The primary factor used in the model, which is the most important one in defining the price, is the clients’ behaviour. Specifically, it is the demand curve, which defines the interdependency between the sold volume and the price as well as other factors. However, as the market situation changes steadily, it is essential for the business to respond to the changes as soon as possible, and to adjust the prices of goods by various deviations and trends.

How does this work in practice? First of all, large data volumes are processed by a computer algorithm. Subsequently, the prices are changed by the algorithm depending on the given e-shop customers’ behaviour. The data taken into consideration are not only related to the previous day, but represent a summary of numerous data for a number of weeks, months or even years. As such, the aggregated data are independent of deviations which ensures more stable results. Conversely, for the shops this means greater margin.

What types of data need to be gathered? Typically, these data include sales at the level of demand, that is, who bought what, how many items, in what colour and at what price. Based on these data types, future demand can be forecasted. Another data type taken into account are the competitors’ prices. Here, the price comparisons are made via various search engines. Last but not least, the relevant website data, such as website attendance and the number of clicks per product, are assessed. Naturally, also the supplier-related data, such as the acquisition price, the margin, and the like, are taken into consideration.

Can these data also be used for other purposes? The volumes of data gathered to define optimal prices can also be used for marketing purposes. As such, companies can address customers in a personalised way and adjust discounts and other special offerings according to the customers’ preferences. Based on the data, companies know what is worth investing in and what not.

What is the ideal frequency for price changes? Once a week. However, it is certainly possible to make the changes on a daily basis, and even in real time ‒ for instance, Amazon can change the prices every ten minutes.

Is the price going to be of key importance also in the future? More and more, the competitive advantage is going to be about the customer service and the added value an e-shop is able to offer to its customers. In particular, this means loyalty programmes, free shipping, express delivery, interesting website design or user-friendliness in terms of the easiest way of purchasing (such as a pre-prepared shopping basket, which includes products a customer usually buys). As a result, the price does not have to be decisive.

Can this method also be applied in brick-and-mortar shops? If electronic price tags are available, then the pricing changes are also possible here. What matters here is how often traditional shops reprice their products. Also, finding a balance for the frequency at which repricing is necessary is of great importance.

Are you still asking yourself whether the pricing optimisation is really necessary?
You should bear in mind that:

• The market conditions change quickly;
• You don’t always have to battle your competitors;
• The ideal compromise between conversion and the margin exists;
• Prices of a great number of products cannot be processed manually; and
• The time spent on all-day repricing can be used more efficiently.

What if all shops use dynamic pricing, which will result in the same prices everywhere? What will then be the distinctive feature for the customers? Even if selling the same product, the final price can be different in various e-shops, as every e-shop has a different target customer portfolio.

So what is the most important step to take on the way to dynamic pricing? That all employees in a company find their way to the tool. As new technology goes hand in hand with the simplification of work, at the same time it disrupts established processes. Therefore, it is necessary to persuade your IT department that the external code is not harmful. On the other hand, your suppliers need to be told that even if the sales price may be one crown higher above the market value, this does not necessarily have to be disadvantageous. Moreover, your sales department is to be persuaded that a fixed price is not an ideal solution and that in the end, the results attained all across the portfolio are part of one whole.

The strength of your data will manifest itself the moment you are able to use it to your advantage. At Deloitte Advanced Analytics, we use machine learning, predictive modelling and other statistical methods, advanced data visualisation or text mining to strengthen the role of data.