PAPERZZI

Ten papers across the collection point to the same problem: the tools banks and governments use to judge small businesses often miss what matters. In one study, financial ratios misclassified about half of struggling firms at exactly the moment the judgment mattered most. Two Dutch research teams independently arrived at a similar answer: judge a business as a story over time, not a snapshot, and put the founder’s resilience at the center.

Eight studies, from several continents, describe a strikingly similar pattern: buying AI produces signals, not results. Value tends to appear when people and knowledge structures turn AI into new ways of working, employees feel ownership of the change, and humans keep the judgment to question what AI produces. Meanwhile, the readiness tools used to pick which firms deserve support appear themselves immature.

Egypt’s digital payment system appears to have transformed because entrepreneurs, banks and the central bank built trust and rules together over nine years, not because of any single invention. Other papers in the collection suggest AI credit scoring stalls at micro scale because decision rights sit in four disconnected institutions, and that work on finance, reporting and statistics keeps landing on the same design principle: let small firms share their own data, with consent, across banks, insurers and governments.

The nine papers in this track point to one shared diagnosis: entry programs and microloans exist, but growth capital and scale support do not. Women become investment-ready and still meet unsuitable products and risk-averse funders. And the field may often measure the wrong thing: firm-size expansion is not the only ambition worth counting.

Six papers in unrelated fields describe the same gap: support exists for micro-firms and for large corporates, and nothing holds the middle. The proposed remedies look similar everywhere: bridge institutions that hold the middle, from venture builders to pilot-to-procurement programs to thresholds that would finally make mid-sized firms visible.

In one comparison, two territories with nearly identical resources delivered very different results, suggesting composite scores cannot tell a high-input region that wastes its resources from a modest one that converts them fully. The papers propose a shift: measure conversion efficiency and capability position, and protect the mid-sized manufacturers whose skills are hardest to rebuild.

In one study, Turkish SMEs went green and digital when they entered new export markets, not because of grants. Other papers point to public procurement that buys solutions that do not yet exist, giving small firms a pilot client instead of a subsidy. Together they suggest buyer pressure does what green and digital subsidies assume but often fail to deliver.