This is the first installment in a six-part series on designing and building profitable EV charging stations.

Site selection is the single biggest determinant of charging station profitability. Get it right, and you build a foundation for strong utilization and returns. Get it wrong, and no amount of operational optimization will fix it.

This is the first in a six-part series on designing and building profitable EV charging stations. We're starting with demand because everything else builds on it. Understanding who will use your chargers, how often, and why sounds straightforward, but in practice, it rarely is.

The factors that drive utilization interact in complicated ways, and the relationships that hold in one context may not apply in another.

The five core demand drivers

Most site selection frameworks start with the same fundamentals. These matter, but on their own, they don’t tell the full story.

1. Population density

More people generally means more potential drivers. Dense urban areas tend to see higher station utilization than rural locations. But density alone doesn't guarantee demand. A dense neighborhood with low EV adoption or widespread home charging may underperform a less dense area with the right mix of other factors.

2. EV adoption rates

The number of EVs in an area sets the upper bound on potential demand. Markets like California, where EV penetration exceeds 25% of new car sales, behave very differently from markets where adoption is still in single digits. Both current penetration and growth trajectory matter.

3. Demographics

Income levels, housing types, and commute patterns all influence charging behavior. Higher-income areas tend to have higher EV ownership, but they also have higher rates of home charging, which can reduce public charging demand. In many cases, housing type is more predictive than income alone.

4. Roadway and traffic patterns

Traffic determines visibility, accessibility, and the types of trips your station will serve. Highway-adjacent sites cater to long-distance travelers who need fast charging to continue their journey. Urban sites serve daily commuters, local errands, and commercial drivers. The same hardware can see wildly different utilization patterns depending on traffic context.

5. Competition

The presence and quality of nearby charging infrastructure shapes your potential market share. But competition cuts both ways: in emerging markets, being first can mean educating the market and building loyalty. In mature markets, differentiation on speed, reliability, or amenities becomes essential. Understanding not just where competitors are, but their utilization rates and pricing, provides crucial context.

Why these relationships aren't always predictable

Here's where site selection gets complicated. These five factors don't operate independently. They interact in ways that can amplify or cancel each other out, depending on the specific context of your site.

Traffic volume may dominate performance for a highway corridor site, where drivers are passing through with limited options. Local EV adoption may matter less in that context.

In contrast, for an urban retail location, local EV ownership can drive utilization. A site with moderate traffic but strong local adoption may outperform a high-traffic site in an area where few residents drive electric.

Competition creates similar contextual effects. In some markets, limited infrastructure creates opportunity. In others, oversupply suppresses utilization. The relationship is not linear. It depends on market maturity, driver behavior, and the specific value proposition you're offering.

This is why simple scoring models often mislead. Weighting each factor independently ignores how demand behaves in the real world.

The hidden demand segments

When most people think about EV charging demand, they picture the typical EV owner: someone with a single-family home, a garage, and a Level 2 charger they plug in every night. This profile represents about 80% of EV charging by volume, and it happens almost entirely at home.

But public charging stations don't serve home charging. They serve the gaps: drivers who can't charge at home, drivers who need a fast top-up mid-journey, and drivers whose usage patterns demand more than overnight charging can provide.

Two segments matter disproportionately for fast-charging utilization:

Renters and apartment dwellers
Nearly 35% of U.S. households rent, yet only about 5% of rental properties offer EV charging access. This creates a massive gap between EV ownership potential and charging availability. Research shows that disadvantaged communities, which disproportionately include renters, have access to 73% fewer public chargers per capita than more affluent areas. For this population, public fast charging isn't a convenience; it's a necessity.

As multi-unit dwelling charging grows (projected to reach 17% of the market by 2030), understanding where renters live and where infrastructure gaps exist becomes essential for site selection.

Rideshare drivers
This segment punches far above its weight. Rideshare drivers are adopting EVs five times faster than the general public, driven by fuel savings of roughly $0.10 per mile and regulatory mandates in markets like California and New York City. The average rideshare driver charges approximately five times more frequently than a typical retail customer.

The numbers are striking: one charging network saw rideshare throughput grow from 11% of total volume in 2021 to 24% by 2024. Another saw daily sessions at a single location jump from 50 to 500, most of them rideshare vehicles. Sites near airports, downtown cores, and heavy pickup/dropoff areas are seeing this effect most acutely.

For site selection, this means looking beyond residential EV registration data. A neighborhood with moderate EV ownership but high rideshare activity, or a large renter population with limited home charging access, may generate more public charging demand than a wealthier area where most owners charge in their garages.

Why predictive modeling matters

At this point, the complexity is clear. You're not evaluating five independent inputs. You're evaluating an interconnected system where the importance of each variable depends on the context of all the others.

This is why intuition and spreadsheet models fall short. Human judgment can't reliably weight dozens of interacting variables across hundreds or thousands of potential sites. And static scoring models that treat each factor independently miss the contextual relationships that actually drive utilization.

The only way to properly account for this complexity is a predictive model trained on actual charging behavior, one that has learned the relationships between demand drivers and utilization from tens of thousands of real stations and has been benchmarked against live data.

This is the approach we've built at Stable. Our demand estimation models are trained on utilization data from over 50,000 chargers and are continuously updated as the market evolves. We combine EV adoption trends, demographic data, traffic patterns, competition analysis, and energy costs to generate site-level forecasts. And critically, we benchmark these predictions against actual utilization to validate that our models work.

At the portfolio level, this approach delivers remarkable accuracy: error rates under 2% for portfolios of 25+ sites. Individual locations will vary, but across a portfolio, the signal becomes highly reliable.

What's next

Choosing the right areas is the first step. In Part 2 of this series, we’ll explore forecasting growth to understand not just where demand exists today, but where it’s heading over a five- to ten-year investment horizon.

In the coming weeks, we’ll also cover:

  • Estimating energy costs
  • Optimizing incentives
  • Station sizing
  • Amenities and site-level differentiation

If you’re evaluating new markets or portfolio expansion, you can explore Stable Evaluate for free to see projected utilization across your target areas.

Get started with Stable Evaluate for free