Pricing strategies have always been at the heart of all retailers’ operations. Prices are set aiming to maximise profit. We understand from price elasticity of demand that if you decrease price, you can most certainly expect to see an increase in sales volumes (other than the luxury product of course where increase price doesn’t damage demand until no one can afford it anymore). If 1% change in price yields a >1% change in sales volume, we say it’s elastic, otherwise inelastic.

However, in reality, at some point the marginal demand change will be diminished as the prices continue to drop. I.e. they go from elastic to inelastic. In addition, when we account for margin, as prices decrease, margin will also decrease. So given the stock level, there is a point of sales volume and margin where we can optimise profit and minimise stock level – like the graph demonstrated below:

To find out the optimal price is difficult for 3 reasons:
1) The price elasticity varies product by product. For example, people might be less sensitive to a 5p decrease of a £1 product than to a £10 change of a £200 product despite both being a 5% price decrease.
2) It varies from time to time. E.g. customers are more likely to pick up a meal deal for 2 in the evening just before dinner than other time of the day! How might that change their attitude towards price?
3) It varies customer by customer. What is a customer’s price sensitivity towards top quality yogurt if she is to buy them for her kids? Different customer segments will react differently to prices.

Essentially, there is a different optimal price for each customer at each time point when considering a particular product. Therefore dynamic pricing is needed to meet potential optimal prices. Optimising prices requires processing and analysing large scale of sales datasets with cutting edge analytics techniques to ensure accurate decisions on a continuous time scale.

Once you have the tools to support pricing decisions, the next step is to execute them. In brick and mortar stores, it’s almost impossible to distinguish price by customer. To make frequent price changes on product level and in different time zones are also proven to be difficult because of large amount of staff time required. Tesco trialled electronic shelf pricing back in 2012 and this can potentially become a big leap forward in realising dynamic pricing in store.

In the online world, this becomes a completely different story. The availability of internet has enabled infinite flexibility towards change. Prices can be altered at any time effortlessly. The unique view of every customer allows you to offer them personalised prices. Customers online browsing behaviour are also much easier to be tracked comparing to when they are in store. This, in turn, can further assist retailers’ pricing decisions.

Amazon, the world’s leading online retailer, uses ‘robot pricing’ – a high frequency algorithm driven technology to set their optimal prices. Most recently, they changed as many as 2.5m prices within a month. It’s not just Amazon, many other leading retailers are gradually catching up too, for example retail giant Best Buy offers data driven price recommendations but also allow their store managers to make their own decisions to achieve better results.

A full scale dynamic pricing system is still very much a rare thing in the world of retail. And it might not be suitable for everyone. In the multi-channel offering scenario, it’s unclear how more frequent changes in price online comparing to in-store will impact customer perceptions. However, a data-driven dynamic pricing system to some extent will undoubtedly help retailers better manage their pricing strategies and reduce ineffective efforts.