Recently, I’ve been answering a lot of questions about about RecallCheck, which was an Android based mobile application to allow uses to scan UPC bar codes on packaged food and find out about any recent recalls involving that product. Developing and marketing RecallCheck, I worked with my business partner Scott. With the project is now over, I’d like to cover some of the learnings and reasons I believe we were ultimately unsuccessful.
In February 2009 when we began our project the news was filled with stories about Peanut Corporation of America (PCA) salmonella outbreak. The FDA would eventually list nearly 4,000 separate products (with peanuts as an ingredient) as recalled.
The problem with the recalls was that it was difficult for a consumer to know if the jar of peanut butter on their shelf had been recalled. For most, the procedure was they would have to first find out from the press or web that there was a recall for products involving peanuts. They would have to find all the peanut containing products that might be in their kitchen. Once they had rounded them all up, a Google or other web search would be necessary to determine if that specific jar of peanut butter had been recalled. Of course, some people could just through them away.
RecallCheck was intended to simplify that process for consumers. With this app, they only had to scan the bar code with the camera on their Android phone and the phone would check the bar code (also known as UPC) against our database over the Internet, and let them know whether the product was recalled. If there was a recall, it would present all the identifying information and a link to the specific recall notice. Given all the attention recalls were getting, we thought a lot of people would be interested in the product.
Boy were we wrong. All told we sold less than 100 for $5 (and then minus Google’s vig of 30% for $3.5 to us).
The reason we didn’t do better is that despite all of the news reports about food recalls, they actually impacted very, very few companies or consumers in the USA. I’m going to present two pieces of evidence to support my assertion.
After we had compiled a years worth of data (2009) we thought we might put together a research report (available for sale here). As part of this report, we compared the number of firms involved in making recalls with the number of firms in various food related categories in the 2007 US Business Census.
The graphic is useful because it visually shows that fewer than 2% of the firms involved in food manufacturing, wholesale, or retail initiated a recall in 2009. This includes the frims behind the 4000 products listed by the FDA in the peanut recall.
A more recent piece of evidence comes from the US Center for Disease Control (CDC). One of the principal tasks of the CDC is gathering and publishing statistics on various forms of illness. While they don’t break out salmonella statistics by source (e.g. packaged food, restaurant, poultry) they do give an overall estimate of salmonella cases per 100,000 residents.
If we take the CDC’s estimate of 15.19 per 100,000 for 2009 and compare it with 2009 population (305 million) we get an estimate of 46,000 cases of Salmonella. It is useful to note that this is down from the year before. Without minimizing the suffering of those who are sick, this is a very, very small percentage of the US population. And keep in mind, the CDC numbers are for everyone who got Salmonella, regardless of source. Some of those ill will have gotten it from a source other than packaged food.
While I feel a little foolish for not having understood the inflated nature of the recall threat in the media, I did learn a great deal about business, development, and marketing that I hope to deploy in my next position.