There’s No GLP-1 for Digital Transformation
By Erin Hennessy, Chief Member Experience Officer at PSECU
Simple Subscribe
Subscribe Now!
Most people know what it takes to get healthy. Eat vegetables, get enough sleep, exercise, and take mental health breaks. None of it is secret and none of it is quick, which is why an entire industry sells cleanses, overnight fixes, and miracle shots that promise results without the work.
Financial institutions approach digital experience the same way. Leaders know the fundamentals, then buy the newest tool anyway and wonder why it underdelivers. The cause usually isn’t the technology itself, but the data underneath that lives in silos and means something different to every team that touches it. Unhealthy data undermines every digital investment built on top of it, and there’s no GLP-1 for clean data.
Credit unions are starting to feel the burden. They still lead banks on overall satisfaction, scoring 725 on a 1,000-point scale compared with 657 for retail banks, according to J.D. Power’s 2026 U.S. Credit Union Satisfaction Study, but that lead is slipping. Member satisfaction fell four points from last year, and 59% of members now hold a checking account at another institution. J.D. Power calls the pattern “soft switching,” in which members open second and third accounts elsewhere and gradually move money into them.
Key insight: Digital channels are where the weakness usually appears. Banks outperform credit unions on mobile app performance and website satisfaction, according to the American Customer Satisfaction Index, and credit unions fell further behind banks on overall satisfaction.
The usual explanations are familiar: Credit unions are risk-averse, big banks are bureaucratic, and fintechs are fast. The industry has repeated these beliefs long enough that we now accept them as facts and use them as a ready excuse. Speed comes from operating choices any institution can make at any size, and those choices depend on healthy data. When pursued together, data transformation and digital transformation become an experience transformation that members feel every time they log in.
Speed Is an Operating Choice
Fintech speed is far less glamorous than its reputation. It comes from a series of operating choices repeated every day, starting with a small number of clearly prioritized outcomes, empowered owners for each one, and cross-functional teams that stay together after a launch. Add fast access to clean data and short approval paths, then layer extensible platforms and testing rigor on top. That speed comes from removing ambiguity and handoffs. It requires neither longer hours nor bigger risks.
Many institutions build invisible barriers for themselves, adding layers of process and approval that exist only because no one questioned them. Those processes often slow decision-making and blur ownership. The alternative is to hire strong people, give them data and resources, and let them go. They will make mistakes, as everyone does, but they will learn from those mistakes faster than any committee would.
Key insight: I’ve seen fast execution at very different scales, as senior vice president of marketing at Rocket Mortgage, at credit unions, and while helping launch First Women’s Bank as a startup. Size had little to do with how quickly any of them moved.
What A Culture of Delivery Looks Like
Rocket organizes its culture around philosophies it calls ISMs, many of which focus on getting things done. One reads, “Innovation is rewarded. Execution is worshipped.”
Another, “Launch and learn,” holds that taking longer to make a decision doesn’t make the decision better. Credit union values tend to sound different. “Be trusted” is a common one, and although there is nothing innately wrong with this, trust is an outcome, not a goal. Members trust an institution that consistently delivers, keeps trying new things, gives its teams ready access to data, and runs on resilient systems. Those are the how and why behind trust, and nothing stops any financial institution from building them. The only barrier is human choice.
PSECU has a long history of moving early. In 1995, it mailed floppy disks to members so they could bank from home. Its recent selection of Candescent’s digital banking platform, following a competitive evaluation, marks the start of a multi-year effort to build more personalized, connected member experiences. The working posture is to launch, listen to members, and change course when needed.
AI Won’t Fix Bad Data
AI is the newest shiny object, though financial institutions have been using it for years.
Most of that AI is modular, built into a specific product or deployed as a chatbot, and highly programmed. According to Wipfli’s 2026 State of the Credit Union Industry report, 79% of credit union executives are researching or implementing AI, largely through tools embedded in existing vendor platforms. Wipfli noted that many credit unions use AI simply because it’s available, without a deliberate strategy behind it.
AI can only work with the data it is given; without a sound data foundation, even the best vendor tool will underdeliver. The way to catch up is to ignore the noise around each new tool and fix the foundation first.
The results so far are mixed. Only 28% of national bank and credit card app customers use virtual assistants, according to J.D. Power’s 2026 digital banking studies, and satisfaction with those assistants drops sharply when customers try to resolve a problem, dispute a charge, or report fraud. Those are the moments when people decide whether to trust an institution, and having a human at the helm makes them feel reassured.
Agentic AI raises the stakes on data. Bad data used to only produce a bad report. Agentic systems feed each output into the next decision, so ungoverned, messy data compounds its errors with every cycle. Even at a fintech the size of Rocket, data arrives from many sources, each source defines and stores it differently, and different teams interpret it differently. Cornerstone Advisors’ Ron Shevlin put it plainly in the firm’s 2026 banking report, writing, “There is no AI strategy without an effective data strategy.”
Key insight: Institutions don’t need to wait for a perfect future-proof architecture to begin improving how data is governed, accessed, and used. At PSECU, we’re building data architecture that is accessible, reliable and usable across the organization to improve the digital interactions our members experience with us every day. Treating data and digital as separate initiatives was always the wrong conversation. Together, they make up the member experience.
Key insight: All of this comes back to the fundamentals. Data structure, governance, and resiliency are the digital banking equivalent of eating well and lifting weights. The work is unglamorous, and the payoff takes time. Once the foundation is healthy, though, it gets exciting because every new tool built on top of it can deliver its full impact.
Credit unions are not destined to move slower than banks or fintechs. How fast an institution moves depends on the choices its leaders make, starting with the next one.
