Dawn Intelligence called the rise while the data pointed down.
At the May 2025 cutoff, North Norfolk was running 14% below its level a year earlier. Every trend in the recent data pointed the wrong way.
Dawn Intelligence forecast a rise anyway, made before the month began and logged: 62 domestic abuse offences for June 2025. The police recorded 59.
Over the six months that followed, the area held 23% above the same months a year earlier. Not a one-month blip, and not a trend line continued.
Consider Lucy, somewhere in North Norfolk. The Dawn Intelligence platform has never met her, and holds nothing about her.
On the numbers, her district had been having a quiet year. Domestic abuse was running well below where it had been twelve months earlier, month after month. Every ordinary reading of that area said the same thing: this one is improving, look somewhere else. That is when a rota thins and a budget line moves. Nobody is wrong to move it. Nothing in the record says otherwise.
Dawn Intelligence said otherwise. A month before June began it forecast a rise, against a trend that had pointed down for half a year. The rise came, and it held for the six months that followed.
Dawn Intelligence was never reading Lucy. It was reading the area she lives in, and the area was turning. A call like that buys the month before. That is the only month in which anything can still be put in place.
What it changes. A call made against the recent trend is the one a commissioner cannot get from a trend line. Matching a rise already under way is useful. Calling one while the data still points down is forecasting.