The stakeholders liked the ideas. They were happy with the research and encouraging about the products taking shape. Then Bob Petrie's team asked a different question: would the business spend its own money on the next step? Too often, the answer was no.

For Petrie, Director of Innovation at Citi Ventures at the time of this conversation, that gap exposed a problem with strategic alignment. An innovation team could have enough independence to explore promising opportunities while still being too far from the priorities of the business expected to adopt them.

In this Innovation Rockstars episode, he explains what the team learned from that experience. After nearly two decades in startups, Petrie brings a direct view of the work: corporate innovation requires entrepreneurial judgment, but it also needs sponsors, controls and a credible route into the organization that will carry the business forward.

Stay independent enough to explore, connected enough to matter

Petrie distinguishes the incubation work from venture investing. While the investment side spends much of its attention outside the company, incubation depends heavily on understanding internal business partners, their strategies and the disruption they face.

Sitting outside the operating business creates room to pursue opportunities that struggle to compete with immediate delivery needs. An operating team has customers to serve and an existing business to maintain. Money available for those demands is difficult to divert toward an uncertain future opportunity.

But independence can become distance. Petrie recalls projects receiving positive feedback while the incubator funded the work, only to encounter resistance when a sponsor was asked to commit resources. The concept did not fit that sponsor's strategic direction closely enough.

The response was to bring the program nearer to the business without letting it become absorbed by day-to-day production priorities. Governance and sponsor input needed to influence the work while choices were still being made. Agreement that an idea looks interesting is a weaker signal than willingness to take responsibility for its next stage.

Accept the constraints, then remove avoidable friction

Petrie arrived after a demanding startup turnaround. He had expected a relatively short corporate chapter before returning to another venture, but found he enjoyed the opportunity to combine entrepreneurial work with Citi's capabilities and reach.

He also quickly discovered that building inside a bank was not simply running a startup with a larger balance sheet. Legal, risk, compliance and other control partners brought perspectives that had to be incorporated into the product and its route to market.

He challenges a familiar analogy that brakes exist to help a car go faster. In his view, some processes are intended to slow work down, and that is appropriate when the institution's obligations require care.

“So we want to move as fast as possible, but not fast enough to break things.”

The task is to distinguish necessary constraints from other causes of delay. It requires understanding how the organization works, where decisions sit and how a project can move from the incubator into the business. Startup experience helps, but internal knowledge is equally important.

Design the smallest useful test for this environment

Petrie does not claim to have a simple formula for an MVP. He describes the decision as a judgment about the smallest request that can produce enough information to move forward.

That begins with identifying the main assumption being tested and resisting the tendency to add features. It also requires acknowledging the context. A control partner may need something included that an independent startup would not consider essential to its experiment.

Because putting a product into production carries substantial effort in his environment, the team does more learning beforehand. Qualitative research, quantitative research and clickable prototypes help examine a concept before committing to that step.

For R&D leaders, the lesson is to make the purpose of the experiment explicit. What uncertainty is it intended to reduce, and what would the result change? Calling something an MVP does not resolve those questions or remove the obligations surrounding the test.

Keep activity measures in their proper place

Measuring success presents another tension. An incubator wants economic impact, but decisions outside its control can prevent a project from reaching the market.

Petrie describes trying preliminary indicators such as problem spaces explored, prototypes developed and papers written. Those measures captured work the team could control, but he worried that they could draw attention away from the intended result.

“Those are necessary, but not sufficient to say we've been successful on a program.”

Projects need to progress across several dimensions, including funding, customer engagement, technical feasibility and strategic alignment. The emphasis can vary by project, but ultimately the work needs a route into the business.

That is why he talks about projects graduating from the program. The incubator must help something become useful beyond its own activity, then be able to turn its attention to another problem.

Notice when the machinery finally works

Petrie's standout example is Bridge Built by Citi, a platform connecting small businesses with small and midsize banks. What made the launch memorable to him was how smoothly the pieces came together.

Funding, the sponsor relationship, control-partner involvement and the move into production aligned. He saw the experience as evidence that lessons from earlier projects had been applied. He also valued the participation of minority depository institutions, describing an ambition to combine economic and social benefit.

His account does not make innovation sound effortless. It makes the less visible work recognizable: developing relationships, resolving requirements and keeping a sponsor involved before the handover arrives. After years of grinding through uncertainty, a project running smoothly was itself an achievement worth remembering.