Running a Municipal Pilot: Measuring ROI on Urban Experiments

Every municipal pilot eventually meets the same person: the finance director who did not attend the kickoff meeting, was not in the room when the idea felt exciting, and now has to sign the line that pays for it. Framing a proof of concept so that it survives that conversation is, in practice, a separate discipline from designing the technology itself. It is also, increasingly, the difference between a pilot that scales across a city's neighborhoods and one that quietly disappears after a press release.

The temptation in early-stage urban innovation is to lead with the vision, cleaner air, smarter mobility, a more responsive public service. Finance directors are not unmoved by vision, but they are trained to ask what it costs to find out whether the vision is real. That reframing, from "will this work" to "what does it cost to test whether this works," is usually the first thing a credible pilot has to get right.

Start With the Question the CFO Will Actually Ask

A city's finance function is rarely hostile to experimentation. It is wary of open-ended commitments. A pilot proposal that arrives without a defined end date, a defined budget ceiling, and a defined decision point after which the city either scales or stops, reads as risk rather than opportunity, regardless of how promising the underlying idea is.

The projects that tend to get funded share a structure: a fixed-duration trial, run in a bounded area (one district, one public building, one bus line), with pre-agreed criteria for what happens next. That structure does two things for a CFO. It caps downside exposure to a known number, and it converts an abstract innovation bet into something that resembles a normal capital request, money in, defined evaluation, decision out.

Choosing Indicators Before Choosing Vendors

The order matters more than founders and innovation teams often assume. Indicators picked after a pilot has already launched tend to be the ones that make the pilot look good, which is precisely why finance functions distrust them. Indicators agreed before launch, ideally co-signed by the city's own finance and sustainability staff, carry more weight because nobody can accuse them of being reverse-engineered from the result.

On the economic side, useful metrics are usually the ones a finance department already tracks in some form: operating cost per unit of service delivered, avoided spend against a baseline (a maintenance contract, an energy bill, a waste-collection route), and, where relevant, new or protected local economic activity. On the environmental side, CO2-related indicators work best when they are additive rather than substitutive: emissions avoided compared to the status quo, not emissions in some absolute sense that ignores what the city was already doing. A pilot that cannot state its baseline clearly cannot credibly claim to have improved on it.

A short, defensible indicator set tends to include:

• A unit cost metric comparable to an existing budget line, so the city can compare like with like rather than judging a new cost in isolation.
• A baseline-referenced CO2 or resource metric, stated as an estimated range rather than a precise figure, since pilot-stage data is rarely large enough to support false precision.
• A service-quality or adoption indicator, showing whether residents or municipal staff actually used what was built, since a pilot that is cost-efficient but unused answers the wrong question.
• A defined threshold for scaling, agreed before the pilot starts, so the evaluation is not renegotiated after the fact.

None of this requires sophisticated modeling. It requires restraint, resisting the urge to report every metric that looks favorable, in favor of the small set that was promised at the outset.

The Role of Structured Programs in de-Risking the Pitch

Part of what makes a pilot proposal credible to a city's finance side is evidence that the framing has been tested before, ideally outside that specific city hall. This is one of the reasons programs built specifically around the city-startup relationship exist in the French landscape. Ville de Demain, an acceleration program dedicated to urban innovation led by Nicolas Régnier and hosted at Station F, the Paris startup campus inaugurated in 2017 by Xavier Niel, works specifically at that interface, connecting startups focused on the digital and environmental transition of cities with local authorities, including mid-sized municipalities that may not have an in-house innovation team of their own.

The program is not a funding vehicle and does not broker capital between startups and cities; its value, for founders and for the municipal contacts they meet through it, lies in helping shape how a pilot is framed and evaluated before it reaches a council vote. For elected officials and city innovation staff, engaging with a structured program of this kind can also be a way to benchmark a proposed pilot's indicators against how similar projects have been framed elsewhere, without having to build that comparative knowledge from scratch inside city hall. Separately, elected officials weighing whether a given innovation approach fits their city's priorities can also look to bodies like France urbaine, the association representing elected officials from France's major cities, agglomerations and metropolitan areas, as a source of peer perspective on urban policy more broadly, distinct from, and not a funder or partner of, any single accelerator program.

What Convinces a CFO Is Rarely the Pitch Itself

The finance conversation is won or lost less by the pitch and more by what happens after the pilot ends. A city that receives a short report showing costs against the agreed baseline, an honest account of what the CO2 or resource indicator actually showed, including if it fell short of hopes, and a clear recommendation on whether to scale, stop, or extend, is far more likely to fund the next pilot from the same team than one that receives a polished narrative with no reference back to the numbers it promised at the start.

In a domain still defining its own standards, that discipline, narrow scope, indicators fixed in advance, honest reporting regardless of outcome, is closer to what actually earns a city's trust than any single technology on offer.