Ten analysts a week, or one screen a day

The brief from Noogle was a question: how does the same content perform on Twitter and on Instagram, and why differently? The people asking were the Maps channel managers, who had to decide what to post and when. Their answer lived in a Google Sheet.

The sheet was only as fresh as the last time ten analysts had filled it. Each week they gathered the raw data for every country by hand, so the numbers were always a few days old and the analysts spent their week collecting instead of reading. The brief asked for a Tableau dashboard with more features, intuitive navigation and a pleasant layout, and for zero human input of raw data.

Stop collecting the numbers. Start reading them.

The beforeTen sheets, one week

Here is the week as it was and the day as it became, one card per beat. On the back track, ten analysts' exports have to arrive before anyone can compare a network. On the front, the dashboard already has both networks and three levels to read them at.

The peopleDesign for the person who presents the number

The executives the channel managers reported to were not available for interviews. So we studied them indirectly: a netnographic study of the services they read, such as Forbes, Bloomberg and The New York Times, plus focus groups and workshops with the employees who reported to them. What made a deliverable look diligent in front of that manager became our research question, and it unblocked work that stakeholder calendars would have stopped.

From the persona, shortened.

That last line set the design. Carolyn would never open the dashboard herself. Her channel managers would, and they had to walk into her office with a number they could defend and the post behind it.

The productThree levels on one screen

The hub is a single view of the Noogle Maps social footprint, with each network in its own colour and always side by side. It reads top to bottom in three levels. High: how the quarter is doing against the last. Mid: how it got there, ten quarters at a time. Low: the posts that did it, down to one tweet. The filters at the top follow you down every level, so the comparison is never lost.

Mid levelWhen to post

The view channel managers asked for most answered the brief's first question: what time to post. It splits owned content by the day it went out. On Twitter, Thursday posts earned a 0.45% engagement rate, about twice most other days, while Wednesday posts drew the most impressions and not the most engagement. On Instagram the week was flatter: Thursday and Friday led at 5.9%, Saturday trailed at 4.4%.

Mid levelWhat moved, on which network

Every chart carries its quarter-on-quarter change, because the question in the room is always "is that up?" Read across both networks and the story of 2017 jumps out. In the fourth quarter, Twitter impressions rose 296% and Instagram's average post reach rose 112%. The low level shows why: in December 2017 the account posted 55 tweets for 10.3 million impressions, more than any other quarter in the record.

Mid levelWho engages

The content tag comparison set any tag against any other: the audience who engaged against the network, the asset type against the feature mentioned. It showed a different crowd on each network. Lifestyle consumers led everywhere, but Twitter skewed technical, with 23.8% tech enthusiasts and 13.5% developers, while Instagram had the most students and educators.

The processWireframed in sprints

I ran the product as an agile project and wireframed it in sprints with the developers. The dashboard was one of four sections in a larger social analytics product: owned channel analytics, social listening, a benchmark library and the Noogle Maps benchmark. Each opens on an overview, with deeper views behind the arrows on the left, so every section is learned once.

The outcomeContent made for each network

With both networks side by side and fresh every day, the differences stopped being a finding in a report and became a working habit: channel managers began to make content for each network. The analysts moved from gathering data to analysing content and strategy. Four months after launch, engagement on both networks was up between 25% and 45%.

The engagement result is my own account; I do not have Noogle's own measurement to link.