Client Market
Research
Led the design of a focused MVP, utilizing a multi-stage feedback strategy to earn buy-in for the full vision.
- Role:
- Lead Designer
- Team:
- 1 Designer, 1 PM, 1 Eng
- Company:
- GoDaddy
Summary
Problem
Pros weren’t using the Hub to manage their clients.
As a growth team we saw an opportunity to build value by providing tools that support the start of a client project lifecycle. The vision consisted of an AI powered business dashboard that could generate and track helpful documents like proposal contracts, project summaries, and industry research.
Approach
Start small and build signal to expand.
I worked with the team to narrow scope to a single AI powered document generator, I strategized how user feedback could gain buy-in from stakeholders to build out the full vision, and I conceptualized how the future vision could look.
Reflection
The tool felt awkwardly placed on it’s own.
Discovery on the tool was low, and it would feel awkwardly placed until we could change more of the client dashboard around it. In retrospect making changes to the structure of the existing client dashboard up-front would have allowed us have scaled faster post-launch.
Outcome
Product
The Hub, a variation of a customer's GoDaddy account that has features and tools tailored for tasks related to client and site management.
What we shipped
A tool that generates market research for the client project, part of "scoping the work." A Pro enters a brief description of their project and our AI system generates market research on the client’s industry.
Concept work
For this project I explored what the client space could evolve into: a full client-project lifecycle management experience, with multiple tools supporting our users at every stage of their work.
Original
The existing experience had many features but no clear hierarchy, making it hard to introduce new tools without adding to the confusion in the space.
Logging in as client is the main action on this page but nested within a menu. Too much going on in the header creates cognitive overload.
The dashboard is cluttered with tools that don’t have a clear mental model for the user to follow.
Concepts
The goal with the concepts was to reduce cognitive load, centralize the most important features of the existing client dashboard with AI document generating tools, and structure the dashboard to fit the entire model for a client project lifecycle.
“Project Companion” as a dedicated space for the AI document generator.
Creating a details card as the main card on the page allows us to clean up the page header.
All other tools fit in a project space organized in order of tasks in the client project lifecycle.
Feedback strategy
As a small growth team that needed to move quickly we narrowed in to a few key tools and then ran lots of user testing to help build a business case for the broader vision.
The strategy had three main types of tests:
- Quick task-based prototype tests for signal.
- Live demo account tests for quality of AI outputs.
- A feedback module with the first launch for continued learning.
Successes from this strategy:
- Momentum: By running signal tests early and often, then sharing back with stakeholders, we were able to gain momentum for the bigger vision and get buy-in for more resources to expand.
- Accelerated learning: Testing with real AI outputs (not mocked data) caught quality issues we never would have found in a prototype test, and gave stakeholders something concrete to react to before launch.
Outcomes
Fast follow: A barrier to using the tool was needing a client dashboard first. As a fast follow, we added the ability to spin up a "sample client" so Pros without existing client dashboards could try the tool immediately. That single change drove an 800% increase in new client instances, and became a key learning for future work on activation and entry points.
Reflection: Discovery on the tool was low, and it would feel awkwardly placed until we could change more of the client dashboard around it. We knew this might happen in the early stages of planning but took the chance anyway. In retrospect with some extra vision work up front we could have scaled faster post-launch.