AdRoll Insights was a tool that helped bridge the gap between deep valuable data and new possibly inexperienced users of the AdRoll Dashboard.I planned it using a very similar framework to Google Analytics insights in which there is a right-hand pop-out bar that categorizes and prioritizes natural language questions about the data that can be found all throughout the dashboard. Questions like what advertising channel did most of my sales come from?AdRoll Insights does exactly that, we prioritize the most important and valuable questions that experienced users of the dashboard have, build the queries that answer those questions then surface them in a right-hand pop-out navigation with links in each question to the relevant report to also help up-level the newer users.
Before digging into a solution for the problem of depth and speed of data in the dashboard I met with experienced users of the dashboard to learn about what was most important to each user (at least 3 AMs, 3 Marketing Managers, and 3 business owners + my years of personal experience)
We identified that metrics directly tied to revenue weighed most heavily on crucial decision making and should be quick, easy and clear to find.
I research other data rich platforms to see how they may have solved for this already and found that Google Analytics had a unique pop out navigation to surface important actionable quesitions/answers.
Many platforms lean into secondary navigation for hidden insights.
These insights need to be: - Easy to find - Easy to Access - Easy to Use - Quick to understand - Prioritized and minimized (easier to contemplate the opposite of too much information)
On brand Clean, clear Professional Non-intrusive to the normal flow of the dashboard
Visual Studio for the Chrome Extension (HTML JS) Redash for the Queries (SQL) Photoshop and Illustrator for design elements
In this MVP I built it in a Chrome extension referencing account and insight specified Redash Queries per item and it was connected via a query-specific API. The account information is variable on the ADV EID in each query so that that could be dynamic and plug into flexible queries.
AdRoll Dashboard users who want to get the most of their data without spending the time to learn the depths of the dashboard.
Get the data that is most important to you quicker and easier than ever before.
Save time Better informed decisions Learn the dashboard faster
pop out navigation with easy to understand natural language insights in the form of questions/answers.
This would likely be rolled out in a subset of accounts (5) with as close of a profile to our ICP as possible. (ideally frequent users) After 30 min of usage in the dash (or 2 weeks whichever comes first) have a retrospective feedback session to learn what went well and what needs improvement Then gradually increase user base to include a greater variety of users (starting with managed accounts moving out to self-service) Regularly scheduled retrospectives even after full release using feedback and the analytics data to inform iterations
No specific time goals
time and role-specific insights i.e. Q4 advertising may surface different insights and it may vary for Marketing managers vs Business owners A library of queries that auto-sort and filter via a machine learning model that pivots on the role, time used, and interactivity with the dashboard
No specific time goals