"HEY, LAUREN, CAN YOU MAKE SOMETHING REAL QUICK?"

The project, like most, did not arrive at my feet in a nicely packaged box with a bow. It began with the head of design coming to me asking for optimized site designs for a large prospective client who needed to be wowed. I am not one to simply slap something together that looks nice, I wanted to make sure whatever deliverables we provided had some actual numbers behind it. In this case, they were coming to Frankly with the hopes that we could give them a content publishing solution that increased engagement and put more money in their pocket.

My first question was "who are their users?".

I was shocked to discover that even though this is one of the largest news media groups in the world, they had no idea who their readers were nor what they wanted. 

My second question, then, was "can they get us in contact with some of their users so we can conduct testing and interview sessions?".  

Again, another roadblock - their legal parameters protected their users' privacy to such a degree that it would be impossible to get in direct contact with their readers through them as a resource. 

WHAT IS OUR CURRENT CLIENT ONBOARDING SYSTEM AND HOW CAN WE IMPROVE IT? 

Before jumping into this project, I wanted to first visualize our flow from requirement gathering to new site launch with new and existing clients. Whatever we learned needed to be implemented by others in the company, making their lives EASIER rather than more complicated with a new system that better ensures adoption. Coworkers are users too!

I had already been working with multiple news media groups in creating templates for their local stations located across the US, and had been through many meetings and iteration reviews that all seemed to need the same things in different packages. I had also made it a point to interview everybody in the company who regularly interfaced with our clients, and had joined multiple onboarding and training sessions at some of the stations to get a sense of what the design process was from beginning to end. I began seeing patterns of need, roadblock, frustration points, and resource drain. 

We had just recently started moving in a more modular direction with our design options, yet how we were presenting them to our clients was far too abstract. It may be easy to visualize for a designer, but not for a client. In addition, we needed to help guide them toward GOOD design decisions, backed by data. We needed to take more control and offer them fewer options that promised higher impact. 

WHY DO PEOPLE READ THE NEWS?

Before jumping into pixels, I wanted to first understand the underlying motivations behind users frequenting local news sites, and see if there might be patterned indicators of their preferences and behaviors.

I recruited several users across my identified demographics who each reported to frequent one or several local news sites, and conducted one hour interview sessions with each to uncover motivations, behaviors, reading/brand preferences, accessibility limitations and device usage. I kept the remaining 100 survey responses (answered by hopeful participants) to include in my research and to share with the team. These sessions were conducted in our San Francisco office.

I used four sites to test: our client's, the users preferred local news site, and two typical stations with different design layouts.

My sessions typically start with open-ended get-to-know-you questions, allowing the user to tell me things I might otherwise not even think to ask, which I implemented here. I then followed up with a task list, last impression rating, and a "draw your ideal experience" exercise on the whiteboard.

A typical session looks a bit like this.

KEY FINDINGS

I was faced with potential location bias: we were testing in San Francisco - most of our clients are based in the south & east coast. Could there possibly be some information I'm not able to gather by pin-holing my view to SF? That in mind, I set up two more sessions in New York and Philadelphia, this time inviting some coworkers to sit and observe, even participate a bit in the interviewing process to get a sense of their skill and begin teaching my research methodology.

I wanted to make sure we could continue internally testing without my having to be there to lead every session, so I trained a handful of colleagues in our sister offices to cover multiple areas.

I am a big supporter of the 5 user rule in usability testing. Aside from it being time/cost efficient, it allows you to be extremely agile in your research methods. However, in order for it to work, you must A) select your users intelligently, B) have a tight focus on your research objectives, and C) supplement with larger numbers of response, like a survey, so you can make sure you can really observe a strongly supported standard of deviation. 

HURRICANES SWEPT IN ANOTHER USER

In the wake of hurricanes Harvey and Irma and the Napa fires, I had been witness to some of my dearest friends struggling to navigate (and help their loved ones navigate) one of the scariest moments of their lives.  

I watched as people used Facebook to locate and help people in need. I saw an incredible spike in communication/mapping/weather tracking app downloads, as well as sign ups for news notifications. Friends from ABC News told me they had experienced a 300% traffic increase during Hurricane Harvey - this is clearly an area that could use some attention.

I conducted 8 interviews with people who were either directly affected by the hurricanes or who had been the main resource for helping their loved ones navigate the devastation.

SO, WHO ARE WE DESIGNING FOR?

Based on motivations and design preferences across 60 million potential users, I came up with four solid personas, taking care to flesh out their motivations, typical journeys, potential points of friction, device usage, and potential design solutions for review.

At this stage, I worked closely with our engineers and PMs to get a sense of what our current feature roadmap was to make sure my solutions were within the parameters of our current abilities. 

PATTERNS REVEALED

At this point, I was now able to uncover a laundry list of best practices to keep in mind when designing these site templates, as well as begin to formulate focused questions to ask stations while trying to identify a station's users' preferences at the time of design onboarding.

I would share more, but there's only so much I can reveal about our findings. Here are few of the big ones that made a decent impact.

NEWS STATIONS

At this time, ongoing user research was underway and it was time to do the same kind of bucketing with our current and prospective clients as I had done with the end users.

Not only are we designing a frontend experience, but we have to continuously keep in mind how stations are publishing their content and what kind/quality of content we have to display.

Fortunately for me, Frankly has an incredible Client Services team in place, many who have been journalists and digital producers. I took this time to sit a few of them down to get a rundown of the different "types" of local news stations that are out there, breaking them down by size, content + design, and marketing angles.

Paired with my previous on-site research, I came away with some great initial personas.

We have been able to present early design options to clients during the onboarding process. Having gone from showing individual module features to suggested configurations (much like graduating from single lego pieces to full models), we have been able to cut our client conversations down tremendously.