Customer Data Platform

Project summary

My proudest accomplishments with this project was leading the end-to-end design for a highly technical and complex product, establishing relations with the product team as their first design partner, and advocating for user research that prevented us from releasing features to the wrong set of users.

Problem

Data integrations were managed solely by the product support team, which was a security risk and cut into the company resources

1. Strain on resources:

   1. Support team didn’t have enough team members to accomadate volume of customer requests

   2. Communication between support team and customers was slow and inefficient

   3. Even the support team struggled to understand the UI, requiring them to frequently revisit internal documentation

2. Security Risk:

   1. Customer password credentials were known to company employees

   2. Credentials were being stored on third party platforms rather than securely in-app

Solution

The proposed solution was to have integrating data be a self-service feature – giving users the autonomy to do this on their own

Simply changing the permission settings to allow users to access this feature would not make this solution successful. This was because our users are less technical than our internal teams who have originally done this task for years.

Goals

Ideally, this solution aligned with both the business, product, and user needs

• User Goal(s)

  • Understand how to create/edit new output connectors with little to no intervention from the company support team

  • Successfully provision credentials for external data systems with little to no intervention from the company support team

• Product Goal

  • Significantly reduce work load on Support team around output connectors and credentials management

  • Mitigate security risk around storing customer credentials

• Business Goal(s)

  • Be on-par with other competitors in the market that also offer this process as self-service

Discover

Difficulties recruiting technical users led way to internal research to better understand hypothetical expectations and workflows

I had many assumptions throughout this project, and the tricky thing was that it was hard to recruit for these extremely technical users to get primary data on their expectations and technical understanding.

I strongly advocated for user research, so I brainstormed with product and UXR to determine which set of users would best represent these technical users: the internal Support team and Customer Success Managers

Define

Internal research highlighted where the main pain points existed, leading way to a new, ideal experience to design for

Old Workflow

New Workflow

Wireframes

Because there were 15+ different types of integrations, I had to make my designs scalable and reusuable to minimize design & engineering effort

My goals for wireframing was to determine an overall, scalable layout and to simplify the technical content by workshopping the copy with PMs and engineers.

High fidelity

Aiming for scalability & design and development efficiencies, I identified reusable patterns

Railroad and Navigation

There were currently 15+ different output connectors that I categorized into 4 main types and at least 6 ways to provision credentials. Instead of following the old experience of creating separate workflows for each, I designed components could be reused consistently across all of these different variations. This consistency not only streamlines my design work but was also used in the developing design system and established a consistent and familiar user mental model.

Progressive Disclosure

I discovered during my scrappy UXR sessions that users felt overwhelmed with the volume of parameters and information that were presented to them. Very few users used all parameters, so I analyzed qualitative and quantitative data to categorize the formatting fields into two sections and reduce cognitive load: basic (most commonly used formats) and advanced (least commonly used formats).

Small but impactful guardrails were used to increase user confidence and minimize user error

I explored ways to design and implement contextualized help to users because this was such a technical workflow. These small changes delighted users in usability testing, confirming my hypothesis that these small changes had a big impact in user confidence.

Introducing friction to increase user comprehension

Increasing the steps within the flow, and thus increasing the interaction cost, was a design trade-off to allow for chunking - a psychological principle that breaks up information into smaller pieces to allow for better comprehension. User comprehension was essential for the user task rate to be successful. I was also aware to keep the new number of steps within the recommended 7 +/- 2 as described by Miller’s Law.

Outcomes & metrics

• task completion rate for support team: 89-100%

• task completion rate for customers: 22-89%

  • low completion rate pivoted product strategy to only release certain steps of the flow to users

  • retain most technical step for internal/support team

• customer satisfactory scores

• support team satisfactory scores

• trust and partnership built between design and product

• designs have been shipped in Q4 2021