Transformational change is difficult, whether inside a large corporation or a startup. The challenge is always the same: how does an organization rebuild its processes and technology platforms while still executing on what is due today. This is the story of Morgan Stanley's Next Best Action platform, from incubation through launch, and the balance it took to get there.
Every day, the recommendation engine had to process millions of client accounts, each holding hundreds of financial instruments, for tens of thousands of financial advisors, prioritizing thousands of relevant content pieces and alerts generated daily, all while maintaining compliance and control.
Morgan Stanley's Next Best Action platform has been covered in the press. It's an internally-built AI engine that delivers timely, customized messages to clients and prospects, guided by the Financial Advisor.
When the goal was conceptualized in 2009, the inspirations were Netflix, suggesting the next movie to watch, and Amazon, suggesting the next product to buy. But the next best action in financial advice had to clear a higher bar: more accurate, more compliant, more sophisticated. A suggestion based on preference or past behavior alone would often be inappropriate.
Picture the world in 2009. BlackBerry and Nokia still outsold Apple. Processing large volumes of data was hard enough that the industry just called it "the Big Data problem." Conversational AI lived in science fiction. Getting from there to here took a series of changes across many years: organizing content, data, advanced analytics, and human workflows to the point where, today, a conversational interface can draw on all of it to hand a financial advisor the next best action in the context of their specific client's situation.
Executing transformational change and continuing business as usual requires a fine balance of business choices.
Leadership
Morgan Stanley's leadership had a clear vision: where they wanted the organization in ten years, and why the change mattered. That clarity came with a matching commitment: the investment was treated as a force multiplier for the future, not something to justify against next quarter's top or bottom line. The budget was set against near-term financial goals instead, sized to what the business could reasonably commit rather than what a direct ROI case would have demanded.
The early initiatives were incubated deliberately, insulated from the pressure of disrupting business as usual, but not isolated from it. Core business teams were involved in the design from the start, since this was their roadmap too, just further out. And the project was led by top talent chosen specifically for it, with a small core team built to get an MVP into the world.
Design & Execution
Execution began with mapping the current workflow. The surest way to drive adoption is to minimize disruption to what people already know how to do. Alongside that sat a long-term architecture, clear enough that everyone building toward it understood the destination, even while the roadmap to get there stayed modular.
A modular approach simplifies the execution roadmap while retaining flexibility as solutions evolve. Staying true to the long-term architectural vision is key.
Not every gap got a permanent fix. Band-aid solutions, backed by robust and scalable operational processes, bridged the distance between what the long-term architecture called for and what the modular build could deliver in the moment. And throughout, a single operational owner sat inside the incubator, accountable for the end-to-end process until it matured enough to hand off to the core business.
People & Communication
None of it worked without constant communication: the vision, the approach, the status, repeated often enough to cut through the ambiguity that comes with any change effort. The more people who understood where things stood (senior leaders, the incubating team, the supporting organization around it), the more support the effort had to draw on.
Feedback needed to arrive early, too, without putting anything at risk. In financial services, robustness isn't optional. A modular approach made room for real feedback while keeping the exposure of any one experiment small.
What This Means for Growth-Stage Companies
Growth-stage companies rarely face a challenge at Morgan Stanley's scale. But they face its shape constantly. A cross-functional project (the kind that touches sales, finance, product, and technology all at once) routinely stalls not because any one team resists it, but because no one owns the coordination between them. It's the same gap covered in other Insights pieces: no single department can solve a cross-functional problem alone, and assuming otherwise is usually the first mistake.
The other half of the balance is designing for a future the company hasn't grown into yet. It's tempting to build only for what is needed this quarter, and just as tempting to over-architect for a scale that is still years away. Morgan Stanley's approach split the difference: a long-term architecture clear enough to hold the vision, built through short-term, modular steps that could adapt as the picture came into focus. That is the same balance every growth-stage company is weighing when it decides how much infrastructure to build ahead of the need it's serving today. The operating model that got a company here rarely survive contact with what is next. The discipline is building toward the next stage without freezing the one you're standing on.