helped people identify the roles that may satisfy them, but I could then help them identify the ideal companies for them that have these open roles, help them get a job there and then help them flourish in that job.

I have a whole host of ideas to explore for each of these steps.

For discovering roles, I can further expand on the app I have today by adding more roles. I could record short videos of people doing these roles to help people explore them more effectively. I could also further explore the conversation route.

For discovering the ideal company to work at, I could help people explore the right industry, the attributes of a company that matter to them and the jobs that are available today and their ability to access these jobs and how to fill any gaps that exist. Even providing a whole market view on what jobs are available in the UK could be a massive value-add as I’m not sure that exists in a single place today.

For helping people get their ideal job, I can provide a service to help them create an amazing CV, write a killer job application and help them stand out. Plus prepare them for their interviews and coach them throughout the process.

Finally, for helping people flourish in their job, I could create a mentoring platform so that people can be mentored by others more experienced than they are in the same job. I could help them progress in their job (either in their existing company or another company). And I could help them further develop their skills.

Making Money

There are lots of ways to make money and build a scalable and sustainable business in this space too. We could charge applicants for some of these services, either up front or after they get a job (so tied to results). It could even be a % of their first paycheck. We could also charge employers for advertising jobs, recruitment fees, providing a mentoring service or helping their employees develop skills. And we could get referral fees for referring users to career and life coaches.

So I’ll probably test whether I can make money from my existing product in the coming months. I may then keep my eye out for opportunities for funding and collaborators to join me on this journey. Meanwhile, I may spend more of my time on Soultime and less on this project so I’m not constantly switching between the two of them.

A Short Update on Past Projects

You may be wondering what has happened to the malaria chatbot and education web app.

The malaria chatbot still isn’t live yet, but is progressing. We hit some complications when it comes to payments and privacy and decided that we would have to integrate with the Against Malaria Foundation’s payment system. We’re very close to completing that work now, so hopefully it won’t be too long before that is live.

The education web app grew to about 450 teachers who created accounts. However, the engagement metrics aren’t very strong. We only have about 20 monthly active users. I think this means that it’s not solving a very strong need for teachers and I’m not sure that iterating on it will get us much closer to product-market fit. The two teachers who I have been working with on this project are continuing to explore why teachers are not using it to see if we can create value for them.

Thanks!

Thanks for reading! If you’ve not tried the app yet, give it a go at www.wouldyouratherbe.com.

I’ll share more on my work at Soultime soon!

--

--

Phil Hewinson
Phil Hewinson

No responses yet