Getting Beyond Hooking Up and Linking In — Interview with Felix Tin – Director of User Experience at TeamFit
Felix Tin joined as Director of User Experience back in April 2014 and helped us make the transition from team collaboration to team building (see Why TeamFit). Felix came to user experience or UX from the computer engineering side (he has a degree in systems design engineering from the University of Waterloo). He is that rare bird who combines a deep understanding of systems with a feel for the human side of design (and that is not just ‘human factors’ but the humane).
UX design calls for a lot of different skills. The ability to really get inside people’s heads and understand what motivates them, how they communicate with each other; how to interview users without shaping the responses; how to design a system of user interactions; how to create habit forming experiences, it seems endless. And you need a lot of different experiences to do it well.
Felix has done UX for all sorts of different applications: hardware at Logitech, mobile at AirG and two-sided markets at Plenty of Fish. He brings all of this experience to bear on TeamFit.
What are three critical things that you focussed on in the UX design for TeamFit?
This is a two-sided market, addressing both sides is a real challenge. Identifying and then understanding the two sides was critical. In our case, the two-sides are people who work on project teams and the people who need to pull these teams together. They have many shared concerns but there are also places where there are disconnects. Each side has information the other lacks and that leads to poor decisions on both sides.
Talking to people on both sides of the market was critical. We had to get to the pain points. Pain is personal and you have to build some trust and credibility before people will really share their pain with you. We had to go beyond just the pain and understand what people were doing to solve that pain today.
We found that there are multiple models for team assembly. Far more than we were aware of before we talked to people. The model for team assembly has a big impact on the needs of consultants. For example, in companies where the team builder takes into account the consultants career aspirations the consultant does not need to be as aggressive about marketing themselves internally and fighting to get on the right projects.
Another issue is “What is the core concept?” We started by organizing our thinking around ‘teams,’ but after many conversations with users we moved to organizing around projects and project requirements. The team is the people who do the projects. People think about this more in terms of getting a team for a project and getting on a project than the more subjective idea of joining a team. This is especially true in our target markets of business consulting, systems integration and professional services, where projects (cases, implementations) are the core unit of work.
We are also proposing new and better ways for people to evaluate skills. People told us very strongly that existing methods don’t work. Social skill endorsements are too removed from actual work to be believable or even useful. The formal taxonomies and processes of corporate skill management systems are too rigid and top down to provide a lot of value.
There is a lot of technology behind what we are doing (Bayesian probabilities, clustering, trust models) but we still need to do a lot of work to understand this for the user. The best math in the world doesn’t help until we make it meaningful to people so that they can really understand each other and make decisions.
Our approach is much more granular than anything that exists on other endorsement systems. Skills are endorsed in the context of specific projects. This is great, it makes the endorsements much more meaningful, but we also need to be able to pop up and see the person as a whole. We are all more than just the string of projects we have worked on.
What are some of the different design directions that you think TeamFit could go?
We have already learned a lot, from ourselves and from early users and can see what is working well and where some of the big gaps are. We have to guard against just the temptation to just add features. On our Trello board, we actually have a theme for ‘things to remove or simplify.’
But certain concepts are bubbling up in importance.
How do we create a habit-forming service? We are all students of Nir Eyal and his book Hooked. We are building something that as so much value that people always have it open and come back everyday. The whole cue-routine-reward loop is an important part of our design thinking.
We want to get to a point where we can show the gestalt of a person, a group of people or an organization. (The word ‘gestalt’ came from one of our users; he used it to mean something like ‘get a feel for a person, their skills, their work style and experience at a glance, and primarily visually). Getting to this will take very good data visualization on our part. We are looking for compelling, visual ways to show who a person is and how people work together.
Right now a lot of our focus is on data collection, helping people to enter information about what projects they have worked on, what skills they have demonstrated, who they have worked with, how successful the projects have been. We are also working on integrations with services like Slack, Asana and Basecamp to make it even easier for people to gather information on all of the different projects they have worked on.
But data collection is just the first step. The more data we have the more focus there will be on building up the machine learning on our back end and then delivering insights to users through dashboard and visualizations.
What did you learn from other positions that you are applying at TeamFit?
We have sometimes described TeamFit as ‘eHarmony for teams’ or ‘Match.com for for teams.’ (But then we have also thought of it as ‘Moneyball for teams’!). This is kind of amusing for me as I was director of User Experience for Plenty of Fish, which is about finding romantic fulfillment. TeamFit does a similar thing for professional relationships; it is a similar problem at heart (smile). So I am taking lots of ideas about what makes a two-sided market work and what predictions are most important in that context.
For TeamFit we think our TeamMatch™ score (how well a person meets the needs of a team given the other people who are on it) and our TeamFit™ score (how well a team as a whole fits the requirements for a project and likely the team is to succeed on the project) are the critical predictions, the ones that have to really work and to add value.
Of course TeamFit is about groups of people rather than couples, so more information is available, but more interactions are possible. This makes the analytics, visualizations and above all the simplification, cutting to essentials, all the more important.
