Tariff design is where charging networks quietly succeed or fail commercially. The same site, the same hardware, the same drivers — and pricing decisions can swing it from a loss-maker to a compounding asset, or from a community amenity to a local resentment.
Yet most operators inherit their tariff structure rather than design it. Here's a framework for doing it deliberately.
Start from the three constraints
Every public tariff lives inside a triangle:
- Cost recovery. Energy (increasingly volatile), grid connection charges, hardware amortisation, maintenance, platform and payment costs. Know your fully loaded cost per kWh per site — sites differ more than operators expect, and a blended average hides which locations subsidise which.
- Local alternatives. Drivers benchmark against the chargers ten minutes away and — for those who have one — their home rate. Public charging carries a justified premium; the question is how far it stretches before behaviour changes.
- Utilisation response. Price is a demand lever. A rapid hub at 15% utilisation with premium pricing might earn more at a lower rate and 30% — or might not, if demand is inelastic commuter need. You only learn your site's elasticity by changing prices measurably — which requires per-site utilisation and revenue data good enough to read the response.
Choose structures drivers can predict
The menu of components — per-kWh rates, session fees, time-of-day bands, idle fees, subscriptions — rewards restraint. Two principles:
Per-kWh should carry the weight. UK regulation requires pricing displayed in pence per kWh, and it matches how drivers think about "filling up". Session fees and time-based components have legitimate uses (covering fixed payment costs; discouraging bay-hogging on AC), but a tariff dominated by them reads as obfuscation.
Complexity must pay for itself in changed behaviour. Off-peak windows that shift depot-adjacent demand into the night: worthwhile. Five overlapping components a driver can't mentally total before plugging in: a trust tax. The test for every added element — will a driver change what they do because of this, and can they understand it at the charger?
Idle fees: the sharpest double-edged tool
Idle fees solve a real problem — a full battery blocking a rapid bay is lost revenue and a queue of frustrated drivers — and generate the industry's worst headlines when done badly.
Done well: a grace period after charging completes, clear signage and app notification, capped exposure, and automatic waiver when the data is ambiguous. That last one is an engineering point as much as a policy one — an idle fee computed from a session record with a missing stop event is how a £400 invoice ends up in the local paper. Idle fees should only ever be charged from sessions that pass integrity checks, and the tariff engine should refuse otherwise.
Done badly — no grace, no notification, applied overnight at a residential on-street charger where moving the car at 3am is unreasonable — they teach drivers to avoid your network. Context matters: idle fees belong at rapid hubs, rarely on overnight residential schemes.
Differentiate by audience, not by trickery
The defensible version of price differentiation is transparent segmentation: a fleet rate for contracted volume, a resident's rate on council schemes, a tenant rate at workplace sites, roaming visitors at the public rate. This is where multi-tenant tariff machinery earns its keep — each audience sees its own clear price, and the segmentation is a published feature, not a discovered trick.
What ages badly is asymmetry drivers discover: app price quietly lower than contactless, roaming rates wildly above native ones. Sometimes there are real cost differences — payment processing on contactless is genuinely dearer — but the gap needs to be explicable in one sentence at the charger, or it reads as a penalty for not being a member.
Operational hygiene: one tariff source, full history
Two platform requirements that sound dull until their absence bites:
- Single source of truth. The price on the charger display, in the app, in the open-data feed and on the invoice must come from one tariff record. Maintained separately, they will diverge — and every divergence is a refund plus a compliance question.
- Versioned tariffs with effective dates. Prices change; sessions straddle the change; disputes arrive months later. "What was the tariff at this site on 14 March at 23:40?" must be a lookup, not an archaeology project — and scheduled future changes ("new rates from 1 April") should be staged in advance with an audit trail of who approved them.
Treat pricing as an experiment you're running
The operators who win commercially review tariffs like a product: quarterly, per site, against utilisation and margin data, with small reversible changes rather than annual shocks. The prerequisite is the data loop — session, revenue and utilisation figures trustworthy enough to attribute a demand change to a price change.
Price with your costs known, structures a driver can total in their head, sharp tools handled carefully, segments in the open, and one versioned source of truth underneath. Everything else is iteration.
AmpNexus manages versioned, per-site, per-audience tariffs with full audit history from one source of truth — see the Portal or book a demo.