Posts Tagged ‘Adam Clarkson’

The Role of Research in Building the New Chartbeat Publishing, Part 2

December 6th, 2013 by Adam

A quick recap of Part I of this post…
  • We knew Chartbeat Publishing was strained by UX debt
  • We were psyched to introduce some major new functionality into the product
  • We knew that we had to consolidate all our research and fill in knowledge gaps before the full design and dev process
  • We were working out of a glorified bomb shelter next to a demolition site, an atmosphere of adversity which likely hastened our eagerness to redesign

The research effort culminated in the construction of a massive affinity diagram or mental model, which neatly organized all of the chaos of a newsroom into a taxonomy of actions and behaviours. The top level of the mental model consisted of four main categories:

  1. Developing content - actions associated with actually creating content
  2. Assessing content - actions associated with measuring traffic to content
  3. Assessing audience - actions associated with measuring identity and quality of traffic
  4. Developing audience - actions associated with systematically building audience

Taking a look at the mental model, the “developing content” and “assessing content” categories were fairly concise. We had a pretty thorough understanding of the workflows, processes, and product opportunities. But for “assessing audience” and “developing audience,” things were a little murkier. There were a myriad of complex activities that seemed disorganized and in need of rationalization — we had unearthed all kinds of social media tricks and hacks, experiments in link partnerships, attempts to infiltrate Reddit, newsletter optimizations, Outbrain gambits, and a whole slew of other tactics.

And the survey data backed up our feeling that there were more people working on audience development and using Chartbeat than we had originally thought.

We reached two conclusions:

  1. We needed to sit down again with the publishers doing the most progressive work in the area of audience development and try to figure out what we’d missed, if anything, the first time.
  2. And, in parallel, we needed to prototype some ideas that came out of our own hypotheses about how to measure audience quality in a simple way.
Read the rest of this entry »