Burger Chains Can’t Beat Coffee Shop Success in Unique Branding

Burger Chains Can’t Beat Coffee Shop Success in Unique Branding

Who’s Loving Who?

In my previous two posts[1][2], I looked at the impact of different strategies based on geography and type of marketing in the Quick Serve Restaurant (QSR) space. Three of the largest chains in Canada (Starbucks, McDonalds, and Tim Hortons) were using very different strategies. In those posts, I focused exclusively on their differences in technique.

Another question that arose from this analysis is whether these different strategies actually help to separate the brands in people’s minds. Do people consider Starbucks different from Tim Hortons, or are they both simply places to get coffee? Are McDonalds customers also patrons of other QSRs, like Burger King? In this post, I will answer some of these questions using cross-mention analysis.

Cross-Mention Analysis

In August 2013, I returned to my trusty Twitter sample[3] consisting of 14,195,003 Tweets from 57,867 Canadians between January 15, 2013 – May 31, 2013. I analyzed the Tweets for mentions (including common nicknames and aliases) of Tim Hortons, Starbucks, McDonalds, and many other Canadian QSRs.

Over the entire period, I collected a set of unique authors for each QSR. I then took the intersection of these sets to determine which authors discuss both brands. Figure 1 plots the pair-wise overlap of mentions for McDonalds, Tim Hortons, Starbucks and Burger King.

cross_mentionsThe Raw Numbers

First, look at the cross-mentions for Starbucks, McDonalds and Tim Hortons (top three graphs of Figure 1). Each have about 25% overlapping mentions with the others: a quarter of authors mention both brands. 75% of authors do not mention one of the competitors. Does this indicate that the audience for each of these brands is well segmented with minimal cross over?


Figure 1: Cross Mentions for Starbucks, McDonalds, Tim Hortons and Burger King

Looking at Burger King reveals a more interesting trend. Here we can see a large asymmetry. Over 50% of Burger Kings’ audience also mention McDonalds, compared with 10% of McDonalds’ audience referencing Burger King. This raises the question whether Burger King is seen as a substitute for McDonalds, without much clear differentiation between the two brands. To find out, I analyzed the Tweets behind these cross-mentions.

Text Analysis

To get more of a sense of what people are talking about when they mention rival QSRs, I looked at which terms are most correlated with mention pairs (Table 1). In the left-hand column are the tweets from “fans” of the QSR and in the top row is the cross-mention term. Even though the cross-mention posts are symmetrical (the same posts will appear in both sets), the terms that are used in the remaining posts are different. This leads to uncovering different sets of correlated terms depending on which QSR corpus is being analyzed.

McDonaldsStarbucksTim Hortons
McDonaldstoppled dispensaryforeign derogatory globeandmail
Starbuckstoppled 7eleven8am 3pm globe pipeline 20oz
Tim HortonsWalmart derogatory intoxicated#ftw digital Quiznos

Table 1: Top correlated terms

Overlapping mentions for these three QSRs are rarely a direct comparison of the product or brand. Many are tweets that are grouping the QSR into categories through a third entity (7-Eleven, Walmart, Quiznos), referencing news stories that mention both QSRs (Globe, Globe and Mail, toppled), or terms common to the industry (20oz, 8am3pm). Some notable examples of common tweets that directly compare restaurants are #ftw (= For the Win):

Never a line at Starbucks and their coffee doesn’t taste like toilet water Tim Hortons #yeg” Starbucks #ftw

or pipeline:

I prefer Tim Horton’s to Starbucks. Please don’t read that as support for the Keystone XL pipeline.

Direct comparisons seem limited to Starbucks and Tim Hortons, indicating that Canadian consumers see these brands as most in direct competitions. Interestingly, neither of these two brands is seen as a direct replacement for the other. Instead, when cited together, they appear to be polarizing opposites.

Applying the same analysis to Burger King and McDonalds, which had a significantly skewed cross-mention analysis, yields very different results (Table 2). Analyzing the text from the Burger King corpus reveals that word associations are indicative of a replacement product. Most of the terms are from tweets treating McDonalds and Burger King as synonymous:

I’ve never seen a McDonalds or a Burger King under construction. They just show up out of no where

@mcdonalds @burgerking Love McDonalds Pizza. Crumbly crust.

Burger KingMcDonalds
Burger KingLovin’ construction crumbly Popey

Table 2: Burger King and McDonalds Cross Mentions

Looking at tweets from the McDonalds fans provides a very different perspective. The most significant cross-mention terms all refer to the news event where Burger King’s Twitter account was hacked and replaced with a McDonald’s logo. This asymmetry in overlapping mentions mirrors the discrepancy from the cross-mention analysis. Burger King fans tend to view McDonalds as a replacement brand, on equal footing with Burger King. Conversely, McDonald’s fans do not feel the same way about Burger King. For McDonald’s consumers, Burger King is only significant in the context of current events.

Featured Image courtesy of Wikimedia Commons, the free media repository


Footnotes    (↵ returns to text)

  1. http://xplane.us/where-you-live-impacts-where-you-go-for-coffee/
  2. http://xplane.us/mcdonalds-finds-a-little-free-coffee-goes-a-long-way/
  3. K. White, J. Li, N. Japkowicz, “Sampling Online Social Networks Using Coupling From the Past,” 2012 IEEE 12th International Conference on Data Mining Workshop on Data Mining in Networks, pp. 266-27

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.