T-Mobile Is Running Research That Wasn't Possible Before

From Knit's session at Quirks Chicago 2025 with Tara Kenneway, Senior Brand Strategist & Insights, T-Mobile

‍

‍

T-Mobile's insights organization is not small. Ninety-eight people. Multiple sub-teams. A remit that spans the Magenta and Metro brands, home internet, fiber, finance products, and devices most people don't even know T-Mobile makes. The questions never stop coming — and increasingly, neither does the expectation that answers arrive in days, not weeks.

That gap — between what leadership demands and what traditional research can realistically deliver — is exactly what brought T-Mobile to Knit. At Quirks Chicago 2025, Tara joined Knit co-founder and CEO Aneesh Dhawan to share what two years of partnership has actually unlocked.

WHAT DREW T-MOBILE TO AN AI-NATIVE AGENCY

Trust first, technology second
‍When T-Mobile started evaluating AI research partners, they weren't just looking for speed. They had serious questions about data governance — what they could share, with whom, under what conditions. And they were specifically drawn to agencies that had built their model around AI from the ground up, rather than bolting it on.

"We were really curious to talk to folks like Knit where they were building it from the ground up from the AI angle — and it started to feel really different."

— Tara Kenneway, Senior Brand Strategist & Insights, T-Mobile

What sealed it: Knit's researcher-driven model, which embeds a senior researcher into every client relationship to build context over time, then uses that context to guide the AI across every step — from scoping through reporting. The output doesn't feel like it came from a machine. It feels like it came from someone who's been paying attention.

BUSINESS OUTCOME #1

Friday ask. Monday report.

The scenario is familiar to anyone who's ever worked in insights: senior leadership surfaces a critical question on a Friday, expects answers before the week is out, and the traditional options are grim. You can run an overnight poll — but it'll be gen pop, no open-ends, directional at best. Or you can push back on the timeline and lose the moment.

T-Mobile no longer has to make that call. When leadership came to Tara's team with an urgent question, they got into field that Friday night. By Monday, they had a full report — not directional data, but real findings, formatted the way their SLT expects, with the depth their leadership actually demands.

"We had a report and a depth of insights in the way that our SLT demands. Now we're starting to spoil them a little bit."

— Tara Kenneway, Senior Brand Strategist & Insights, T-Mobile

The speed isn't just about infrastructure. It's about accumulated context. Because Knit has built up a detailed understanding of how T-Mobile wants findings presented — what frameworks they use, how they think about cohort differences, what their stakeholders need to see — there's no rebuilding that scaffolding on every project. It's evergreen.

BUSINESS OUTCOME #2

Not just faster — actually better

Speed would be enough for a lot of teams. But what T-Mobile found was that quality went up, not down. One study cited in the session included over 500 survey respondents and hundreds of video responses — a genuinely substantive study, not a quick pulse check.

The depth comes from something Knit built specifically for T-Mobile: a co-developed values framework that now lives permanently in the platform. Any new study can invoke it with a click, and findings come back already structured around the dimensions T-Mobile cares about most.

"Because we've trained the synthesis to tell us what we need — the nuances between cohorts, the depth on value — that all comes back in the first round. You don't have to layer it on."

— Tara Kenneway, Senior Brand Strategist & Insights, T-Mobile
‍

BUSINESS OUTCOME #3

Research that wasn't possible before

Over the past quarter, Knit and T-Mobile have been co-building a product called Digital Personas — a tool that lets teams query simulated audiences grounded in T-Mobile's own first-party data.

T-Mobile spent 18 months assembling the largest data collection in the company's history: a single enterprise-wide segmentation model, stitched together from dozens of data sources, findable in their media tech stack for the first time. Nine distinct customer clusters. The challenge wasn't the data. It was activation — getting hundreds of teams across a massive organization to work from the same customer understanding.

"We wanted one easy interface that we could share, socialize, and democratize. Then we had this epiphany: everyone's doing it anyway. People are just setting up their own chats. At least let us build something we can have governance over."

— Tara Kenneway, Senior Brand Strategist & Insights, T-Mobile

Digital Personas is that interface. Teams can query one segment or all nine simultaneously, ask follow-up questions in natural language, and get directional reads that inform audience sizing, concept prioritization, or target identification — without consuming a full research phase to get there.

‍Audience sizing and prioritization
‍Instead of spending a full phase figuring out who an idea is best suited for, teams can identify bull's-eye and shadow targets directionally before committing to a larger study.

Testing dark projects
T-Mobile routinely works on locked-down products that can't go into traditional qual environments. Digital Personas gives them a way to vet concepts they would previously have launched on instinct alone.

Finding the data behind what you already know
‍Researchers use it as an internal cheat sheet — surfacing the specific study behind a finding they know to be true, then navigating directly to cite it formally.

A QUESTION THAT ALWAYS COMES UP

Is it replacing human research?

Short answer: no — but it is replacing phases of projects. Discovery research that used to consume phases one and two can now be compressed or skipped entirely. The savings get reinvested into more rigorous execution downstream.
‍
The integration runs in both directions. Knit is currently running primary research for T-Mobile that feeds back into the Digital Personas tool, keeping the simulated understanding of each segment grounded in fresh data. The more the platform ingests, the more precise the personas become.

"We're going to always augment. By nature AI is going to be reflective. We never see it replacing all of our research. It's really just an augment — and a very valuable one."

— Tara Kenneway, Senior Brand Strategist & Insights, T-Mobile

After about six weeks in beta, Tara's team had to start turning their phone ringers off. Multiple teams are already using it, and colleagues across the organization are lining up for access.

Want to see how Knit works?
Watch the full Quirks Chicago session or book time with the team to see what AI-native research looks like in practice.

[goknit.com →]   ·  [Request a demo →]

‍