12 Weeks. 8 People. 1 Goal.

The Challenge: Develop a product to help people in the sharing economy reduce their risks.

It was quite the undertaking to be given. Driven by the Innovation Forum – a cross-SBU group of innovation leaders at Liberty – and supported with guiding autonomy, we set out to conquer our challenge. But what we quickly realized was that our broad challenge lacked focus (duh!). Instead of hunkering down in slews of internet research to determine our direction, we went broad with a hunch workshop.

Hunch Workshop

A hunch workshop is a design tool used to get everyone’s feelings and intuition out in the open. Based on our own experiences, insights, and judgement – and supported by initial evidence – we came up with 16 different hunches. Hunches varied from “Sharing Economy platforms have difficulty with sourcing and retention” to “Engaging with strangers makes people nervous, decreasing Sharing Economy use.” Each hunch was made up of 3 parts and posted on one large flip-chart sheet:

• The Hunch Itself: A statement that is ‘user-focused’ and people can generally relate to. The simpler, the better.
• Supporting Evidence: A few bullets that support the hunch. Hunches should have a basis of evidence, and not just opinion.
• Concepts: 2 – 3 Concepts that could be quickly dreamed up and solve that hunch. This shows that there is broad opportunity in which to innovate.

After generating all of these hunches, we needed a way to pare them down to create the focus we desired. In simple form we brought back in our stakeholders for a readout. Each hunch was given just 4 minutes for presentation and discussion, ensuring our hunches remained succinct, with a clearly identifiable problem and potential for broad solutions. All participants were asked to confirm or challenge our hunches based on their own experiences and evidence-driven knowledge. After the challenges and confirmations were aired and all hunches were heard, we used a standard dot method to vote on ideas that we felt had the most ‘legs.’

So what qualified as a leg to stand on? A hunch needed to be solving for a real problem. It needed to have broad enough opportunity in which to explore, yet narrow enough to provide focus. And finally, Liberty should have an unfair advantage to pursue such an opportunity. From those 16 original hunches, we had solid reactions (a high number of votes) on 7 of them and were able to group a few of those together to come to 3 we would explore further:

1. How might we help sharing economy providers navigate independent contractorship?
2. How might we help people feel effortlessly safe to explore peer to peer transactions?
3. How might we humanize sharing economy interactions for better experiences and reduced risks?

At this point, we were just about a week in and had narrowed in on our targets…or had we?

Key Takeaway: Collaborate up front to expose all of your initial ideas. Using evidence and voting to verify or refute each hunch with your stakeholders helps everyone get started on the right foot!

Sourcing Users

Using human-centered design can be a lot of fun. At its core, HCD (as the kewl kids call it) encourages its proponents to go talk to users, empathize with them, understand their wants/needs/desires and build great new products to fulfill both their obvious and latent desires. Great in theory, but it can actually be quite hard in practice. Finding the right users to talk to can oftentimes be quite difficult, and this project was no exception. So we broke our users down into 3 pertinent user groups:

Now, I know what you’re thinking. Users (demand) should be SUPER easy to come by, especially in a young, professional city like Boston. True, but finding the extremes of this user type proved to be hard. You see, according to Design Thinking methodology, speaking to extreme users is seen as incredibly valuable. Take a person who hates and refuses to use Uber, for example. They will highlight issues – such as inconsistent background checks – that average users have concerns about yet overlook. Improving the experience for them will truly improve the experience for everyone. Or, on the opposite end of the spectrum, consider the hardcore Uber passenger who has given up their car in favor of constant uber rides. What kind of awesome nuggets of wisdom could you find from those people?!

To source all of our user types, from extremes to average, we tried a variety of methods. We took out ads on Reddit. We created listings on Craisgslist. We created our own research group and advertised research opportunities on Facebook.  We even passed out free donuts to strangers in hopes of finding a diamond in the rough. All of this was met with limited success and often poor quality candidates. Our best feeder turned out to be a Boston-based website that specializes in exactly our needs. UserInterviews.com is a platform that sources users of all kinds for a small fee. While candidates were not always perfect, we saved countless hours and headaches leveraging that platform to find each type of user listed above.

Additionally, we leveraged the expertise of GLG – a professional research sourcing company – for a lot of our platform research in order to have anonymous authority to the conversation. Being anonymous was the goal because we did not want to introduce any biases. When people think of insurance, they typically have a negative reaction which can change their attitudes and behaviors. This may lead to people to not be very forthcoming with you as they may mask their real habits (such as being overly cautious when driving or discussing their drivers’ habits). Initiating dialogue on an even footing gives you a great chance to best understand how these user groups truly interact, which is the basis for all of our innovation efforts.

Talking to Users

Once sourced, the fun part began. We created moderator’s guides to use as a reference when conducting our conversations. These moderator’s guides weren’t just a list of questions, rather they were broken into sets of activities. Our guides typically included games such as a getting to know you mad-lib, a fun card sort exercise, and even a game involving Starburst candy. We used different mod guides for each user group in order to focus in on assumptions we had made based on each user type and to get to know them uniquely even more.

Discussions with users came in two main forms, group discussions and one-on-ones. Group discussions, or Whine & Dines as they are sometimes called, happen in a relaxed group setting. We orchestrated them around dinner time and provided heavy hors d’oeuvres to create a comfortable and open environment where everyone could share freely. Research has shown that people tend to be happier and more engaged when they are eating, so we used that to our advantage and had great results. In around 2 hours worth of time we were able to gather initial opinions and understandings from a small group (6-10) of individuals that would inform our future conversations, opportunities, and prototypes.

Individual discussions, or one-on-one interviews, were great resources for more in-depth discussions. We used interviews to more deeply understand our users, dive deeper into certain topics, and to get out of our surroundings to see theirs. We took about uber rides, stayed in an AirBNB, brought in a dog walker for the team dog, and began to see first hand what these experiences were like. However we weren’t content with just this. From the provider angle we felt the need to really understand them more deeply, so we undertook an intense form of observation – immersion.

The Immersive Experience

Immersion – or ethnographic research – involves integrating yourself in to the community you wish to serve. You do this in order to more deeply understand the needs and joys of a person in this community. In this case, we assumed roles as sharing economy providers: Uber drivers, Postmate deliverymen, Handy house cleaners, Instacart delivery people, and Wag dog walkers.

For one week we put our typical jobs on hold and became the sharing economy providers we wished to serve. The time spent doing this was both frustrating and rewarding. We experienced first hand the difficulty of maneuvering a big city at all hours of the day and night as well as the liability of leaving your car double-parked to make a delivery.  Or how about the awkward encounter of cleaning a stranger’s kitchen as they’re trying to make lunch?

All of these experiences gave us a completely different perspective on the lives of the sharing economy contractor – both the good, the bad, and most importantly: what was missing.

Divergent Themes

Through our various discussions, interviews, and experiences we were able to put our initial hunches to the test. What we found was not that our hunches were good or bad, but that they were great conversation starters that led into brand new areas we hadn’t explored. Our original hunches were molded, modified, split and and re-imagined as we chatted with more and more users (demand), providers, and platforms letting them all guide the way. From our original hunches we diverged into a set of 27 different themes based on hundreds of quotes & interactions.

Convergent Opportunities (Jobs to be Done)

But what the heck can anyone do with 27 themes? Very little, actually. That’s why it’s especially important to converge on opportunities. This is done in the same way we widdled down our hunches from 16 to 3, only this time the voting remained insulated within our team. Why? Because we had done all of the interviewing and had the closest relationships and experiences with the users. We were no longer voting on what we believed, but on what we heard, observed, and experienced. As advocates for our users, we voted with their needs front of mind and consolidated our themes down to 7 different opportunities, also known as “jobs to be done.”

After doing this, we took a step back and began to think outside of just desirability. Until this step we had been fully focused on generating empathy and concepts that would delight users and solve their problems. But now, tasked with digging deeper, we had to be real with the opportunities in front of us.

After many conversations we made the tough decision to drop users as a group of focus. We would no longer create concepts and materials specifically targeting their needs. We made this tough decision because users, as a trend, did not believe the risks inherent in the sharing economy were worth their direct resources of time & money. They tended to be less aware and more carefree about risks and opportunities in this space, instead identifying providers or platforms as the ones who they thought should provide experience improvements. Therefore, our proposed solutions would have to address opportunities from the perspectives of these two user groups while indirectly benefiting the users themselves.

Eliminating users (demand) as a group brought us down to 6 opportunity areas in which to focus. And the job of generating new divergent ideas based on our 6 new areas began.

Rinse and Repeat

And so the process of talking to providers and platforms began again. At this point we split up into two teams each diving deeper and gaining a bigger focus on one particular user group. Each day we would share back our findings to the group so everyone remained involved in all aspects of the project, however this approach allowed for deeper insights to be gleaned due to the added focus. This approach was not standard, but it worked for us as a team and that’s the big takeaway – do what works for you!

We diverged and converged two more times focusing mainly on desirability. During this time we performed many more one on one interviews and dug deep into how our user groups actually acted. We took more Uber rides, stayed in an AirBNB, brought in the team dog for a walk with Wag and tried our best not just to interview these people, but to experience a critical part of their lives through their eyes. As we continued to diverge and converge, our questions began to shift from broad concepts on empathy and understanding turning eventually into questions on delivery mechanisms, particular functions and features as well as mapping out potential for a great customer experience with the product.

Prototypes

With the fine tuning of our questions came a tuning of our concepts. As our questions got more granular, our prototypes did too. We had come a long way from our hunch sketches (which we did show to real users) and created higher fidelity sketches, storyboards, wireframes, and even now clickable prototypes. The higher fidelity of prototype were created to test deeper assumptions and bring our proposed experiences to life. With the added detail in our prototypes came additional detail in the questions and responses from the users.

Concept Testing

At the end of our last diverge/converge cycle we had narrowed in on a few concepts that we felt had strong indicators for success. To test our qualitative data gathering from the past couple of months, we submitted many various and specialized versions of our final ideas to a rigorous internal concept testing process.

What this testing process does is give a more statistically comparable viewpoint on the desirability from potential users of each concept. We had over 100 individuals rank and give feedback on our concepts in survey form and used this to guide our final understanding and report out. This did not dictate what we would end up recommending, but rather provided additional insight and  quantitative data points from which we could draw additional conclusions. This, combined with our previous interviews and our feasibility/viability research, helped to shape our final recommendations.

Feasibility, Viability and Unfair Advantage

The final two weeks of our sprint were spent diving head-on into the world of feasibility, viability, and Liberty’s right to play in this space. We tested each of our concepts out by interviewing many internal employees with expertise or an analogous viewpoint that would relate to each of our concepts. We dug into whether our ideas were technically able to be done (one was not, and died there), would be theoretically viable and profitable, and if these concepts felt complimentary to Liberty’s offerings in a way that would create a competitive advantage.

Beyond discussions, we dug into the numbers. Research was done on potential market size and revenue projections with input from various sources. We used PwC & Deloitte projections and research, leveraged information gleaned from our discussions, and scoured online and offline resources for credible data points to affirm or challenge our assumptions that each concept was a viable business for Liberty to pursue. At the end of two weeks, we had 3 concepts that we felt met all of the criteria for successful pilot offerings.

Our Concepts

Our concepts are not the focal point of this piece. Rather, it was the learning gained on leveraging the Design Thinking methodology that made these results possible. For confidentiality reasons, I cannot release the final insights and concepts. But what I can say is that you should definitely check out GigUp and be on the lookout for other new Liberty (and non-Liberty branded) ventures.

Still Left to Explore and Improve…

Project Frontier was a fantastic learning experience. Not only was it incredibly successful with 3 final recommendations we have confidence in, but we learned plenty about the process along the way. If you’re looking to run a similar sprint, we recommend a few things:

• Have Roles: These don’t have to be static throughout the entire sprint, but it’s a good idea to nominate a project lead, a project historian, a user recruiter, and a couple of design researchers (especially if one is good at visual & prototype design).
• Be 100% dedicated with an ‘out of office’ mentality. 100% dedicated = 100% critical.
• Have business leaders ‘on call’ to verify feasibility/viability assumptions throughout
• Break your challenge down into smaller steps. Broad challenges, like this one, should output specific challenges and not necessarily fully baked concepts. We recommend following a 3-sprint system: Research Sprint, Design Sprint, Development Sprint.  Depending on the topic and scope, each sprint could last a variable amount of time.
• Have all stakeholders and participants clearly aligned on the requirements of the Design Thinking mindset and the budget/activities required. This may involve some travel to meet with users, perform an analogous experience, and do tasks not typically performed in typical research endeavors.
• Define a purpose, particularly when it comes to fulfilling performance review objectives. Members of the Sprint team want to know and be given credit for the important nature of this work.
• Use tools that are fit for purpose. Use Agile tools like a Kanban board to organize the team’s work, make the team calendar very visible, know when to pay platforms for user sourcing, and have company-agnostic materials so as not to introduce biases.

We did not explore everything we could have. In reality, there just isn’t as much time as would be necessary for such a growing industry. And that’s okay. In the future we hope to take a deeper look at the rise of the sole proprietor as well as the growth of skilled labor within the sharing economy – two topics we only briefly touched upon. Celebrate what you can accomplish and don’t spread yourself too thin on what you cannot.

Final Thoughts

Personally, I would highly recommend becoming involved in innovation and partaking in a sprint like this when given the chance. Sure, I’m biased, but the experiences you get from stepping out of your day job and into the shoes of your users is unmatched elsewhere. And for those people who don’t get an opportunity for a 12-week sprint like this, that’s okay. These are rare, yet the principles we followed can be brought back into any work environment and used in almost any context. So give it a try. Just remember the ABC’s of Design Thinking: ABC = always be creating.

• Liberty Mutual

• May 31, 2017