A look into how Coca-Cola is using AI to best adapt to consumer preferences.

Coca-Cola has a problem few other fizzy drinks firms face.Because they now supply in markets all over the world they see that there are local differences in preferred quantities of flavours, sugar and calorie contents.As Coca-Cola expand they may appeal less to consumers in the new markets as they will not have the same information relative to the local competitors who have had time to accustom to consumers tastes and preferences.

Coca-Cola plans on tackling this problem using AI.In order to train any ML model some data is required. In this case Coca-Cola used social media to better understand specific information for different areas regarding where,why and how consumers were consuming Coca-Cola .To do this, Coca-Cola analysed over 120,000 pieces of social media content to understand the demographics and behavior of its customers and why they were discussing the products.Furthermore to better analyse this data the firm has set up 37 social centres to gather and go through data. After the centres work out what posts create the most positive attention, they can then work on creating more of that type of content in order to promote the brand and compete in other markets.

A look at Coca-Cola’s popularity on social media, via Statistica.

A look at Coca-Cola’s popularity on social media, via Statistica.

Another large source of information is simply what choices consumers have previously made using the relatively recent electronic vending machines where consumers can customise flavours.There are millions of these vending machines around the world meaning Coca-Cola has large amounts of regional data on preferred flavours and other local preferences.

Coca-Cola’s new modern vending machine

Coca-Cola’s new modern vending machine

All this information is then used in AI systems.This may be in small level appliances such as in individual vending machines or to effect large scale operations, such as how the formula for drinks should be for each region.Another interesting application of AI was in the technology required to process images in order to enter codes for loyalty/reward schemes.This increased the amount of people enrolling in these schemes simply because of how easy it is to do it.

Due to Coca-Cola ending up knowing the same (or more) as their local competition regarding the consumers exact tastes and preferences, in the united states this was manifested in the launching of the “Cherry Sprite” as these techniques uncovered a demand for it, which was not matched by any supply. Due to the easier to use loyalty systems and increased targeted advertising on social media they have also seen favourable shifts in consumers tastes and preferences for Coca-Cola.

A look at Coca-Cola ‘s ever growing popularity, via  https://www.statista.com/statistics/235593/us-regularly-purchased-soda-brands/  (Statistica)

A look at Coca-Cola ‘s ever growing popularity, via https://www.statista.com/statistics/235593/us-regularly-purchased-soda-brands/ (Statistica)

We can expect Coca-Cola as well as other similar firms in similar industries to take advantage of the AI revolution and expect similar results, as AI opens new doors and could be argued to lead to an increase in consumer sovereignty as firms start to track more specifically what consumers exact wants are however ti could also be said that the targeted advertising that may come along with the AI revolution will mean that producers may have more of an effect on what consumers buy that the consumers themselves