Ground risk: what to ask before you commit to a site

Developers price planning risk, market risk and financing risk with sophistication. Ground risk is often priced with a contingency percentage and a hope. Yet the ground is the one input you cannot change: you can redesign the building, but not what it stands on. The projects that go badly wrong underground usually asked the right questions too late.

Questions that cost little before purchase

• What does the existing investigation data actually show — and does anyone hold it in a form that can be interrogated, or only as archived PDFs?
• Where are the soft zones, the buried channels, the fill? Screening-level settlement and bearing analysis on existing logs will not replace design, but it will redraw the risk map.
• What would the foundation strategy plausibly be, and what does that mean for cost per square metre?
• What did neighbouring investigations find? Ground does not respect title boundaries.

From data room to decision

The obstacle has never been the questions; it is the effort of answering them inside a transaction timeline. Historic logs arrive as scans in a data room, and building a view from them took weeks a deal does not have. This is exactly the work AI structuring compresses: parse the logs, build the model, run the screening analytics — in time to affect the price, the contingency, or the decision to walk away.

Our position at Greenshift is simple: ground truth should enter the deal, not just the design. The earlier the model exists, the more decisions it can inform — and the same model then serves design, construction and, eventually, the asset's operating life.