The brand is a mid-market UK retailer selling into a specific consumer category — I have agreed not to name it — with roughly £80m of annual direct-to-consumer revenue. The marketing director — call her Silke — has been at the brand for eight years and made the specific decision this piece describes in early 2023, three and a half years ago. The decision was, at the time, controversial within the business. Three years on, the same decision is being described inside the business as one of the most consequential strategic moves the marketing function has ever made.
The customer
The customer segment in question represented, at the time of the decision, approximately 12% of the brand's revenue and approximately 18% of its contribution margin. Per-customer, the segment was the most valuable segment on the brand's books — high average order value, high frequency, favourable margin profile, low return rate. On the standard commercial metrics, the segment was the sort of customer any brand's finance function would want more of.
The problem was that the segment was, on every qualitative measure the marketing function tracked, the wrong customer for the brand.
The specific qualitative problem was this. The segment engaged with the brand primarily as a value proposition — they bought at scale during promotions, engaged actively with discount messaging, responded strongly to price-driven communications, and were, in aggregate, unresponsive to the brand's editorial and lifestyle content. The brand's positioning was, and remains, specifically not a value proposition — the brand competes on quality, editorial voice, and a specific lifestyle claim that is distinctively not price-driven.
The commercial consequence of serving the segment, on Silke's analysis at the time, was that the brand's marketing function was, in effect, running two parallel programmes with divergent creative directions: the promotional programme that served this segment, and the brand-editorial programme that served the rest of the customer base. The promotional programme was, gradually, distorting the brand's public perception in the market — the promotional content was more frequently seen, more heavily distributed, and more directly measurable than the editorial content, and the brand's perceived positioning had, over the previous three years, drifted in a value-oriented direction that was measurably damaging in the brand-tracking work.
The decision
Silke's proposal, brought to the CEO in Q1 of 2023, was to systematically de-emphasise the promotional programme that served the segment — to reduce promotional cadence, tighten promotional depth, and, critically, to accept the near-term revenue decline that would follow.
The projected revenue impact, on the finance team's model, was approximately £6-9m over the following twelve months, primarily from reduced volume in the promotional segment. On any conventional commercial analysis, the decision was indefensible: the brand was walking away from a demonstrable, high-margin, high-frequency customer base for reasons that were qualitative rather than quantitative.
The CEO's response, in Silke's account, was initially sceptical and eventually decisive. The scepticism came from the specific projected revenue impact and the specific difficulty of defending it to the board. The decisiveness came from a longer conversation about the brand's medium-term trajectory. The value drift, as the CEO came to understand it, was going to cost the brand its ability to defend premium pricing across the customer base within three to five years. The £6-9m near-term revenue loss was, in this framing, the cost of preserving a much larger longer-term price defence.
The decision was approved. The promotional cadence was systematically reduced starting in Q2 of 2023. The specific segment gradually disengaged from the brand — some customers reduced purchase frequency, some stopped purchasing entirely. Twelve months into the implementation, the brand's revenue was down approximately £7.4m against the pre-decision baseline, roughly in line with the finance team's projection.
Every brand of a certain age has a customer segment that is, on the finance model, profitable and, on the brand model, corrosive. Serving both is not possible over the long run. The specific act of choosing between them is the strategic work most brands avoid, because avoiding it is possible for years before the cost becomes visible.
Three years on
The compensating effect, which took approximately eighteen months to become fully visible, was substantial.
The brand's core customer base — the customers who engaged with the editorial positioning rather than with the promotional messaging — showed measurable improvements across the brand-tracking metrics. Brand recognition rose. Unaided recall rose. Price sensitivity in the brand-tracking work declined. Willingness-to-pay indicators improved. All of these are relatively slow-moving metrics that respond over eighteen-to-thirty-six-month windows to changes in the marketing function's actual output. On all of them, the brand's position by the middle of 2024 was materially stronger than it had been in early 2023.
The commercial consequences of the improved position materialised through 2024 and 2025. Average order value in the core customer base rose approximately 14% over the two-year window, driven primarily by increased willingness to purchase premium product tiers that the brand had, in the value-drift period, been struggling to sell at list price. Retention in the core customer base rose approximately 8 percentage points, driven by the increased alignment between the brand's marketing output and what its actual customers valued. Wholesale demand, from the boutique retailers the brand also supplied, rose measurably as the brand's public positioning became clearer and more defensibly premium.
By the end of 2025, the brand's total revenue had returned to its pre-decision baseline and was continuing to grow. Contribution margin per pound of revenue had risen approximately 4 percentage points, driven by the improved product mix. The specific segment that had been walked away from was now approximately 4% of revenue — a substantial reduction from the previous 12%, and one that, on Silke's ongoing tracking, is broadly stable. The brand has not attempted to win the segment back.
What the CFO thinks now
The CFO — who, in Silke's account, was the most reluctant participant in the original decision — has become, over the three years since, one of the strategy's most active advocates. Her published view, in an internal memo Silke shared with me (with permission), is that the decision was "the single highest-leverage marketing decision this business has made in ten years, and one that would have been impossible to justify against the finance model that measured only the near-term revenue impact".
The CFO's current position, on which she is not shy in internal conversations, is that most brands operating at the price tier and stage of maturity of this business have a similar customer segment that is quietly damaging their medium-term positioning, and that most CFOs at similar businesses would resist the decision to walk away from that segment for the same reasons her 2023 self had resisted. She now spends a meaningful share of her external speaking engagements arguing that finance functions need to develop the analytical vocabulary to properly cost the long-run impact of brand-drift, so that decisions like this one become more defensible in prospect rather than only in retrospect.
The specific vocabulary the CFO now uses, which Silke described in some detail, involves distinguishing between customer-level profitability (which is directly measurable) and brand-level profitability (which requires the finance function to place a specific value on brand-tracking metrics). Their internal model now includes a "brand-drift cost" line that translates specific declines in brand-tracking indicators into projected pricing-power impact over three-to-five year windows. The model is imperfect. The specific act of having it, on the CFO's own view, has changed the internal conversations they now have about strategic marketing decisions in ways that produce, on aggregate, better decisions than the previous framework did.
The decision to walk away from the segment, in this larger framing, was not primarily a marketing decision. It was a strategic decision about what kind of brand the business wanted to be over a ten-year window, and a corresponding acceptance that certain kinds of near-term commercial opportunity were incompatible with that longer position. Most businesses, in most sectors, will face similar decisions if they are willing to look at them honestly. Most will not, because the framework for looking at them honestly does not, in most businesses, currently exist. Building that framework, in Silke's view, is the most consequential piece of strategic marketing work available in most organisations right now.