Marketing within fair-treatment and disclosure constraints

# Marketing within fair-treatment and disclosure constraints

In October 2024 the CFPB ordered Apple and Goldman Sachs to pay more than $89 million over the Apple Card: $19.8m in customer redress, a $45m penalty for Goldman, $25m for Apple. Part of the finding concerned customers who believed that buying an iPhone on the card would automatically be financed interest-free, and who were charged interest instead. No headline claimed otherwise in so many words. The harm sat in how the purchase flow was built and explained.

That is the constraint this lesson works with. It does not ask whether your copy is accurate. It asks whether the people you reached ended up better off, and whether you could have seen in advance that some of them would not.

The standard: good outcomes, not clean copy

The FCA's Consumer Duty, in force from 31 July 2023 for products still on sale and 31 July 2024 for closed books, added a new Principle to the Handbook: a firm must act to deliver good outcomes for retail customers. Three cross-cutting obligations sit under it: act in good faith, avoid causing foreseeable harm, and enable customers to pursue their financial objectives. Four outcomes carry the specifics, covering products and services, price and value, consumer understanding, and consumer support.

Its predecessor, Treating Customers Fairly (2006), was mostly a process test: had you followed and documented the right steps. The Duty is an outcomes test. You can run a faultless approval workflow, clear every disclosure requirement, and still fail because the cohort you acquired ended up in a worse position than before they bought.

For a marketer that reorders the work. The rulebook mechanics (what a promotion must carry, how a rate must be expressed) come later, and the advertising and disclosure lessons in this module own them. Before any of it, you answer:

• Who is this product for, and who is it explicitly not for?
• What harm is foreseeable for the people at the edges of that group, not the median buyer?
• Could someone who bought after seeing only this creative say what they hold and what it costs them?
• Is the price defensible against the benefit, in writing, before the campaign runs?

Target market is a marketing decision

Product governance requires a firm to identify the target market for a product by the needs, characteristics and objectives it is built for, and to pick a distribution strategy that reaches that market rather than a wider one. In most banks this document is drafted by product and filed. That is a mistake, because it silently sets the boundary of every audience you are allowed to buy.

The useful half is the negative target market: the people for whom the product is predictably wrong. A 0% balance-transfer card is wrong for someone with no existing balance and a habit of revolving. A high-interest instant-access saver is wrong as the answer to a customer holding a maturing five-year bond. Writing that list down converts a compliance artefact into a targeting brief, and it is the only defence when a growth target pushes the media plan outward.

Creative sits inside the same boundary. The ASA ruled in December 2020 that four Instagram posts Klarna had placed with influencers were irresponsible: they encouraged shopping on credit as a way to lift a low mood during lockdown. Klarna sells buy-now-pay-later credit, so the mood-and-spend framing was its natural creative territory, which is exactly why the ruling matters. The claims were not false. The pairing of an emotional trigger with a credit product aimed at a young audience was the problem.

Price and value can kill a campaign before a brief is written. In February 2024 the FCA reported that GAP insurance paid out only around 6% of premiums in claims and asked firms to pause sales. Most of the market did. No amount of creative rescues a product whose value case collapses under a single ratio.

Foreseeable harm is a design test

Foreseeable harm is not the same as intended harm or likely harm. It is the harm you could have anticipated for a person inside your stated target market who behaves in an ordinary way. Which means the test runs on your journey, not just your ad.

A harm register for a campaign reads something like this:

> Product: 24-month instalment plan on device purchases. Target market: existing cardholders buying hardware over £300 who intend to repay in fixed instalments. Foreseeable harms: (1) customer assumes instalments are automatically interest-free and does not select the plan at checkout; (2) customer selects the plan and later revolves a separate balance at the standard rate; (3) instalment ends but the direct debit continues. Mitigations, owner and monitoring metric for each.

The Apple Card interest-free confusion is harm (1) in real life. The fix was never a footnote in a banner. It was a checkout step that made the default explicit and a statement that showed which balance was accruing interest. Marketing owned the assumption customers arrived with; product owned the moment it was resolved. When those two teams never sit in the same room, the gap becomes redress.

The instructive part: a compliant promotion and a harmful outcome coexist comfortably. The CFPB's own record is the cheapest evidence of what regulators actually punish. Browse the CFPB enforcement actions database and note how often the failure is operational rather than editorial.

The channels you do not control

Your distribution strategy is part of the product's fairness case, which makes affiliates, comparison sites, price-comparison position and influencer partners your responsibility even when a third party writes the words.

The failure mode is incentive design. Pay an affiliate on approved accounts and it will optimise for whoever converts, which tends to be the segment closest to your decline boundary or furthest from the intended use case. The campaign hits CPACPACost Per Acquisition: the total cost to generate one customer or conversion, computed by dividing total spend by the number of acquisitions.View full definition →. The cohort shows early arrears at several times the book average four months later. Under an outcomes test, that is a distribution failure attributable to you, and the affiliate contract is the evidence trail.

Two practical constraints follow. Cap or claw back on cohort quality, not just on conversion. And read your partners' creative the way a regulator would, since a distributor's exaggeration becomes your net impressionimpressionThe total number of times an ad or piece of content is displayed, regardless of clicks. Each display counts as one impression, even to the same person.View full definition → problem.

Eligibility, and the offer nobody gets

Product eligibility is where fair treatment quietly reshapes the funnelfunnelThe customer journey from awareness to purchase, typically Awareness, Interest, Consideration, Decision, Action, with prospects narrowing at each stage.View full definition →. Advertise a headline rate that a small minority of applicants actually receive and you generate three costs: wasted media on people who will be declined, a hard credit footprint on their file, and a cohort of accepted customers priced far from what they expected. Soft-search eligibility checks before a full application exist to remove the second of those. Use them.

The harder version is explanation. New York's Department of Financial Services investigated the Apple Card after public claims of gender bias in credit limits and concluded in March 2021 that it had not found unlawful discrimination under fair lending law. It was still critical of the experience: customers could not get a usable account of why they had received the limit they did. Legally clean, poor outcome. The consumer understanding obligation does not stop at the sale, and a black-box decision the front line cannot explain fails it.

Where a campaign needs to exclude customers showing signs of difficulty, the flagging and suppression machinery belongs to the consumer-protection lesson in this module. What matters here is upstream: whether the product was ever meant for them.

A pre-brief that survives review

Run this before creative, not after.

1. Restate the target market in one paragraph, in your own words. If you cannot, product governance has given you a document nobody can act on. Send it back.

2. Write the negative target market as an exclusion list. These are audiences, lookalike seeds and keywords you will not buy. Get product and compliance to sign the list, because it will be challenged the first week volume dips.

3. List foreseeable harms with an owner each. Anything you cannot mitigate becomes either a product change or a reason not to run.

4. Check the value case survives one sentence. Price against benefit, stated plainly. If the sentence needs three qualifiers, the campaign will need thirty.

5. Define post-sale metrics before launch. Complaint rate, early arrears or dormancy in the acquired cohort, cancellation inside the 14-day withdrawal right, and share of customers who used the feature they were sold. Conversion alone cannot tell you whether the outcome was good.

What the constraint buys you

Trust is the asset the foundations lesson treats as the thing bank marketing manages, and this is the machinery that either builds or spends it. The point of the pre-brief is that it moves the argument to a week when changing the answer is cheap.

The alternative price is visible. Apple and Goldman paid over $89m, Goldman was barred from launching a new consumer credit card without a credible compliance plan, and a partnership marketed on simplicity spent years defending its clarity. Redress is the small number in that list.

Key Takeaways

• The test is outcomes, not accuracy: good outcomes, an identified target market, no foreseeable harm. A campaign can clear every disclosure rule and still fail it.
• The negative target market is the most useful line in the product document. Turn it into an exclusion list for audiences and keywords, and get it signed before media planning.
• Foreseeable harm lives in the journey. The Apple Card interest-free confusion was a checkout default, not a headline.
• You are accountable for how distributors sell. Affiliate incentives paid on volume will find the segment your target market excluded.
• Legally clean can still be unfair: New York's regulator found no unlawful discrimination in Apple Card limits and still faulted the firms for being unable to explain them.
• Define post-sale metrics (complaints, early arrears, dormancy, 14-day cancellations) at brief stage. Conversion cannot prove an outcome was good.