The Platform Pivot Dies Indoors

Based on the research ofFraser, Altman and Özcan, "Platform development in incumbent firms: The interplay of data and actor interests in adding platforms to existing businesses," Strategic Management Journal, 2026

Fraser, Altman, and Özcan followed a financial incumbent as it tried to launch a platform connecting small and midsize businesses with lenders, work published in Strategic Management Journal. The firm had the strategic rationale, the executive sponsorship, and the technical roadmap that most accounts of platform transitions treat as the hard part. What actually shaped the project, according to their process model, was a sequence of five data bottlenecks: constraints that emerged from the interplay of technological limits and control tensions over how data gets aggregated, shared, and used. Resolving one bottleneck didn't end the problem. It reconfigured the constraints and produced the next one. The paper's deeper move is theoretical: it extends resource dependency theory by showing how the peripheral team running the platform clawed back power not through authority, which it didn't have, but by demonstrating value creation with whatever thin slice of data it could get its hands on.

The Threat You Prepare For Is Not the Threat You Get

Every incumbent that announces a platform pivot braces for the same fight. Analysts ask whether it will cannibalize the core business. Boards ask whether a nimbler startup will out-execute them. Strategy decks model competitive response. None of that framing prepared the firm in this study for what actually happened, which was an internal standoff. The business units that originated the SME and lender relationships controlled the data the platform needed to function, and those units had every incentive to keep it that way. Sharing data with a new internal platform meant losing the leverage that came from being the only place a colleague could go to get it. The platform team was not fighting a competitor. It was fighting its own organization's accounting of who owned what.

This is not a quirk of one case. It is close to the default condition of open banking, the regulatory regime built specifically to force this kind of data to move. The UK's Commercial Credit Data Sharing scheme, legislated under the Small Business, Enterprise and Employment Act 2015 and brought into force the following year, compels nine designated banks to share SME credit and current-account data with credit reference agencies on a customer's request, precisely so smaller lenders can compete for that SME's business. A companion rule, the Bank Referral Scheme, goes further: it requires banks that reject an SME for a loan to refer that business, with consent, to a finance platform that can match it with an alternative lender. That is close to a government mandating the exact platform the anchor paper studies, and still the data did not move on schedule.

Open Banking Is the Field Experiment the Paper Predicts

When the UK's broader Open Banking rules took effect, requiring the nine biggest banks to open customer data to outside apps by January 13, 2018, six of the nine asked the regulator for more time. Reuters reported that only Lloyds, Danske Bank UK, and Allied Irish Banks had the technology ready; Royal Bank of Scotland, HSBC, Barclays, Santander UK, Bank of Ireland, and Nationwide Building Society all sought extensions ranging from a few weeks to a year. A fintech founder told Reuters plainly why: "There's very little incentive for (banks) to play ball with this, because it's just a pure downside scenario." A senior bank executive separately estimated the rules could cost the industry 6 to 10 percent of revenue, and the Bank of England projected aggregate profit losses above £1 billion by 2023. The firms counting on the data felt the stall directly. Funding Circle, a peer-to-peer SME lender, had been preparing to hire more than 80 developers to build on open banking; by the end of that year it redirected the investment elsewhere rather than build against data it could not reliably get. MarketInvoice had its integration built and sitting idle, waiting on banks that weren't ready.

That was 2018. Five years later, a 2022/23 Omdia IT Enterprise Insights survey of the industry found the obstacle had not moved outward to competitors; it had stayed exactly where it started. Sixty-five percent of respondents named extracting data across their own existing data silos as a top open-banking challenge, and 60 percent named integrating APIs with legacy and core systems, both ranked above external competitive pressure. Omdia's analysis added that most UK banks were still prioritizing faster product launches on their core systems over investment in their open-banking strategy. A full decade after the legislation, Britain's own Treasury was still trying to fix the plumbing: a September 2025 consultation on enhancing the CCDS scheme found multiple incompatible data-format specifications still in circulation among lenders and credit reference agencies, forcing firms to re-enter the same SME data in different formats and degrading its quality in the process. A regulator with the power to compel data sharing by law has spent roughly ten years and still hasn't gotten the internal plumbing to agree with itself.

The platform team was not fighting a competitor. It was fighting its own organization's accounting of who owned what.

Why the Smallest Business Pays for the Standoff

The costs of this standoff don't stay inside the incumbent's org chart. They land on the businesses the platform was supposed to serve. A Federation of Small Businesses survey found that only 15 percent of small UK firms were sharing their business bank account data electronically with any third party, most of them only to feed accounting software, not to shop for better financing. Two-thirds said they would not consider it at all, with roughly 40 percent calling it unsafe and a similar share unsure what benefit it would bring. FSB chairman Mike Cherry pointed to the reason: "The financial crash casts a long shadow. A lot of small business owners still don't trust lenders to do the right thing." Small firms were being asked to trust a data-sharing architecture that the banks supplying the data hadn't finished building for themselves.

Where the sharing did happen, the research on the UK's mandated scheme shows it worked exactly as the theory predicts. An Edinburgh Business School analysis of the CCDS mandate found that SMEs just below the scheme's £25 million turnover threshold were substantially more likely to open a new lending relationship with a non-bank lender than firms just above it, and the effect was about 50 percent larger for non-bank lenders than for banks, evidence that the previously withheld cash-flow data, not just credit history that was already public, was what had been protecting the incumbents' grip on their customers. Mandate the sharing, and smaller lenders can compete; leave it voluntary, and the data stays exactly where the unit holding it benefits most from keeping it.

The Rule for Any Incumbent Platform Launch

The anchor paper's contribution is to show that this is not a one-time integration problem you solve and move past. It is a sequence: five bottlenecks, each one's resolution reshaping the constraints that produce the next. An incumbent that budgets for "the data migration" as a single line item in a project plan has misread the shape of the fight entirely. The real project is renegotiating, repeatedly, who inside the firm gets to decide what data means, who it belongs to, and what it is worth, while competitors who were never the actual threat get three paragraphs in the board deck they did not earn.

Before your firm's next platform pivot, ask less about who outside the building might copy you and more about who inside it is quietly deciding not to help. The paper's five bottlenecks and a decade of UK open-banking data agree: that is where the launch actually gets decided.