Mappsto.re

I've been experimenting with some custom map tiles. The above are courtesy of Stamen and Polymaps. 

So far, user feedback has indicated that whilst pretty, people prefer the standard Google Maps Roadmap. 

The herd of sheep. How, as a user, can we be expected to process and deal with this many data points? The entire map is obscured.

As I'm working alone on this project, I'm operating a mix between Agile and Lean methodologies. 

I've been undertaking some design A/B testing - respondents answer one question at a time, and have only two choices at each step. 

I've settled for a web app, with a Ruby On Rails backbone. 

Put simply, some early iOS tests took a lot longer than I expected. I was working with Parse, but even with their CloudCode and Push features, small changes were taking longer than I'd like to push out to testers. To A/B test 5 different interaction styles with 30 users, for example. 

Web is simplistic, and suits RAD. So here goes. 

Which analytics tools do you use? In particular for funnel and cohort analysis: on my radar at the moment (in addition to Parse, Heroku in built) are:

- Flurry

- Geckoboard (for monitoring)

- GoSquared

- Google Analytics

- KissMetrics

I've been meaning to try Phonegap out for some time now, so spent a few days looking into it. First impressions were really positive: as developers, any service that allows us to to code in a language we are comfortable with is a huge perk, but once the application was compiled and on my phone, I understood why this technology was considered promising but far off industry acceptance. 

I recently spoke to some of the team at Sky, who were able to optimize Phonegap applications to the level that the latency was non-existent.

Over the last few days I've been pushing out a survey for user inquiry.

If you'd like to contribute, head over to SurveyMonkey. I've had 112 responses, with the following interesting insights:

1. 83.8% find out about local news and events online. 

2. 74.0% don’t mind being located in order to provide live, locally relevant information. 

In addition to surveys, I've been talking to as many people as I can in order to ascertain usage habits, popular applications and so forth. 

This is proving extremely useful at this stage of my project, providing insight into how your everyday consumer perceives the different tools they interact with on a day to day basis. 

I've spoken to about 10 people so far, and the same trend keeps popping up - I ask them which applications they use most often, and travel is a key category. At the very end of the interview I ask a deilberately biased question: "Which of your applications have a location based feature?" It seems to throw people, and they miss out many of the obvious examples like Instagram, Twitter, Facebook, Foursquare and Eventbrite. 

I recently tested out Silverback, but have been searching for something similar built for mobile platforms. 

UX Recorder is a take on the same problem: screen, front facing camera and audio recording for the iOS platform. 

The only catch is that it can record web browser sessions only. The performance seems adequate, and unintrusive - the only cue that the front facing camera and microphone are recording is a small red dot.

The business model is to charge for "session credits", which allow you to record for longer and export your rendered clips to a variety of services. Overall, a tool which I will use for the web based client, but still not suitable for testing native application.

The main challenges I have identified thus far: 

1) Overloaded location based market - of all applications out there, the SoLoMo tagline must be the most overused. 

2) Mobile UI - maps have not truly been tailor made for the mobile platform (yet). 

I've been experimenting with a new application called Pop App over the last few days. Pop allows you to turn static images into interactive prototypes. It includes simulated functionality for iPhone interaction elements like back, dismiss, slide up and so forth. You can see a video of one of my user tests above: this particular individual said he found it "intuitive to use" and subsequently gave some really insightful suggestions for where I should position my UI elements.  

It's somewhat similar to a few other tools like Prototypes, but crucially it's free, and also encourages paper prototyping over full proof designs. 

Silverback is an industry standard usability testing tool. Here at UCL, we've been cooking up something similar up for mobile, so I thought I'd test our Mac counterpart's functionality out on some of my browser-based map tests.

I was impressed - the product itself is very lean and efficient, but could benefit from a little more intelligence behind the scenes; I'm sure there are a huge amount of cues which can automatically be generated from system and browser logs. Overall, I was pleased, and will be using this when undertaking user tests. 

I recently stumbled upon LiveView. What an incredible tool.

If you're developing designs for iOS, this is an essential feedback tool. Ever feel that what you're designing might look slightly different on you iPhone screen? That's not a problem any more - LiveView connects to your devices over WiFi and screencasts directly to them. 

Simply download the application for both your laptop/desktop and iOS devices, launch it on both and you're set. The devices do need to be on the same WiFi network, so stop dreaming of broadcasting your design changes to your co-founders phone, 5000 miles away. 

The quality is lossless, so the frame rate is only just noticeable. It's not fast enough for video streaming (averaging around 6fps for me), but absolutely perfect for design feedback.

You can screencast to multiple devices at once, and even set up interactive mode where you can control your laptop or desktop directly from your iOS device. Neat. 

Interactive chloropleth map on Why Not The Best. 

+ High level view easily understood

I had a great session today with Vicky Cullen of uMotif and UCL Decide.

I've been working with Vicky for some time, and she's ironed out a really useful process of user mapping which I would recommend to any fledgling startup.

It's easy to assume you understand exactly how your product will be used, but often, the most popular use cases aren't the ones you intended. 

Similar to surveys, Vicky had me drawing out the user's journey myself, rather than interjecting or influencing any of my assumptions. Once we'd completed the user journey, we went back through to identify key trigger points and scenarios at which consumers might open up Mappsto.re.

Next was an exercise in critical thought - how do they currently 'hack' to fulfill the functionality that Mappsto.re will offer. Is what I'm offering going to be that much better than their current process?