Targeted Support

Targeted Support, Untargeted Reality: What Healey's Energy Handout and the FCA Have in Common

Two stories landed in GB this month that, on the surface, have nothing to do with each other. One is about government energy support. The other is about financial services regulation. Read them together and you'll notice the same mistake sitting underneath both: confident assumptions standing in for evidence.

The handout that missed its target

John Healey's energy bills support was designed to help people who needed it. That was the stated intention, and it's a reasonable one. Energy costs have hurt low-income households hardest, and targeted support is, in principle, the right tool.

But a City AM analysis found something the policy wasn't supposed to allow: homes worth millions of pounds qualifying for the handout, while households in poorer areas get nothing.

This isn't a story about bad intentions. It's a story about bad targeting mechanics. Whatever criteria were used to decide who qualifies clearly didn't map cleanly onto who actually needs support. Wealth and postcode don't always correlate the way policy design assumes they will. When they don't, the money goes to the wrong doors.

The result is a scheme that looks generous on paper and arbitrary in practice.

The FCA's version of the same problem

Three years into the Consumer Duty, the FCA has published its findings on how firms are assessing "fair value" — whether the price a customer pays is reasonable relative to what they get.

The regulator's conclusion is blunt: some firms are doing this well. Many are not. And the gap between the two groups comes down to one recurring issue — firms relying on assumptions instead of evidence.

The FCA flagged three specific failure points:

Weak evidence. Some firms could describe their product in detail but couldn't demonstrate, with actual customer data, that the price matched the benefit. Description isn't evidence.
No segmentation. Firms treated their entire customer base as one homogeneous group. That makes it impossible to spot whether some customers — by risk profile, channel, or vulnerability — are quietly paying more or getting less than others.
Poor customer understanding. Even a fairly priced product fails the fair value test if customers don't understand what they've signed up to. The FCA specifically called out "pay as you save" energy equipment agreements that implied guaranteed savings over terms lasting up to 25 years, when outcomes actually depend on equipment performance, future energy prices and the customer's own circumstances.

Same root cause, different sector

Put the two stories side by side and the pattern is obvious.

Healey's scheme assumed that certain criteria would reliably identify people in need. They didn't — because the scheme never segmented its target population accurately enough to separate genuine need from superficial markers of it.

The firms the FCA criticised made the identical error. They assumed their target market behaved as a single group, so they tested fair value against an average customer who doesn't actually exist. Hidden inside that average were customers paying more, understanding less, or receiving worse outcomes than the headline figures suggested.

Neither failure was about bad faith. Both were about skipping the hard, unglamorous work of checking assumptions against real-world evidence before acting on them.

What good targeting actually requires

The FCA's findings point to a practical discipline that applies well beyond financial services. Firms — and policymakers — that get this right tend to do three things:

Segment before you generalise. Break the population down by the factors that actually matter — income, risk, usage, vulnerability — rather than relying on a single proxy that's easy to measure but weakly correlated with the outcome you care about.
Evidence the outcome, not the input. A product description or a policy rulebook tells you what was designed. It doesn't tell you what actually happened to the people subject to it. Only outcome data does that.
Check whether people understand what they're getting. A technically fair price or technically correct eligibility rule still fails if the people affected don't understand the terms or the scheme doesn't reach who it claims to.

The FCA is explicit that this doesn't require complex modelling. Smaller firms, it notes, can often draw on information they already hold — complaints, arrears, customer feedback, product usage — to build a credible evidence base. The same logic applies to government schemes: the data to target support accurately usually already exists somewhere. The failure is in not using it rigorously enough before launch.

Why this matters now

Both cases will cost someone. The energy handout costs public money and public trust — taxpayers funding support for households that don't need it, while households that do need it get nothing. The Consumer Duty failures cost firms regulatory scrutiny, remediation, and reputational damage, on top of the harm done to customers who were quietly paying more than they should.

In both cases, the fix was available before the failure happened. It just required someone to ask harder questions of the design, rather than trusting that good intentions would produce good outcomes.

Key takeaways

Targeted support fails when eligibility criteria don't correlate with actual need — as shown by luxury London properties qualifying for energy support meant for struggling households.
The FCA's Consumer Duty review found the same pattern in financial services: firms treating diverse customers as one group, missing unfair outcomes hidden inside the average.
Evidence beats assumption in both policy design and fair value assessment — description of a product or scheme is not proof it works as intended.
Segmentation is not optional. Whether distributing public funds or pricing financial products, failing to break down the target population by the factors that matter will let unfair outcomes slip through.
The fix doesn't require new technology — in most cases, the data needed to target accurately already exists. The discipline required is using it before launch, not after the damage is done.