The models’ limitations
Each of these four models can contribute significantly to the price definition, but are such prices the right prices and do these models rightly respond to retailer’s needs? Each model has its limitations, so using only one of these models does not lead to the desired right or optimum pricing for retailers.
The ‘cost-plus’ model risks triggering a vicious circle: costs determine price, price determines volumes, volumes determine costs. Moreover, fixed costs cannot be allocated. Retailers would face the problem of defining the right criteria for allocating significant operating costs to thousands of SKUs. The model is also blind to external factors.
The ‘value-based’ model has the problem that defining the perceived consumer value is not an easy job and is usually expensive. Furthermore, it is the marketeer’s job to increase the perceived value rather than setting the price to match it.
The ‘competition-driven’ model looks at the market but is risky because of its short-term view, its blindness to costs and customer value and because it could start a negative loop leading to a final price war, ultimately resulting in the erosion of margins.
The ‘optimum pricing’ model is too theoretic and based on assumptions which are beyond reality: perfect information, rational behaviour, and certain market conditions. It seems to be a passive model where all the drivers are given and does not explain how to really create value. Furthermore and most importantly, this model lacks self determination and is not proactive. The optimal price is a given, the model does not explain how to increase profitability.

Complexity and peculiarity
Given these limitations and the complexity and peculiarities of modern grocery retailing, standard pricing models do not provide retailers with exactly the right insight needed to improve pricing decision making. In some industries there is a certain level of autonomy in managing price, which happens independently from other marketing levers. However, in grocery retailing price positioning results from the sum of base prices, promotion intensity and assortment choices. Price is not a ‘stand-alone’ lever. Instead it is strictly synergic to the other marketing Ps (product, placement and promotion).
This makes it difficult for retailers to adopt certain pricing models. For the ‘value-based’ model it becomes more difficult to define the perceived consumer value: continuous consumer expectations for EDLP or HI-LO pricing could in fact impose too low a starting point in price definition to enable consistent profits. When it comes to the ‘competition-driven’ model, just following competition is not enough. Prices cannot be defined omitting the price-product relationship (positioning through assortment choices) and the price-promotion relationship (positioning on basic prices versus average prices including promotions). In the ‘optimum pricing’ model, the key relationship is price-product-promotion (managing the margin mix through assortment, promotions and merchandising).
Another peculiarity to grocery retailers is the lack of information on costs of goods sold. This makes it difficult to comply with below-cost legislation and to assess pricing leverage (regarding the bandwidth in following competitors’ pricing moves). Insufficient information on actual purchase prices and lacking cost information, which prevents cost allocation at SKU level, hampers the operation ‘cost-plus’ model. The ‘value-based’ model is affected as application of different mark-ups and below-cost strategies are impossible due to lack of information. The ‘competition-driven’ model cannot function due to lacking cost data to assess real pricing leverage, and the ‘optimum pricing’ model is affected by staggered variable costs for retailers as each manufacturer uses its own pricing method. As a result a retailer should consider all these prices – which have an impact on the retailer’s variable costs – to define its own optimum price. This is both inefficient and cost intensive.
Finally, and perhaps the most important issue for grocery retailers: standard pricing models are designed for less complex situations. Retailers do not just determine a price per item but define prices based on their overall positioning – promotions strategy, category price range, SKU price, different price levels per geography or level of competition.

Arguments for pricing decisions
Empirical research into retail pricing recognises the fact that previous year’s prices are the most relevant factor for decision making. Perhaps because pricing decisions based on changes in purchase prices seem to matter little compared to last year’s prices. Figure 2 puts the arguments for pricing decisions into perspective.

The real issue is that among different factors, managers consider “last year’s prices” to be of minor influence on the company’s profits (Figure 3).  In this context, recycling and reviewing price and promotional plans become widespread because of:
• Success of previous promotion activities (without considering whether these are still appropriate or first-best option).
• “The-best-I-can-think-of” attitude, given the lack of accurate and accessible pricing information or analytical capability.
• Budgets aimed at maintaining the status quo.
• Fear of loss greater than the will to gain (and incentives based on previous years’ results).
• Decision anchoring (reluctance to see or admit passed failures).

 Processed-based and  systematic
None of the standard pricing models has all the answers needed for finding the right price. This, however, does not make these models redundant. Identifying a balanced mix of elements from each model, it is possible to define a price that best fits the retailer’s context and requirements.
Hence, retailers must start looking at pricing as a structured process, supported by pricing intelligence systems and pricing diagnostic. Today, retailers – and especially smaller companies – do not use these tools in a systematic way, at their loss.

In general, an initial path towards the right definition of price could start from all major relevant pricing variables:

• Store clusters and formats: the evaluation of customers’ needs in order to determine the store layout in terms of incidence of single categories and so on, could be a start to take into consideration the ‘value-based’ model.
• Category roles and guidelines: the category portfolio approach could be a start to take into consideration the ‘optimum pricing’ model in a more strategic context that aims at maximising profit per category.
• Assortment and promotion clusters: this could be a start to take into consideration the ‘cost-plus’ model.
• Competitor benchmark: this takes into consideration the ‘Competition driven’ model. 
The right approach for each retailer is to balance each variable of the retail process and the different models and to apply this in a structured way.