dm and the Anti-Discount Model: Paying More to Sell for Less

The most counterintuitive retail model in Germany pays more per hour than its competitors and has lower labour cost per unit sold. dm demonstrates that the standard retail assumption, that wages are a cost to minimise, is an accounting truth and an operational error. This case study belongs to the retail pillar of the Germany Company Stories hub and contrasts directly with the hard discount model.

Why does staff turnover cost so much in retail?

Because the visible cost, recruitment and training, is the smaller part. The larger costs are hidden in operations: an inexperienced employee is slower, makes more ordering and shelving errors, handles customer questions worse and requires supervision from colleagues who are then less productive themselves.

In a store with high turnover, a substantial share of the workforce is permanently in the learning phase, which means the store never reaches the productivity that its staffing level should deliver. That gap does not appear in any line item.

Stock accuracy is a specific example. Errors in ordering, receiving and shelf replenishment produce both out-of-stocks, which lose sales invisibly, and overstocks, which produce markdowns. Experienced staff make far fewer of both.

The cumulative arithmetic frequently favours higher wages, and it is rarely calculated because the costs sit in different budgets. Wages are a controllable line that a manager is measured on; lost sales from out-of-stocks are not attributed to anyone.

What does store-level autonomy actually mean?

Devolving decisions that most retailers centralise: local assortment adjustments, ordering, scheduling, store layout within a framework, and hiring within the team. Head office sets the range and the prices; the store decides much of how it operates.

The operational argument is information. A store team knows its own customers, its local competition and its seasonal patterns better than a central planner with a national dataset, and decisions made locally are faster and better fitted.

The motivational argument is stronger. Autonomy is among the most reliable predictors of job satisfaction and retention, and a retail role with genuine decision authority is a qualitatively different job from one executing a planogram.

The cost is variance. Devolved decisions produce inconsistency between stores, which conflicts with the standardisation that retail efficiency normally depends on. The model works only where the variance is contained within a clear framework and where the categories reward local judgement.

Why avoid promotional pricing?

Because promotions are expensive to run and they teach customers to wait. A promotional cycle requires forecasting, additional inventory, display labour, price changes, signage and post-promotion markdowns, and it trains customers to buy only when items are discounted.

Everyday low pricing removes all of that operational cost and produces steadier demand, which improves forecasting, reduces safety stock and smooths labour requirements. Steady demand is worth a great deal in a business with perishable working capital.

The strategic effect is price trust. A customer who believes prices are consistently fair does not comparison shop item by item, which is the most valuable position a retailer can occupy and the hardest to establish.

The difficulty is that the transition is painful. Moving from promotional to everyday pricing reduces sales during the adjustment period while customers recalibrate, and most retailers abandon the change before it works.

Is the model transferable to other retailers?

Partially, and the conditions matter. It works best in categories where product knowledge affects the sale, where assortment is broad enough that local judgement adds value, and where the customer relationship is repeated frequently.

It transfers poorly to formats where the operating model is deliberately deskilled. A hard discounter does not need experienced staff because the assortment is narrow, the layout is fixed and the customer requires no advice, which is the whole point of that model.

The general principle does transfer: identify where employee capability actually affects the customer outcome, invest there, and deskill elsewhere. Most retailers apply a uniform staffing philosophy across formats where the answer should differ.

The second transferable element is measurement. Companies that measure only controllable cost lines will always cut labour, because the offsetting benefits appear in metrics nobody owns. Changing the measurement changes the decision more reliably than changing the philosophy.

How does it compete against discounters on price?

By competing in categories where discounters are structurally weaker. Health and beauty products involve brand preference, advice, regulatory constraints and a long tail of specialised items that a narrow-assortment format cannot serve.

The assortment is therefore the defence. A customer buying a specific dermatological product, a particular supplement or a niche baby item cannot find it at a discounter, and once in the store they buy the routine items too.

Private label plays a role here as well, providing a value tier that competes on price where products are commoditised, while the branded assortment provides the reason to visit.

The result is a format that sits between the discounter and the pharmacy, and it demonstrates that price leadership is not the only viable retail strategy, provided the alternative advantage is genuine rather than a claim about service quality that customers cannot detect.

What should a CFO take from this?

That labour in customer-facing operations should be evaluated as an investment with measurable returns rather than as a cost to be minimised, and that the measurement system usually prevents this.

The practical exercise is to model total cost of a role including turnover: recruitment, training, the productivity ramp, error rates during the ramp, and the supervision burden. For roles with turnover above roughly a third annually, the fully loaded cost of a cheap employee frequently exceeds that of an expensive one.

The second point is that interdependent models require full commitment. Paying above market without devolving authority produces expensive employees doing prescribed work, which captures the cost and none of the benefit.

The third is patience. Retention benefits accumulate over years, so a model change of this kind produces higher costs immediately and returns gradually, which is why it is easier for a private company than for a listed one, a theme running through the governance pillar.

How does this model handle scale?

With difficulty, which is the honest limitation. Devolved decision-making works when the organisation can hire and develop store leaders capable of exercising judgement, and the constraint on expansion becomes people rather than property or capital.

That produces a slower growth rate than a centrally controlled format could achieve, and it is a deliberate trade. A store opened without a capable team damages the model in a way that a standardised format does not, because the standardised format does not depend on the team.

The practical mechanism is internal promotion. Companies running this model typically fill store leadership almost entirely from within, which requires a pipeline built years in advance and makes the training function a strategic capability rather than an administrative one.

The international dimension adds complexity, since employment law, wage norms and management culture differ, and a model built on autonomy and above-market pay must be recalibrated in every market rather than transplanted.

What does the model imply about performance measurement?

That store-level metrics should measure outcomes rather than compliance. A store team with genuine authority cannot be evaluated on execution of instructions, because the instructions are theirs.

The workable measures are availability, shrinkage, customer satisfaction, employee retention and profitability, with the store team choosing how to achieve them. That requires management comfort with variance in method, which most retail organisations lack.

The supporting requirement is transparency of data. Devolved decisions need good information, so store teams must see their own numbers in near real time rather than receiving them monthly through a regional layer.

The cultural point is that autonomy without accountability produces inconsistency and accountability without autonomy produces disengagement. The model requires both, and organisations that adopt one without the other conclude incorrectly that the model does not work.

How does this compare with the German training system?

It depends on it. The dual vocational system supplies retail apprentices with structured training combining workplace experience and formal schooling, which produces employees with genuine qualification rather than task-specific instruction.

A model built on capable, autonomous store teams needs that pipeline, and companies operating the model typically train far more apprentices than they need, treating the surplus as an investment in the labour market they draw from.

This is the same institutional foundation that supports the industrial specialists described in the hidden champions analysis, and it is the element that transfers least well to countries without an equivalent system.

Does the model survive online competition?

Better than most retail formats, because the categories involved combine routine replenishment with advice-led discovery, and physical stores serve both well.

Routine replenishment is genuinely contestable online, and subscription delivery services compete directly for it. What is harder to replicate is the discovery of a new product with staff advice, and the immediacy of buying something needed today.

The practical defence used across the sector is convenience of location combined with a click-and-collect capability, which uses the store estate as fulfilment infrastructure rather than treating online as a separate channel.

Frequently Asked Questions