The truth about the beer and diaper story

When data was just emerging, I often heard the story of beer and diapers from manufacturers or big data advocates, vividly and vividly illustrating the unlimited business opportunities brought by big data analysis to business. If you haven't heard it, let's say it again:
"Diapers and beer may sound irrelevant, but this is the result of mining historical data and reflects the laws at the data level. This relationship makes people feel Confused, is this a real pattern? After a follow-up investigation, the researchers finally found out that there is a reason: Some young dads often go to the supermarket to buy baby diapers, and 30% to 40% of new dads drop by to buy a beer to treat themselves. Walmart Beer and diapers were then bundled together and, unsurprisingly, sales both increased.”
—From Tu Zipei’s “Big Data: The Coming Data Revolution and How It’s Changing Government, Business, and Our Lives”
First time When I heard this story, I was actually confused. According to the author's limited understanding of American society, the laws of many states in the United States do not allow supermarkets to sell alcohol products. Only in recent years have some states passed laws allowing supermarkets to apply for alcohol sales licenses. But at that time, attracted by the upsurge of big data, there was no time to delve into it. I recently participated in some events related to big data, and this story is endless. Finally decided to explore the source of this story.
It turns out that this story has also been sung in European and American countries, and the versions that appear are not the same. But there are also many voices questioning this. Among them, a professor named Daniel J. Power wrote an article here, which should be the most comprehensive and justified to reveal the truth behind this story. Interested students can read the full text in English: What is the "true story" about using data mining to identify a relation between sales of beer and diapers? (http://www.dssresources.com/newsletters/66.php). The keynote reads:
· The chain is not Walmart, but Osco Drugs, a pharmacy and convenience store operating primarily in the Midwest.
· The company started a data mining project in 1992 and is considered a pioneer in data mining.
· The project at the time collected information from more than 12 million shopping carts from 25 stores.
· The data did find: Between 5pm and 7pm, customers bought beer and diapers.
· Store management did express interest in this finding. But in reality, there is no redesign of the product placement, and no bundled sales of beer and diapers. However, the company conservatively reworked its sales process to put the right item in the right amount at the right time.
That's the truth of the beer and diapers story, so it's not really a good story for data mining. Everyone talked about this story with great interest, and it was nothing more than a wish to express: data mining technology is a good helper for business decision-making! (Author: Shirley Xie Source: TechTarget China)