Intuit: Saving PM’s 8 Hours Per Week Through AI Workflows

We all want to use AI more. But how do you turn the promise of AI into something that saves product managers hours every week? That was the challenge Intuit, the company behind TurboTax, QuickBooks, and Credit Karma, brought to Irrational Labs. 

We designed a 6-week AI transformation program that combined hands-on learning with behavioral science strategies to increase the speed and productivity of Intuit product teams using AI.

The results were immediate: 

• 100% of participants said they now use AI more effectively. 
• 85% reported using it more often
• On average participants said they saved 8 hours in the week after the training.

Scaling an approach like this across Intuit could shift how entire teams adopt and apply AI in their work.

The Opportunity

Going beyond the chatbots

Intuit saw the potential of AI, and product managers were already experimenting with chatbot tools. The challenge was not sparking interest, but helping PM’s move beyond one-off uses into deeper, more valuable applications. 

Our pre-course survey underscored the gap between experimentation and lasting impact: half of PM’s estimated that AI could accelerate 50% or more of their weekly work. Yet most were using it only for lightweight tasks such as meeting notes or announcements, and fewer than 30% reported feeling confident applying AI to workflows.

The opportunity was threefold:

1. Embed AI directly into daily workflows and operations
2. Share efficiency gains across teams instead of keeping them siloed
3. Focus on high-leverage areas where AI could deliver outsized wins

There was no lack of desire. PM’s were hungry for this shift, but like most organizations, the reality of day-to-day priorities got in the way. It is always faster to stick with the status quo than to reinvent a process and get good at a new one.

Barriers that stood in the way:

• Time to figure it out: PM’s didn’t have the time to sort through which tools to use or build the habits to make them stick.
• Space to create: Without room to experiment, it was hard for PM’s to see how AI could skyrocket productivity or open up new possibilities for their work.
• Lack of coordination: To unlock real gains, teams needed to be moving together, not experimenting in isolation.

The Solution

Building AI Habits Through Hands-On Learning

To overcome these barriers, Irrational Labs designed a 6-week pilot program for 20 Intuit PM’s across multiple teams. Our approach was rooted in behavioral science and focused on a learn-by-doing methodology that reduced friction and built lasting habits.

Our training program was built on four key pillars:

1. Optimize your use of the LLMs for your key jobs: Prompt techniques  are still the foundation of effective and high impact output. For example, we taught PM’s six strategies for crafting clear, context-rich prompts that generated higher-quality prototype solutions.
2. Automate via assistants: Participants learned to design reusable, custom Gems tailored to their specific needs, automating tasks like drafting PRDs and synthesizing research.
3. Redesign workflows: Every participant  identified multiple opportunities to streamline their daily and weekly workflows. By the end of the program, they may progress on building these new flows.
4. Form lasting AI habits: Using proven behavioral science principles, we provided the recipes to turn experimentation into consistent habits across the organization. 

Participants got hands-on experience with a curated set of tools, including Gemini, Gems, Figma Make, ChatGPT Agents, and Zapier.

The sessions empowered PM’s to build functional prototypes for real-world Intuit challenges, including:

• Competitor pricing research in 3 clicks: An automated agent that gathered, analyzed, and visualized competitor pricing data.
• Customer feedback synthesis and roadmap creation: : Teams built a new workflow that automatically collected, categorized, and synthesized multi-channel customer feedback. This was then turned into a structured product roadmap summary. What once took hours of manual work now took minutes. 
• Zero to one consumer prototype: New product creation used to take weeks or months. In these sessions, teams produced functional, quality UI mockups for new, customer-facing tools. 

Teams discovered two things. The first was expected: They sped up or eliminated some day-to-day work altogether. The second was less expected: they found completely new ways to add value to the organization that wouldn’t have been possible before AI.

The Impact

PM’s Reported Saving 8 Hours in a Week Through AI Workflows

The program delivered immediate results, transforming how Intuit’s PM’s work and unlocking significant productivity.

Key outcomes, according to our pre- and post-course surveys:

• Participants reported saving an average of 8 hours in the prior week by integrating AI into their workflows.
• 100% would recommend the course to their colleagues.
• 100% agreed they now use AI more effectively in their work.
• 100% said AI helped improve the clarity and thoroughness of their deliverables.
• 85% reported they now use AI more often.
• Among the participants who completed both pre- and post-course surveys, tool adoption increased significantly:
  • Figma Make: 15.4% → 69.2% (+53.8 percentage points; about a 350% relative increase).
  • Loveable: 15.4% → 46.2% (+30.8 percentage points; about a 200% relative increase).
  • ChatGPT: 84.6% → 92.3% (+7.7 percentage points; about a 9% relative increase).

Lessons Learned

How to Make AI Adoption Stick

This pilot points to several principles for building AI adoption inside organizations:

• Hands-on practice drives change. PM’s learned more when they worked directly in tools than when they only observed demos. Training that gives people space to practice leads to faster adoption.
• Coordinated structure supports consistency. A weekly rhythm with a group of people and repeated exposure helped PM’s turn experimentation into daily use.
• Focus on high-value, high-friction areas. Automation workflows are harder to learn, but are also where teams can unlock the biggest gains.
• Small frictions block adoption. Tool access and setup need to be addressed up front to make it easy for teams to apply what they learn.

These lessons extend beyond Intuit. They show that successful AI adoption requires behavioral design that makes AI easy to try, repeat, and use every day.

Looking to scale AI adoption across teams? Contact us: [email protected]