Service Design · Zero-to-One · 2016–2019
TurboTax Live
Company
Intuit
Role
Staff Product Designer
Timeline
2016 – 2019
Scope
Service Design · 2-Sided Market · Zero-to-One
01

The Problem

TurboTax had spent four decades perfecting DIY tax software — and it worked. The product was genuinely good. But 60% of Americans still filed with a human accountant, not because the software was hard to use, but because the stakes were too high to go it alone. Tax filing peaks between 8pm and midnight in February. That's when a filer hits a question they can't answer — a rental property deduction they're unsure about, a 1099-G they've never seen before, an error code with no path forward. At that moment, the only options were to guess, abandon, or drive to H&R Block tomorrow. TurboTax had broken links — friction points where the software ended and a human being needed to begin — and no product existed to bridge that gap.

The brief wasn't to improve the UI. It was to invent a new category: on-demand access to a real credentialed tax professional, embedded directly in the filing experience, at the exact moment a customer needed one. We weren't improving a product — we were designing a two-sided service that had never existed before.

The Problem · Market Landscape
The trust gap that $40B of DIY software couldn't close
DIY SoftwareStrong product. Weak service. No human fallback.
CPA / Tax ProHuman. Trusted. But expensive and scheduled.
The Gap40% of filers abandoned here — no on-demand expert
TurboTax LiveDesigned to own this space
40yr
TurboTax existed before Live — no expert access
60%
Still filed with a human despite good software
2am
When the tax question hits that no software answers
The real problem: TurboTax worked. The product was fine. The failure was the service — the moment a filer hit a question they couldn't answer and had nowhere to go.
The Problem · Failure Modes
Where the happy path became a dead end
The Wall — unanswerable question in the flow
Customer hits a tax question they can't confidently answer — home office deduction, rental income, 1099-G. Software provides a tooltip. Customer doesn't trust it. Paralysis.
→ 40% abandonment rate at friction moments
Broken links — error codes with no resolution
Import errors, mismatch warnings, "your return needs review" — each message was a dead end. No path forward. No human to call. Only the option to stop.
→ 22% of sessions ended on an error state
The "just review my return" ask with no answer
Confident enough to file alone — but not confident enough to submit without a second set of eyes. No product existed for this person. They had to pay $400 for a CPA or just guess.
→ The exact opportunity TurboTax Live was designed to own
Peak filing: 11pm in February. No one to call.
Tax filing peaks between 8pm and midnight in February and March. Zero on-demand tax expertise existed. The only option was an appointment — days away. Customers closed the laptop.
→ On-demand became the non-negotiable design constraint

The trust gap: 60% of Americans chose human tax pros despite strong DIY software — the failure wasn't the product, it was the absence of a human fallback at the moment of maximum anxiety

02

Research & Discovery

We ran research weekly — roundtables with current filers, sessions with tax professionals, and a practice we called "Follow Me Home": we went to people's houses and sat with them as they actually filed. No lab, no moderation script. We watched real people, in real environments, at real moments of confusion. We built hypotheses before each session, recorded every interaction, and debriefed immediately after.

The pattern was precise. Filers moved through the software confidently until a specific class of question appeared — edge cases with nuance, deductions they weren't sure about, situations the interview couldn't fully capture. In every case, the body language shifted: they leaned back, Googled, called a parent, or abandoned. The need was never "better help text." It was a human being who could say "yes, you qualify for that" with professional certainty.

"I don't need someone to explain the question to me. I need someone to tell me I'm doing it right."
Follow Me Home — service blueprint workshop, actual sticky notes from the design process
ttlive-research.jpg

The actual service blueprint workshop — every touchpoint mapped from the customer filing journey to the backstage CPA infrastructure, using front stage / line of visibility / back stage methodology

03

Design Process

Three of us built TurboTax Live: a product designer, a director, and a content designer — and our amazing tiger team of engineers who rolled up their sleeves and got to work. The first and most important decision was to treat this as a service, not a feature. That meant a full service blueprint — front stage (every touchpoint the customer sees), the line of visibility, back stage (every CPA-side workflow), and the support systems underneath. Two products had to work in parallel. Every customer interaction needed a corresponding CPA experience to match it. The moment a filer tapped "Talk to an expert," a CPA somewhere received a session with full context already in hand: the customer's filing state, their specific question, their tax situation.

We ran the POC in a single tax season with a small pool of credentialed CPAs. Every design decision was hypothesis-driven — build, put it in front of real filers, measure the handoff moments, iterate weekly. The SmartLook feature — screen annotation during a live session, so a CPA could literally circle the field the customer was confused about — came directly from watching filers try to describe UI elements over video and fail completely.

Left: The service blueprint — customer journey from filing start to confident submission mapped across front stage, line of visibility, back stage, and support systems. Right: The 2018 POC launch — CPA Claudell live on mobile, on-demand tax expertise made real for the first time

04

Key Design Decision

Every stakeholder pushed for scheduled appointments. Predictable CPA capacity, manageable demand, safer operationally. We pushed back. The research was unambiguous: the moment of need doesn't happen on a schedule. A filer hits the wall at 11pm on a Tuesday in February, and if the answer is "book an appointment for Thursday," they close the laptop. The anxiety dissipates, they tell themselves they'll come back, and most don't. The product had to meet the customer at the exact moment of friction — or it wasn't solving the problem.

On-demand or nothing. The competitive advantage wasn't cheaper than a CPA or smarter than a tooltip — it was faster than anxiety. The expert had to appear before the filer talked themselves into abandoning. Everything else was built around that constraint.

Left: Scheduled (rejected) vs. on-demand (shipped) — the expert appears now, no appointment, no mode switch, no broken session. Right: The matching experience — CPA credentials and specialization visible before the first word is spoken, building trust before the session begins

05

Final Design

The shipped product: a credentialed CPA or tax expert, one tap away, appearing in a live session with full context of where the filer was stuck, able to annotate the screen directly, answer the question, and watch the filer complete the step in real time. From broken link to resolved — in a single session. What we designed wasn't a customer service feature. It was a new category of tax service: the expertise of a CPA at the accessibility of a click, at a price point viable for tens of millions of filers who previously had only two choices — guess alone or pay $400 and drive somewhere.

TurboTax Live — Sergio, Tax Expert, 19 years, live on screen
ttlive-final.jpg

A tax expert with 19 years of experience, live on screen, in the filing session, at exactly the moment the filer needed them — this is what thousands of on-demand CPAs delivering in real time looks like

06

Outcomes

$13M
Revenue in the POC season — three people, one tax year, first version ever built
$600M+
Revenue at full launch the following year — 46× growth from POC to scale
#1
Flagship product at Intuit — the product that redefined the entire company's direction

TurboTax Live didn't just succeed — it changed what DIY taxes meant. The product we built in one POC season with three people became Intuit's most important product, generating hundreds of millions in revenue and establishing a new category of financial service. Thousands of credentialed tax professionals are now one click away from the moment a filer gets stuck. The broken links that sent 60% of Americans to an accountant's office became the feature that kept them — and the people who would never have attempted DIY taxes at all — in TurboTax. We didn't improve a product. We changed the game.