Three numbers from mid-October worth your time, and one thing I would do about them.
1. Sessions up 26%, utilization flat. That's a healthy market.
Paren's Q3 2026 report counted 50.4 million US fast-charging sessions last quarter, up 26% from 39.9 million a year earlier. National utilization barely moved: 15.9%, against 16.4% a year ago, and it has stayed between 15.7% and 16.5% for five quarters. The reason is supply. Operators added 4,106 ports in Q3, and the US now has 80,964. The top 30 metros run at 21.1%. Everywhere else runs at 11.0% (Electrek has the same numbers).
All of that happened while new EV sales fell about 39% from last year's tax-credit rush, according to Cox Automotive.
Why it matters: I read this as a good sign. Capacity is being built to meet demand, and demand is still growing at a healthy clip. Flat utilization mostly means the sessions are spread across more sites. A new station takes time to ramp, and a lot of these are new, so plenty of them are still in the teething period. This isn't a cliff and it isn't oversaturation. It's how this kind of asset works through its cycle.
If I owned a site pegged at 11%, I would wait. Hold, be patient, and watch how things grow. Look into what adding capacity would take, without spending money on it yet, and plan for the future. There's no reason to get worried or sell.
2. Tesla raised prices 15%. Everyone else held flat.
Tesla's average non-member price is now $0.569 per kWh, up 15% in a year. Members pay $0.396, up 9%. Non-Tesla networks average about $0.533 and have been flat for five straight quarters. Paren puts the national average at $0.540, up 2.7%, and credits the increase to Tesla. Gas, meanwhile, set a September record of $4.33 a gallon (AAA).
Why it matters: this is business at its finest. Tesla has a robust system, people want to use it, and of course it's going to charge for it. That's not the Apple model. It's a sustainable business model. Charging is inelastic to price, and Tesla is the proof. That doesn't mean everyone should raise prices. It means you should know what's actually driving your demand and price to it. I'll come back to this one in its own post.
3. Fastned borrows from its drivers at 6%
Dutch operator Fastned reported Q3 like-for-like charging revenue of €46.3 million, up 47%, and annualized revenue of €395,000 per station so far this year. It's guiding to an operating EBITDA margin of about 45% for 2026. Part of the build is paid for with five-year retail bonds sold to its own customers at 6%, about €337 million raised by June (electrive). It still posted a €13 million net loss in the first half.
Why it matters: paying 6% to build a revenue-producing asset is pretty good. Banks are bad at lending on public charging because nobody can guarantee the utilization, so to a lender the merchant risk looks infinite. When the driver is the lender, the lender is also the user, and that takes a lot of the risk out. Could it work in the US? Maybe. It's crowdfunding, and the compliance side needs working out. More on this one soon too.
What I'd do
If your site is sitting near 11%, don't sell it and don't panic. Hold, learn what's actually driving the sessions you get, price to that demand, and plan your next stalls on paper before you spend on them.
The utilization forecast takes your own numbers on the ramp and the benchmarks page shows what other sites look like. For pricing, start with the business model guide. For money, the financing guide and the SBA loan calculator cover the conventional routes.
