Client engagement · PRODUCT OPTIMIZATION · MOBILE SELF-SERVICE · ITERATIVE RESEARCH
AT&T Device Help
The app was already on millions of phones. The harder problem was helping customers finish troubleshooting without calling for help.
AT A GLANCE
I led a turnkey product and research optimization engagement, including expert review, customer testing, prototype iteration, a practical roadmap and a validated V2 direction delivered in 65 days.
The challenge
Reach was there. Self-service was not working hard enough.
Customers were abandoning troubleshooting in Device Help and calling Customer Care instead. The product had broad distribution, while the team needed to understand where people got stuck and which changes could ship within the existing technology stack.
My role
Sole practitioner, working with a cross-functional team.
I owned strategy and end-to-end research delivery, partnering with product, research, design, engineering, legal and third-party technology stakeholders. Requirements workshops established business goals, technical constraints, support needs and success measures before customer-facing decisions.
How we worked
Three research flights in a 65-day decision cycle.
I reviewed findability, comprehension, navigation, workflow, feedback, error prevention and confidence. I coordinated more than 30 remote mobile usability sessions across the United States, grouped into three iterative flights. Findings informed prototype changes between rounds, so each cycle tested a more focused direction. A 30/60/90-day roadmap connected customer impact with business value, feasibility and speed.
What changed
A tested V2 direction the team could deliver.
The team gained a ranked view of the problems, a prototype shaped by repeated customer feedback and a production direction grounded in technical reality. V2 launched in under 60 days according to the broader portfolio record; the approved case record describes a 65-day engagement window.
Reported results
Resolve two different performance measures before using a single headline lift.
The source archive contains two post-launch figures with different labels: a 46% increase in app usage and a 35% increase in self-service. Until their definitions and measurement sources are reconciled, this page does not present either as a single verified performance claim. Fewer support calls and improved app-store ratings are also reported without a preserved baseline.
What this demonstrates
Research can be part of delivery, not a gate at the end.
This case shows rapid mixed-method diagnosis, iterative mobile research, prioritization under technical constraints and a clear path from customer evidence to a shipped product direction.