Below the deductible.

Premium answers mostly to the capital markets. The deductible is the term most often reported responding to how a risk is managed. That is where the return on sharing what a site knows is likeliest to show up.

Fleet telematics is sold, and bought, as an operations tool: safer driving, maintenance, running costs. Insurance came afterwards. In a Teletrac Navman survey of 600 respondents, reported in December 2025, 84% of fleets gave driver exoneration as a key reason for deploying safety technology, and 65% reported premium decreases after implementing it. Of the 600, 74% now run both telematics and dashcams, and more than 75% say insurance management has become the top driver of adoption.

It is a vendor's survey and should be read as one. But the pattern is plain enough. What began as something the better operators did has become what operators do. It rewarded those who adopted it, and it showed everyone else how.

That is the precedent for construction: see the risk continuously, manage it better, and in fleet the insurance terms followed. The question for construction is which terms.

Premium answers mostly to capital

In fleet, the reported reward came through the premium, and a fleet policy renews every year. Construction has no renewal to do that through, and its premium answers mostly to something else.

A project policy is bound once, for the life of the build. Its premium is typically adjustable on final contract value, and an extension of period normally carries additional premium, but those are adjustments for size and time. There is no scheduled point during the build at which the way the site is run is re-rated.

And the level of premium is set largely elsewhere. Marsh's Global Insurance Market Index recorded an average fall of 6% in global commercial rates in the second quarter of 2026, the eighth consecutive quarter of decline, with UK rates down 8% and property down 12% globally. That release gives no construction figure, but it is safe to assume construction is in there, as it is often lumped into property (with a capital P, as we say). Marsh's explanation was about insurers, not insureds: "Strong insurer profitability, a surplus of capital, lower reinsurance costs, and higher investment returns are intensifying competition and contributing to lower rates."

The level of premium follows the supply of capital far more closely than it follows the conduct of any one insured.

The deductible answers more to the risk

The deductible is different. A rate is benchmarked against a market. A deductible is set peril by peril, against the project in front of the underwriter. It is also the term on which the reported movement is largest and most specific. Premium credits are reported too, but more cautiously.

ENR, reporting on the US market in June 2026, had Justin Levine, CEO of Shepherd, saying that projects using Brickeye's leak-detection and automated shut-off technology can see water-damage deductibles reduced by 50% or more (the two companies' own announcement the same day says up to 50%). The same report had Matt Wagner, head of construction property at Zurich North America, saying that up-front loss-prevention measures can be the difference between a $2 million and a $1 million water-damage deductible on a large project. It quoted Sedat Kunt, Marsh's national builders' risk practice leader, calling site monitoring "one of the most significant shifts in builders' risk underwriting", and it quoted Aon's Jason Behrer on carriers that remain cautious about significant pricing credits until there is more performance data: they "see the potential for it, but they're not yet there".

Shepherd, a US MGU writing on partner capacity, has a programme that includes this. Builders using Brickeye receive a tiered protection classification, based on the contracted scope of IoT services, which can earn premium credits or lower deductibles.

A deductible is the part of every loss the insured parties keep. Carlos Casal, executive vice president of construction at Mast Capital, a Miami developer, put the high end of it to ENR in January 2026: "At some point, the deductible becomes large enough that you're effectively self-insuring the first loss."

So a lower deductible is more than a discount. It is risk the insured parties no longer carry on their own balance sheets.

Seeing below the deductible

There is a phrase I have heard from underwriters when they explain what continuous site visibility is for. They call it seeing below the deductible.

It is worth being exact about who is doing the seeing. It is the risk function, not the people who price or settle, and what reaches the engineer is the condition of the works, not a ledger of anyone's retained losses. What reaches the underwriter is what has always reached them: the risk engineer's judgement, now better informed.

Between surveys, a project is largely visible to its insurers only when it claims. Whatever falls beneath the deductible is borne by one of the insured parties, and little of it is seen in detail outside the project. That is where the frequent losses sit. The severe ones reach the policy, and are too rare to count.

The conditions that produce those frequent losses are also what the risk function sees least of, and they are often the conditions behind large ones. A leak that costs forty thousand and a leak that costs four million can begin at the same unfinished joint.

How the return arrives

Risk engineering is already part of the placement. It buys a finite amount of an expert's attention, spread across the visits and reviews a project needs, and what reaches the engineer in between is second-hand and after the fact. Share what the site already produces, and that attention can go to the conditions that matter now.

The return then comes in three steps.

First, a better chance that losses below the deductible stay small or do not happen. Every pound of that belongs to the insured parties, during the build.

Second, a better chance that the large loss does not happen either, because it can start in the same place.

Third, the deductible itself. The reported US examples are narrower than the argument here: a deductible reduced for a specific device against a specific peril. My argument is that the same logic extends to a risk that can be shown, continuously, to be well managed. That extension is mine, not theirs. If it holds, this is the step that moves risk off the client's balance sheet.

On a project policy the deductible is agreed at binding, before that project has produced any data. So the third step can arrive in one of two ways: upfront, for committing to monitoring, which is how the Shepherd and Brickeye classification works; or on the next placement. A contractor does not place one project. One that can show how its last site was run has something to put forward with the next.

What this does not say

It does not say any of this is promised. That continuous visibility reduces losses is a proposition with a precedent in fleet, not a settled finding in construction, and the examples above are reported from the US market by the people involved. Deductibles also move with the cycle. Marsh's John Donnelly notes insurers in many markets currently competing on "lower deductibles" among other terms, so a reduction earned on the quality of a risk can be hard to tell apart from one the market was giving anyway.

What it does say is where to look. Premium is set largely by capital. The deductible is the term most often reported responding to how a risk is managed, and lowering it is real risk transfer.

Pikt connects to the technology already running on site and gives the risk function continuous visibility of the project. What a site shows is recorded, not assessed: what a source said, and when. What underwriters do with a better-managed risk is their business, and we give no advice on it.