Context
What is Plotline?
Plotline is a SaaS product that enables users to publish tooltips, modals, and bottom sheets (collectively known as in-app engagement) without requiring developer involvement
What were these in-app campaigns?
Typically, these would be onboarding journeys for users or driving new feature adoption. These campaigns could have tooltips, modals, embeds - or any combination of those.
What was my role?
Re-designing the layout and data visualization to address issues with comprehension and consumption of campaign data, along with my fellow designer and the engineering team.
The redesigned dashboard
Versus the previous version - more difficult to consume data and no visual reference to the campaign in question
The Problem
What was the problem?
Initially - the time we spent explaining the metrics on call or Slack to users, which we put up with and the project failed to get prioritized.
Over time - lack of confidence in campaign performances, since they didn't understand how metrics were calculated
Why was this important?
Lack of confidence in campaign performance meant a lack of confidence in the product and dipping CSAT scores.
This also threatened our Net Revenue Retention (NRR) on our path to $1M in ARR.
TL;DR
Users thought their campaigns weren't peforming well when, in fact, they were which led to dissatisfaction with Plotline. We just did a poor job of explaining and representing the metrics.
Goal-Setting
Since difficulty in understanding the data and lack of trust in the data were the key pillars of the problem, we articulated the goal or guiding principle as follows:
How might we make the data easier to consume and inspire confidence in campaign performance for users?
The Process
The process I followed is roughly summarized below. It doesn't come close to capturing the chaos and despair of having to convince the engineering team to build the feature though.
Research and Feedback
A collection of the insights we had gathered from scattered interviews for the Dashboard overhaul
primary users
Research for the project came from continuous discovery with users instead of a dedicated subproject, though largely speaking these were semi-structured interviews over Google Meet.
Our primary users were product managers and growth managers who:
responsible for growth metrics such as feature adoption and user conversion
using Plotline for in-app nudge campaigns
looking for obvious wins and losses in data
looking to easily share info about winning campaigns with their stakeholders
key observations
App Opens, Eligible Users, Target Users, Clicks and CTRs were all part of a funnel but the existing implementation obscured this relationship
Final Design Decisions and Changes
I. Redesigning funnel metrics, temporal views and Unique numbers
observations
Existing implementation obscured the relationship between App Opens, Eligible Users, Target Users, Clicks and CTRs - bad use of Gestalt
Also, some of these metrics were earlier only accessible by downloading a CSV and performing calculations.
decisions
Used Gestalt principles of Continuity and Similarity to establish a relationship horizontally (conforming to the typical flow of time from left to right).
Also gave more importance to Unique Users - which ultimately mattered more to client teams over Total Counts - in the Information Hierarchy
Versus the previous version - more difficult to consume data and no visual reference to the campaign in question
Expanded View
Versus the previous version - more difficult to consume data and no visual reference to the campaign in question
Default: Collapsed View
Versus the previous version - more difficult to consume data and no visual reference to the campaign in question
II. Explanation of Key Metrics
observations
A challenge for users was that they did not understand some of the metrics used to measure a campaign’s reach. They had to read through lengthy documentation on the knowledge center.
decisions
This was solved by explaining these metrics visually inside the product along with hypothetical numbers.*
*These explanations use hypothetical numbers due to the increased cost of dynamically generating them for each campaign without latency
Versus the previous version - more difficult to consume data and no visual reference to the campaign in question
Versus the previous version - more difficult to consume data and no visual reference to the campaign in question