Evidence
MOSAIC has been run on twenty five programmes across the public, private and nonprofit sectors, using only evidence available at the decision point, with every figure sourced to the public record and every verdict written before the outcome was known, covering healthcare, financial services, infrastructure, technology, education and public administration across the United Kingdom, the United States, Europe, Asia and the Middle East. Seven of the twenty five carry a full diagnostic run against every available tool. The remainder are coded to the same schema from the published record.
Each read produces one verdict on pace. Full pace means the evidence supported committing. Phased means commit to one population first, against a named condition. Hold means do not commit until a named condition closes. The map below groups every case by who owned the programme and what the verdict was. Select any case to read it in full.
The case map
Five of these twenty five programmes were told to move at full pace and were right to. The verdict is not a preference or a temperament. It is what the evidence available at the decision point actually supported, and it is written down before the outcome is known so it can be checked afterwards.
A note on the shape of this. The case base leans towards hold because public evidence leans towards failure. Programmes that go well are rarely audited in detail, which makes them considerably harder to source. We would rather name that bias openly than quietly balance the grid.
The full case library
Each case sets out what the organisation was trying to do, what the delivery problem actually was as distinct from the stated one, what they chose to do, and what MOSAIC would have said had it been in the room. Figures describe the size of the decision at the point it was taken. They are not savings Sense Shift claims to have delivered.
The evidence base
Employees willing to change work behaviours to support organisational change fell from 74 percent in 2016 to 43 percent in 2022. In April 2025, 79 percent reported low trust in change.
Just 32 percent of business leaders report achieving healthy change adoption by employees.
Abandonment of AI initiatives rose 147 percent between 2024 and 2025.
More than 80 percent of AI projects fail, roughly twice the failure rate of conventional IT projects.
One in six IT projects overruns cost by 200 percent and schedule by 70 percent.
Large IT projects run 45 percent over budget on average and deliver 56 percent less value than predicted.
Federal IT management has sat on the GAO high risk list continuously since 2015. All twenty five MOSAIC cases are sourced from published audit, regulatory, court and company filings.