Most teams still talk about adding AI into their product. Intuit quietly showed a different direction. They put their product inside AI.
With a multi-year agreement worth more than $100 million, Intuit is bringing TurboTax, QuickBooks, Credit Karma, and Mailchimp directly into ChatGPT. A user can ask for a tax estimate, check upcoming invoices, or trigger a marketing campaign in a simple conversation. No forms. No navigation. Not even a login screen in the traditional sense.
This shift seems small on the surface. It isn’t. It changes where workflows begin.
People are increasingly starting work inside AI surfaces. ChatGPT, Microsoft Copilot, Salesforce’s conversational layer, GitHub Copilot. These tools are becoming the place where the first question is asked.
Intuit recognized something important. The user’s question is the new homepage. If your product is not present where the question begins, you are already behind.
Rather than insisting that users open a dashboard, Intuit is meeting them at the conversational entry point. It creates a more fluid workflow. In many cases it saves minutes of context switching. And maybe more importantly, it positions Intuit closer to the moment of intent.
Once the starting point moves, ownership of the workflow moves with it.
For twenty years, SaaS distribution was predictable. Websites, mobile apps, paid channels, and partner integrations. Now another layer sits between the user and the tools they rely on.
AI platforms are becoming that layer. They are not simply places to generate text or automate tasks. They are becoming the first interface for work.
Stripe already has a similar pattern. Developers can ask real questions about payments inside ChatGPT. Snowflake is moving in this direction through conversational data access. Morgan Stanley advisors query research through an internal assistant rather than navigating a labyrinth of dashboards. Even Microsoft is pushing teams to use Copilot as the primary interaction model for many day to day tasks.
The common thread in all of this is simple. The application no longer insists on being the front door.
Enterprises typically adopt technology in phases. Exploration. Pilot. Platform integration. Broad deployment. Most companies are still somewhere in the middle. Intuit’s move shows what the later phases look like.
Several things begin to shift once AI becomes an intermediary.
The interface becomes conversational without requiring training. Information across systems becomes easier to compare. Workflows that used to require navigating multiple tools start to collapse into a single request.
A CFO might ask for forecast versus actuals and bring in sales pipeline data. The request spans departments. The response is unified. This is the type of interaction that AI intermediaries make possible.
It also raises an uncomfortable point. Vendor switching becomes easier when the AI layer does most of the navigation. A user who never sees your UI is less anchored to it. The differentiation moves from interface design to execution quality, data accuracy, and ability to integrate.
That shift benefits the best engineering teams. It exposes the cracks in everyone else.
Startups have less brand gravity than a company like Intuit. But they have more flexibility. AI native distribution may let them bypass channels that normally take years to build.
A startup that exposes its core functions through clear APIs can appear with disproportionate strength inside ChatGPT or Copilot. A young company that presents a clean set of actions, well documented and reliable, may outperform incumbents whose systems were never built for conversational access.
This is especially relevant in fintech, healthcare, legal ops, logistics, and compliance. These are domains where people begin with a question. They want clarity before they want an interface. AI sits directly in that habit.
There is a simple way to understand what Intuit is doing. They are assuming that more work will begin with a question rather than a click.
Teams that want to follow this pattern can ask themselves a few grounding questions.
What are the simplest meaningful actions in your product.
Are those actions callable through an API.
Can your data be retrieved without complexity or manual cleanup.
Which workflows naturally begin with a question instead of a form.
Where would a conversational layer actually reduce friction rather than add another step.
These questions seem operational. They are actually strategic. They determine whether an AI platform can distribute your product effectively.
The companies that benefit from AI will not be the ones that add the most features. They will be the ones that understand where the workflow begins.
If user intent begins inside an AI assistant, distribution follows that entry point. Intuit’s agreement makes this explicit. Whether you agree with the timing or not, it is clear that conversational interfaces are becoming a strategic channel.
The next wave of product differentiation will come from how teams adapt to this. Some will move early and benefit from being present where users start their questions. Others will wait and hope the old model holds a little longer.
If you want to explore how your product could adapt to this shift or what an AI initiated workflow would look like inside your environment, we are always open to a discussion.
Let’s talk.
