Writing price and savings claims that survive scrutiny

"You will save money by switching to us." Said by sales agents who did not know what the customer was currently paying, that sentence cost SSE £10.5 million when Ofgem penalised the supplier in April 2013 over misleading sales and marketing statements. Nobody invented a number. The failure was that the savings claim went out without the evidence needed to hold it up. The distance between "not a lie" and "substantiated" is where most enforcement in this sector lives, and it is why "save up to" should be the last phrase you draft and the first one you attack.

What follows is the drafting method: how to choose a baseline, how to state consumption assumptions, what goes in the substantiation file, and which wordings have already collapsed under examination.

Why energy savings claims are structurally risky

Energy pricing has a feature most retail categories do not: a moving, regulated reference point. In Great Britain, Ofgem sets a periodic price cap limiting what suppliers can charge on default tariffs. It has been reset quarterly since October 2022 rather than every six months, and it is calculated separately for each of the 14 distribution regions. So a claim like "save £200 a year" carries two hidden variables at once: which quarter, and which region.

Four failure patterns account for most of the trouble:

• A stale baseline. A comparison built on last quarter's cap, or on a withdrawn "average bill" figure, overstates savings the moment prices move.
• "Up to" carried by a tiny minority. If only the lowest-usage 3% of households could ever reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition → the maximum, "save up to £300" misleads the person reading it.
• A single national number over a regional cap. A saving that is true in the East Midlands can be materially smaller in North Scotland or on a prepayment meter, and the ad rarely says which region the arithmetic came from.
• Conditional savings stated unconditionally. Discounts that depend on direct debit, paperless billing or paying on time are not available to a customer who misses one payment, and the customers most likely to miss one are the same group the fair treatment lesson covers.

None of this requires dishonesty. It requires weak substantiation, which is precisely what regulators test.

What the wording has to survive

Which body can fine you, which can only pull the ad, and how those remits overlap across the UK, Australia and the US is the mapmapUsing software to automate repetitive marketing tasks and campaigns, enabling personalisation at scale across channels like email, web, and social.View full definition → the regulators lesson draws. For drafting purposes, three obligations bite regardless of who is holding the pen.

Evidence must exist before publication, not after a complaint. The CAP Code requires marketers to hold documentary evidence to support objective claims prior to distributing the ad. A file assembled in week three of an ASA investigation is already a bad file.

Qualifying information must sit with the claim. The ASA treats the proximity and prominence of the comparison basis as central to whether an ad misleads: a headline saving on screen and its consumption assumption three scrolls down in grey 9pt is a familiar way to lose a ruling.

Baselines can be prescribed for you. Australia is the clearest case. After the ACCC's 2018 retail electricity pricing inquiry found retailers advertising large percentage discounts off wildly different base rates, so that a "30% off" offer could be dearer than a "10% off" one, retailers were required from 1 July 2019 to express offers against a single reference price (the Default Market Offer, with Victoria running its own equivalent) and to flag conditional discounts as conditional. The second-order effect is worth noting: once the denominator was standardised, headline discount percentages shrank dramatically without a single tariff changing price. If your competitive story depends on a self-chosen baseline, a regulator can remove that story overnight.

Real triggers for action

"Cheapest deal" claims. The ASA has ruled repeatedly against suppliers and comparison services claiming a tariff was cheapest without a current or representative dataset behind it. The ASA's published rulings database is searchable and free, and reading three rulings in your claim's family before you write is cheaper than being in the fourth.

Fixed versus cap arithmetic timed badly. Complaints in 2022 and 2023 clustered around fixed-tariff savings measured against a cap inflated by wholesale gas prices: dramatic on the day of booking, unrepresentative of the year the customer would actually live through.

Device and efficiency savings. "Cut your bill with a smart meter" attracts scrutiny because the meter itself saves nothing. The behaviour change it enables might, and evidencing behaviour change across your own customer base is far harder than quoting a device spec.

Sales scripts, not just ads. The SSE case turned on what agents said in telephone, in-store and doorstep conversations. If your substantiation file covers the TV ad but not the call script and the objection-handling crib sheet, the weakest wording in the campaign is the one nobody reviewed.

What substantiation actually looks like

Before any claim goes live, you need a file, not a hunch. For a savings claim it should contain:

1. The exact baseline, dated and named, and defensible as representative: Ofgem's published cap figure for a specific quarter and region, not a hand-picked high-price outlier.
2. Typical Domestic Consumption Values (TDCV), Ofgem's standard usage benchmarks in kWh (medium usage sits in the region of 2,700 kWh electricity and 11,500 kWh gas), so "average household" means the same thing as it does for your competitors rather than something you defined yourself.
3. The share of customers who can actually reach the top-line figure, if you use "up to". The ASA looks for a significant, non-trivial proportion, not a rounding error of very low users.
4. A recency test: is the comparator still live? Wholesale prices and tariff tables move within weeks.
5. The conditions attached, written out: direct debit only, dual fuel only, first 12 months only, exit fees payable if the customer leaves early.
6. Consumption-sensitivity disclosure where the saving swings sharply with usage, placed next to the claim.

A simple worked example

Say you want to claim "save up to £180 a year versus a standard variable tariff."

• Reference price: the Ofgem cap for a medium TDCV household in a stated region, as at the date you publish (check Ofgem's current price cap page rather than reusing last quarter's number).
• Your fixed tariff annual cost at the same TDCV, in the same region, on the same payment method.
• Difference equals your headline saving, but only if both sides are calculated at the *same* consumption level, standing charges included. Comparing unit rates alone and ignoring the standing charge is a common way to manufacture a saving that does not exist on a real bill.
• Then test spread: what share of your target base sits close enough to TDCV for the figure to be broadly representative? A workable house rule (not a legal threshold) is that if fewer than a third land within reasonable range, "up to" is doing too much work and you should publish a range or a lower, achievable number.

Keeping the claim alive after it launches

Most price claims die of old age rather than of complaint. Ofgem publishes the next cap level several weeks before it takes effect, which means you often know, on the day you book media, that your comparison will be wrong halfway through the flight. Two consequences for drafting:

• Give every claim an expiry date and an owner, tied to the next cap reset rather than to the campaign end date. Anything running across a reset needs either a pre-approved alternative version or a hard stop.
• Keep a version log. Regulators and the ASA ask what the ad said on a specific date; "we updated it at some point" is not an answer. Screenshots with dates, plus the baseline file used for each version, settle the question in minutes.

Who signs this off, in what order, and what happens to the launch date when sign-off fails is the operating process the checklist lesson builds. Your job at the drafting stage is narrower: hand that process a claim whose arithmetic, conditions and expiry are already documented in one retrievable place.

🎬 [VIDEO: "How the Energy Price Cap Works" - youtube.com/@Ofgem - Ofgem's own explainer on the price cap mechanism that underpins most UK energy savings comparisons]

Key Takeaways

• Enforcement targets weak substantiation rather than deliberate deception: stale baselines, cherry-picked comparators, unconditional wording on conditional discounts, and "up to" figures only a small minority could reach. SSE's £10.5 million Ofgem penalty in 2013 came from savings claims made without the evidence to hold them.
• Hold the documentary evidence before publication, and put the comparison basis next to the headline claim rather than in a footnote.
• Anchor to a current, named, public baseline: the Ofgem cap for a stated quarter and region, applied at TDCV, with standing charges included on both sides.
• Regulators can standardise your denominator. Australia's reference price obligation from July 2019 shrank headline discount percentages without any tariff changing price, so avoid building a competitive story on a baseline you chose yourself.
• Cover sales scripts and call handling in the same substantiation file as the ads, and give every claim an expiry date tied to the next price cap reset.