Identifying Soda By Taste: Coke/Pepsi/Generic

Several years ago I had a roommate who insisted he would only drink Fuji and Dasani bottled water. All other brands were deemed inferior. I goaded him into making the statement that they tasted better. Then I purchased numerous brands of water, small glasses, and set up a blind taste test. After some deliberation he selected the cheapest competitor, Sam’s Choice spring water, as the best water (to my great delight). It made me think about the role of branding and marketing in product preference.

Last week one of my sisters made a statement about Coke being superior to Pepsi. Not content to let this declaration stand unchallenged, I decided to set up a soda identification test for some of my family. I purchased Coke, Pepsi, and a generic store brand cola called Super Chill. They were all regular colas sweetened with high fructose corn syrup.

The starting line up.

On Friday evening after dinner I set up the test. Prior to taking the test several people were quite confident that they would be able to identify the different sodas, especially the generic option. Some also claimed Coke and Pepsi would be easy to separate.

I gave each person six small Solo cups with a few inches of soda in them. Each person knew they were receiving two samples of each brand. The task was to identify what brand each sample was. On the bottom of the cup I had a discreet numerical indicator to allow me to determine if the answers were correct.

And this happened (rows show participant’s answers):

The soda identification results (correct answers in red).

After submitting answers no one felt confident about identifying the brands. The numbers show this, for the distribution of brands appears rather even and random.

Distribution of perceived identities of samples.

I found it interesting that only three people paired the same brand in their samples, and only one of those three pairs was correctly identified. It also amused me that my sister with the strong Coke preference identified both Coke samples as Pepsi and identified Pepsi and generic samples as Coke.

After the test taste I looked for differences in flavor. I tasted the samples with their identities known, looking for subtle differences in sweetness, carbonation, or scent. At times I could almost convince myself I found differences. Then I would shuffle the cups and completely fail to reorganize them based on taste and smell.

My Conclusions

Cola preference has more to do with branding and marketing than it does taste. I suspect the biggest factor in taste differences is the sweetener (such as real sugar versus high fructose corn syrup versus artificial sweeteners), so if you compare colas with a common sweetener they will taste the same.

I think it would be interesting to try variations of this test that involve participants receiving a calibration sample (one of each brand) and then multiple random samples of the brands to identify. They would need to not only determine the brand of the samples, but also how many of each brand they had.

It would also be interesting to involve ranking the samples in taste preference.

My Advice

Don’t drink soda (but that’s a different argument).

Here is a money saving tip. If you are hosting a party, buy one bottle of Coke and many bottles of generic store brand cola. Use glass or plastic pitchers to serve the soda, restaurant style. Leave the empty Coke bottle in a conspicuous place. People will see the empty Coke bottle and convince themselves they are drinking Coke.

Closing

So do you think you can distinguish between cola brands? Would you like to wager on it?