Big data: Creating the “oxygen for growth”
Introduction
The availability of so-called big data analysis has dramatically transformed many aspects of management practice, including the way retailers and producers have been running their businesses. Examples in the retail/FMCG (fast moving consumer goods) space include improving inventory management, supply chain efficiency, sales forecasting and price management. In many instances, big data analysis may help to better predict consumer behavior and to improve product offerings in order to maximise the return on each transaction in a store or at a specific sales location. The fundamentals of big data are based on new technological innovations (internet of things, cloud computing), socio-cultural evolutions (private information sharing) and opportunistic usage business innovation (loyalty schemes, etc.). In sum, there are many examples of big data applications that affect our business lives. Above all, perhaps, these effects might be particularly profound when it comes to impacting the discipline of marketing (Eevalls, S., Fukawa, N. and Swayne, L.), including finding new routes-to-market.
Cloud computing and big data analysis
As we have alluded to, we now have at our disposal entirely new analytics that can help revolutionize our understanding. While the relevant algorithms in general have been developed for some time, it is the availability of the capacity to analyze large sets of data that represents the break-through (see for instance Bodrev and Kamsaev).
With the near infinite storage capacity with cloud computing, there seems to be no limits to what companies can sort and share. Every 2 days we create as much information as we did from the dawn of civilization up until 2003! (Eric Schmidt, CEO Google). Today’s data centers occupy an area of land equal in size to almost 6000 football fields – (Bernard Marr, big data) – and, the areas of application seem to be endless. For instance Microsoft Cloud features the analysis of large sets of golf shots data, to gain new insights regarding what might constitute optimal golf shots in various circumstances! We see applications in such diverse areas as Pharma & Health, automation, investments, banking, retailing, etc.
But how is the rise of big data affecting the way we manage our business? Are we in a position to use this new extraordinary flow of information and make it a simple, efficient tool to improve our business results? Can we make it simple? Can we, for instance, evolve from Big Data Analysis to Smart Data Analysis?
Let us in particular discuss in some more detail two relevant areas of marketing, namely Pricing as well as Campaign Optimizing, before we briefly point out some limitations regarding big data.
Pricing
Price management is the essential part of Revenue Growth Management (RGM). The capability to generate margins is the real Oxygen for growth.
Margins are a consequence of costs and revenue management, and, in our specific case, big data helps increasing sustainable revenues, through prices management in a very spectacular way.
We are all familiar with the notion of Dynamic Marketing. Dynamic Marketing is in many ways the end of strategic planning, as today’s world offers the capability to correct and change our plans and actions very quickly: from grabbing attention to holding attention, from price promoting to instant price management. The notion of dynamic pricing is one that is very exciting today in order to optimize sales and margins. Based on big data analytics it helps companies to simultaneously targeting different prices to different customers or shoppers. In fact, dynamic pricing is based on customer’s perception of value. Machine learning algorithms automatically spot prices that on average can bring an extra 7% total margins compared to manual price selection.
Airlines are an excellent example. They are able to segment a plane accurately to extract different price offers based on where you sit of which class you are booked in, but also based on the timing of your booking, your personal profile, the competitive offer ad many other relevant criteria’s. The objective is of course to maximize margins by seat and optimize overall results of a plane as a business model. Amazon is another good example. Amazon changes its prices more than 2,5 million times a day in order to capture more transactions and to optimize margins. The fashion industry has also become highly performing in retailing, through managing prices in a matrix model combining timing and shopper segments (fashion fans, regular customers and price sensitive customers). Zara has built its entire business model in its capacity to entirely renew its offer every month. Big retailers are the champions of predictive consumer behavior and price management.
Optimizing
Data analytics helps in three critical domains. Predicting, Understand and Optimizing. We recently worked with a leading beer company, which started using big data to perform better within the retailers’ 4 walls. Their initiative had clearly a revenue management focus. Data analysis allowed us to review past experience in promoted volumes and to predict future results accurately. This was essential for a better understanding of price/promotions activities, optimizing brands and packages offerings and taking corrective actions when and where necessary. This kind of action cannot be created in a silo, ignoring the competitive landscape. Here data analytics helped also predicting competitor’s reactions and their potential effects. For the purists, we managed to do a conjoint analysis on a big scale.
Analysis of shopper behavior is a very important element in marketing today. Data analytics will here help responding to 3 important questions: why do shoppers shop, where do they shop, and how do they shop. Retailers will use techniques such as eye tracing, purchase decisions analytics, traffic analytics to understand better the behavior in the store. In the case of the producer, the key will be to understand the shopping occasions, evaluate their size and potential business impact and prepare the right offer to be captured. What key drivers started the shopping process, what sources of information or influence where consulted, what was the structure of the purchase trip (shopper mission), who influenced in store, what did the shopper do after the purchase?
Analytics will be essential to understand the traffic builders (advertising, digital, WOM…) the basket builders (displays, packaging, on pack promotions…) and the loyalty builders (coupons, loyalty cards, rebates…)!
Some limitations
So what about the places and businesses where big data analytics are not available? Here, I am thinking here about more traditional sectors such as fragmented channels – small retail, moms and pops, bars, restaurants, pharmacies, etc. These sectors are still very important and generally speaking generate more margins than bigger retailers. Data is not readily available, or there are no good methods to capture them. The name of the game in such cases is to combine big data to understand the path to purchase, the pre purchase and the post purchase, and to combine them with more traditional data collection methods in the store (small data). We have for example developed techniques in China to identify and localize the best potential outlets through geo-localization crossed with social networks data.
Conclusions
Clearly big data is affecting our business lives in a massive way today, and we just have touched the surface of what it will bring to business. In retail, we will see the rise of artificial intelligence as shopping assistants, of innovative screens and services and the development of store analytics in real time. Stores will be completely “datafied”, meaning intelligent, and retail spaces will be interactive and responsive. The traditional economy will also be affected. Today, techniques are being developed to merge big data and small data in order to offer producers and traditional outlets a good capacity to manage their future and to optimize their operations. Geo-localization and social network data are essential for these traditional sectors. All of this analysis for one key purpose as far as I am concerned: generating Oxygen for growth.
References and further reading:
Bodrev, A.A., Kamsaev, V.M. (2015). Modern Big Data Analysis Technology. Fundamental Research, pp. 5295-5299.
Brown, B., Chui, M. and Manzika, J. (2011). Are You Ready for the Era of Big Data?. McKinsey Quarterly, 4(1), pp. 24-35.
Duchesei, P. and Laurin, E.J. M. (2013). Decision Tree Models in Profiting Ski Resorts. Expert Systems with Applications, pp. 5822-5829.
Eevells, S., Fukuma, N. and Swayne L. (2016). Big Data Consumer Analytics and the Transformation of Marketing. Journal of Business Research, pp. 899-904.
Laenzliger, M. and Lorange, P. (2017). New Routes-to-Market at Migrolino. Unpublished working paper.
